Starting Over: The Altamont and Blue Ridge 2
                                                                                Pete Caldwell and The Altamont and Blue Ridge 1

I began this blog as a way of collecting and organizing my thoughts regarding my proposed new layout, The Altamont and Blue Ridge2, and I plan to continue blogging into the construction phase on the project. Along the way, I will broach almost every aspect of state-of-the-art model railroad design, construction, and control.

Should you have questions or comments, you can email me at the following address:


From the the feedback I have received, I can see that many are finding my blog entires as they search the Internet for information on various model railroad topics. To aid these searchers as this blog gets longer and longer, I am including an Index that can be seen by clicking the following link:

Go To Index

Table of Contents / Links

July 2015
To Be or Not to Be?
What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
HO vs N Scale 7/17/2015
Further Thoughts Regarding Sound and Lighting 7-23-2015
Giving Some Thought to the Backdrop Wall Design 7/25/2015
August 2015
New Layout Design 8/5/2015
Planning the Flow of the Landscape: Tweaking the Track Plan 
Road Grading, Structure Pads, and Town Terracing: Preliminary Considerations 8-8-2015
The Wiring Scheme: Thinking About Detection Blocks 8/10/2015
The Naming of Power Districts, Detection Blocks and Turnouts 8/11/2015

What Do I Need 8-12-2015 (updated 9-20-2015)
Selecting Vendors 8/13/2015
Further Tweaking of the Track Plan 8/14/2015
Tweaking the Wiring Schematic 8/15/2015
More Work with Block and Turnout Names 8/17/2015
Doubts 8/18/2015
Block L
ists 8/19/2015
Turnout Lists 8/20/2015
Creating a Switchboard in TrainController 8/21/2015
Still More Track Plan Tweaking 8/22/12015
Thinking About the Yard 8/25/2015
Signal Placement and Terminal Track Plan 8/26/2015
Another Design 8/30/2015
September 2015
Further Considerations Arising from a Smaller Trainroom 9/1/2015
Still MoreTweaking of the Benchwork and Track Plan 9/6/2105
Accessing the Hidden Yard 9/7/2015
Chcking Elevations and Grades using XTrackCAD 9/9/2015
More Track Tweaks 9/12/2015
The Luxury of Time 9/18/2015
Rethinking the Removable Corner Mountains  and Resizing the Corner Radii 9/19/2015
Bridge, Viaduct, Over/Underpass, and Culvert Planning 9/21/2015
Road, Terrain, and Topography on the Revised Track Plan 9/22/2015
November 2015
Tortoise Interface 11/4/2015
PM42 and BDL68 Interfaces 11/10/2015
Moving Ahead 11/19/2015
Dwarf Signals 11/21/2015
Labels 11/26/2015
Reconsidering Twisting and Bundling of Power Buses - The Bus Wire Mounting Board 11/30/2015
December 2015

Wiring Color Scheme 12/1/2015
Some Thoughts Regarding Droppers 12/1/2015
Tortoise Interface Prototype 12/3/2015
Rethingking the Train Room Lighting Scheme 12/5/2015
More about the Phillips HUE lighting system 12/13/2015
Preliminary Testing of Phillips HUE System
Phillips Hue Program Up and Running 12/19/2015
More About Marrying the Layout to the Backdrop 12/31/2015
January 2016
Installing and Programming a Phillips Hue Lighting System on the AB&R1 1/9/2016
Architectural Drawing for the New A&BR2 Train Room 1/12/2016
Atlas or Peco Track? 1/13/2106
Designing and Locating the Main Lighting and Electrical Panel
Switching to Atlas Code 55 1/18/2016
Weathering Track 1/19/2016
Testing the Atlas Code 55 Turnouts 1/22/2016
Revisiting the Track Plan 1-24-2016
Reworking and Double Checking the Block and Turnout List Agaisnt the TC Switchboard  1/25/2106
Turnout Templates 1/29/2016
Tweaks to HUE Lighting Sunrise and Sunsets Completed 1/30/2016
Some Thoughts Regarding Cork Roadbed 1/31/2016
February 2016
An Easement Template 2/4/2016
Track Plan Revision at Fitzhugh 2/6/2016
Do-ahead Projects 2/10/2016
Revisiting the PM42 and BDL168 Barrier Block Labeing Scheme 2/11/2016
Relocating the PM42 and the BDL168 on the Layout 2-17-16
Mounting PM42s and BDL168s 2-18-16
Thinking About the Benchwork 2-19-2016
Update tothe Train Room Floor Plan Including the Under-the-Slab Conduit Locations 2-25-16
A Power Bus and a Prototype Mounting Board for DS64s 2-26-16
Modifications to the Tortoise Interface 2-29-2016
Layout Lighting Power Bus and Lighting Distribution Blocks
March 2016
Labeling the Wire Mounting Board 3-1-2016
Installing Turnout Interfaces 3-4-2016
Looking Ahead 3-17-2016
Aux Power Supply 3-22-2016
T Splicing Wire 3-23-2016
Signals and Signal Placement 3-24-2016

April 2016
A&BR2 Specifications 4-11-2016
TrainController Gold 4-10-2016
Servo Controlled Dimmer for Trainroom Fluorescent Lights 4-11-2016
Programming the Lights on the A&BR 4/14/2016
Some File Management Housekeeping for TrainController 4/23/2016

May 2016
TroubleShooting Loconet After a Lightning Hit 5-7-2016

DPM Kits
A DPM Kit Bash 5-24-1016
Windows 10 Problem vis a vis 7.1 Surround Sound Box (ENCAB-8CM) 5-31-2016
Layout Lighting Power Bus and Lighting Distribution Blocks 2-29-2016

June 2016

Prototyping Layout Lighting Distribution Blocks 6-1-2016
Beginning to Lay Track - Peparing to Construct Prefabricated Sections 6 -7- 2016
Weathering Turnouts 6-8-2016
Wiring Turnouts 6-9-2016
Having to Rethink DS64 Turnout Addresses 6-11-2016

Assigning Board Address Numbers 6-13-2016
Laying Track 6-19-2016         
Installing Tortoise Switch Machines and Tortoise Interfaces on the Prefab Section 6-20-2016
Testing Installed Atlas Code 55 Turnouts and the Prefab Section Wiring 6-22-2016
Rethinking the Layout Lighting Wiring Scheme 6-24-2016

July 2016
Signal Protocol, Programming  and Placement  7-4-2016
Breaking Ground 7-5-2016
Footings Poured Today 7-13-2016
Block Walls - Train Room Takes Shape

August 2016
Foundation Walls Complete 8-12-2016
Track Laying Using Xtrack CAD 8-17-2016

September 2016
Slab Poured  9-7-2016
Painting Figures 9-10-2016
Elevation Calculator 9-19-2016
Lighting a Covered Station Platform 9/23/2016
Floor Joists Installed 9/24/2016

October 2016
Contemplating the Completion of the New Building 10-30-2016
Train Room Lighting and Electrical Control Panel 10-31-2016

November 2016

Train Room Lighting and AC Electrical Revisited
Laying Out the Backdrop Walls and the Overhead Lighting Soffit 11/9/2016
Backdrop Wall Initial Framing Completed 11/11/2016
Final Train Room Lighitng and Electrical Plan 11-28-2016

December 2016
Lighting Soffit Framing Complete 12/1/2016
Final Tweaks to Block and Turnout Names and Addresses and to the Power District Layout 12/2/2016
Manipulating Color in HUE Light Programs 12-7-2016
Construction Blues 12-22-2016

January 2017

Wait Time Preperations 1-18-2017

February 2017

Train Room Takes Shape

Creating Signs

March 2017

Train Room Initial Wall Painting Complete
Painting Clouds on the Backdrop 3-20-2017
Rethinking Painted Clouds 3-25-2017
Working Drawings for the Hinged Bench Tops at Altamont 3-25-2017

April 2017

Working with Mountain Silhouettes on the Backdrop

Testing the HUE Lighting System in the New train Room 4-2-2017
Working Drawings for the Benchwork 4-3-2017
Further Thought Regarding Mountain Shilouettes 4-4-2017
Constructing a 4 Foot-long Mountain Silhouette and Checking the Scale 4-6-2017
Rethinking the Tree Line and Completing the Mountain Mockup 4-11-2017
Cloud Painting on the Backdrop Complete 4-13-12017
Test Section of the Bench Completed 4/18/2017
Basic Bench Work Completed 4/19/2017
Taking Stock and Planning the Next Move

JULY 2015
To Be or Not to Be?

If you have explored much of this website, you will probably have noticed that the A&BR is pretty much finished. Of course, these things are never really finished; still, all the track is laid; the bench-work is completely covered with highly detailed scenery;  all the lights, signals, and other refinements that I originally planned are in place and operational; and all of the the DCC control and sound hardware is humming along nicely with all of its software programed and tweaked. For the last year or so, I have busied myself with simply keeping things running smoothly.
Now, operating a model railroad may have its pleasures, but it is the planning and building of these things that really excites me.

Lately, I have been feeling the need to move on. I find myself dreaming of a bigger and better train room. As it happens, I am currently considering adding an addition to my home in the North Georgia mountains. There will be a large (approx 40' x 20') upstairs space to accommodate two bedrooms and a bath. Should I instead use this space for a new train room? Should I take on such an enormous project at my age? 

This is all conjecture at this point. Nonetheless, whether I build it or not, I find myself planning a new layout. This is a complex matter, and so I begin this blog as a way of organizing and documenting my thoughts. At the same time, it quickly becomes a tutorial detailing the kind of thinking that goes into planning and organizing this kind of project. If I continue to the actual building stage, this blog will also function as a detailed "how I build it" tutorial. But for now, it's just away of talking to myself.

I begin the planning process by asking myself lots of questions about the proposed layout. Perhaps the most intriguing of these questions is: 
"What would I change about my current layout if had the whole thing to do over?"

Begun in 2008, the current N Scale Altamont and Blue Ridge Railway occupies one bay of a remodeled two car garage. The train room is 21' x 12', with custom electrical, lighting, and HVAC systems. The DDC layout is controlled by Freiwald "TrainController" Software with train-tracking 4D Surround Sound. Current programed sequences operate ten trains at once in a complex 15 minute demonstration "train show". Do I dare tear this whole thing up and start over?

What Would I Do Differently?
I have been thinking about these questions for several months now - just kind of day dreaming, and playing little "what if" games. A few weeks ago, I decided to make a list of the the things I would do differently were I to do it all again. Now, mind you, I am generally quite happy with the current A&BR. I mean, there is a lot that is right with it, and at first blush, there didn't seem to be much wrong with it. To my surprise my little "do-over" list turned out to be quite long, and the more I think about, the longer it gets. This seems to me compelling. Perhaps I really should start over. 

There are a number of areas of focus in the "do-over" list:

1. Room Lighting. Currently I have dimable daylight fluorescent fixtures with UV bulb filters, all hidden from view in a sofit above the bench-work.I am delighted with this set up. The lighting is warm and bright, and it looks great, but there are a few flaws. 

As you can see in the drawing below, outside of the lighting sofit I have conventional track lighting - two dimable circuits. I originally planned to use these for a kind of hot-side/cool-side colored light to add depth and create twilight etc., but although the incandescent track light cans when angled toward the layout cast nice realistic shadows on the ground, they also cast shadows on the backdrop which is not a desirable effect. If I had it to do over, I would mount these lighting tracks inside the sofit where I could angle them only slightly and point them away from the backdrop. It is also possible that I really don't need these track lights at all.
lighting On the current A&BR I have track lights outside the sofit angled toward the layout. If I had it to do over I would raise these and mount them directly on the ceiling inside the sofit. Or perhaps omit them all together. The warm fluorescents look great and give plenty of light.

Another problem is that the fluorescent dimmers I use are not computer controllable. So I computer automate room lighting mechanically, moving the dimmer sliders up and down using servo motors controlled by a program I wrote for an Arduino microprocessor which executes dimming sequences triggered by inputs from stationary decoders. It's a little bit of a Rube Goldburg, but it works fine. Still, in an ideal world computer addressable fluorescent dimmers would be very cool. 

dimservos Receiving input from stationary decoders, Arduino controlled servos mechanically move room light dimmer sliders on the current A&BR. A bit of a  Rube Goldburg ... but it works fine.

Most likely, I'll control room lighting servos with dimming programs that I will write in C#. These little program will send string serial commands over a USB to a servo controller board like my SSC-32. The individual programs can be called from Traincontroller using automatically triggered push buttons using the "Execute" option in the each pushbutton's operations window. This way I will be able to address and operate multiple servos simultaneously. I have also developed a better way to mechanically move the dimmer faders with servos using small tension rods with clevises on the ends. More on this later.

Lastly, I would add individually switchable, built-in, work lights under the bench and in all behind-the-layout access aisles.

2. AC. I am pretty happy with my current AC wiring setup in general, but if had it to do over, I would add  more outlets and a master outlet on/off switch.

3.Control Desk. When I built the current A&BR I did not design in a control desk. This was a mistake. In any future setup, I would design a dedicated control area, with built-in desk for the computer key board, built-in computer monitors, loconet access, etc.; and I would also move all short circuit indicator lights from the PM42s to a built-in panel in the control area. 

4. HVAC. I would have a special particle-removing dust filter system like the ones used in hospitals etc. built into the HVAC air handler.

5. Separate Undetected Power Bus for Turnouts. I would employ a dedicated turnout track power bus. According to most TrainController Software users and TrainController creator Jurgen Freiwald, it is best in most cases not to have turnouts as part of detection blocks. Rather, they should receive power directly from their own power bus and be designated in the control software as part of undetected routes between blocks. On the current A&BR1, I incorporated a number of turnouts into my blocks. It's not really a problem, but it would be more elegant had I used a turnout bus and made them all routes. The fact is, I could have done a better job laying out the block scheme in general.

6. All of this aside, perhaps the single most glaring issue with the current A&BR is access - both to wiring and in some cases to track work.

Although quite functional, the wiring of the current A&BR is really pretty messy and hard to service. At age 70, I am not quite so inclined to go crawling around under the bench with a soldering iron as I once was, and I have a few tunnels and a hidden yard in which I would probably be unable to address any serious track laying issue owing to lack of adequate access. 
wiremess Wiring underneath the current A&BR works great, and I have experienced very few problems, but it is pretty chaotic looking,  poorly labeled, and difficult to service.

To address the track access issues, I am envisioning a space large enough to accommodate a sort of room within a room, which would feature a narrow access isle around three sides of the layout behind the backdrop, thus allowing open, behind-the backdrop access to all tunnels, especially those in the corners of the room.

To address neater more accessible wiring, I envision a system of recessed cable troughs running just below the fascia board of the bench-work to carry all bus wiring including 12 agw track-power buses, 12 volt layout lighting buses, as well as stationary decoder feeds and signal feeds. On the front of this trough, mounted horizontally for easy access, I propose a long narrow mounting board for stationary decoders, signal cards, other control cards, and the loconet cable. All 16 agw track feeders and 18 agw light feeders would connect to the power buses in the troughs and go out in parallel runs, front to back, through the back wall of the trough, underneath the bench-top to the appropriate track or light droppers. Decoder and signal outputs would likewise run in the trough and feed out to turnouts front to back parallel to secondary feeders. Loconet cable would run along the bottom of the mounting panel below the power cable trough to minimize crosstalk. (Note: I have ditched the idea of the troughs in lieu of different approach - see the post dated 11/30/2015)

trough Access Isle
New cable trough design. Trough will attach to bench legs and run below and slightly behind the fascia board of  the bench. It will be hidden by a black table skirt that attaches to the lower edge of the fascia board and goes all the way to the floor.

7. Structure Lighting. On the current A&BR, I used Christmas tree lights to light structures (gangs of 6 in series with the gangs wired in parallel all running off 15 volt, 5 amp buses). On the new layout, I would change to 5mm LEDs run from a single 5 volt bus. They are smaller and much easier to correctly position inside a structure, and they lend themselves to a much cleaner and more precise installation all around. 

8. Tortoise Interface. On the new layout, I also envision a new design for a tortoise switch machine interface that is clean and easy to install - just an eight position terminal block and a pigtail with a female edge connector leading to the tortoise. This is similar to what I now have, but I soldered directly to the tortoise and did not use the female edge connector which would have made things much neater. Also, the way I handled droppers everywhere on the current layout (and especially on the turnouts) is inherently messy. There is a better way, and I will not make this mistakes again. (Note: see post of 12/3/2015)

9. Individual On/Off Block Power Switches. Sometimes a short can be hard to find. On the A&BR2 I will install single pole sing throw on/off switches on the output of each block sensor. In the case of a hard-to-find short, I can then turn off the blocks one at a time until the short clears, thus quickly locating the shorted block. (Note: I later decided this is more trouble than it is worth,.)

More on these designs later, should I proceed with the project. But for now, I am dreaming of  troughs full of neat bus wires with feeders running out under the layout in perfect parallel lines to neat droppers for track, signals, and lights, and to compact tortoise input blocks - everything clean, tidy, well labeled, well lit - an installation so solid that I will never have have to crawl up under the bench again. Amen.  

Things I Would Not Change

Generally the basic bench work design I have used on the current A&BR is excellent. I use
a 1x6 framework on 2x4 legs  - sturdy enough to climb up on, cheap, and straight forward to build. I'll probably keep the bench height at 42.5 inches, but maybe I'll  raise it an inch or two. I want it high enough to force most viewers to look as much across the layout as down on it, but low enough for most kids to be able to see over the edge. Likewise, I will stick to limiting the bench edge-to-backdrop distance to 34 inches. I might go as deep as 36 inches if I don't raise the height, but I fear my aisles are already too narrow at about 2.5 feet. Three feet would be better. It all depends on how large the room turns out to be. Anyway 34 to 35 inches is about as far as I can comfortably reach. (Note: I later worked it out to allow 3 ft minimum for aisles and even wider in some palces. I think any think less is a big mistake. Also I eventually planned a 36" bench edge to backdrop dimension, which is a lot, but if I keep all track at least 4 inch out from the backdrop and use the extra 4 " to make nice transitions from scenery to backdrop, I think things will be the better for it.)

Likewise the lighting soft design for the current A&BR has worked quite well. Also the 48" sheetrock corner radii in the backdrop wall look great.  (See the Train Room Construction section or this web site for details on all of this).
On top of the bench work, I would construct the plywood and homosote support strips for open roadbed and grades the same way I have in the past, however, I would make them much wider with perhaps a little raised edge inside the tunnels to guard against locos and cars falling onto the floor after derailments, which, of course, only happen deep inside the tunnels. Also in the large space, I can have a larger a minimum turn radius, say 30", and a more gentle maximum grade, say, no more than 2%.. (Note: After the track plan was completed, this turned out to be closer to 28".)

I am delighted with  N scale Peco Code 55 flextrack and with Peco turnouts as well as with the highly reliable Tortoise Slow Motion Switch machines. As I mentioned above, I will make some changes in the Tortoise interfaces. I will also change the way I customize the Peco turnouts for DDC. More on these changes later. (Note: I later decided to go with all new Atlas Code 55 track and turnouts, see post of 1/18/2016.)

As to wiring, I will stay with 12 AGW track power buses, 16AGW feeders (max 3 feet) (
all insulated stranded wire), and 20 AGW solid copper insulated droppers (max 10 inches). Lighting wiring will be less robust, maybe 16 AGW stranded buses and 20 AGW feeders and droppers. Signal wiring for the Digtrax SE8c signal driver boards is accomplisted with 10 conductor ribbon cable (two 3 aspect double headed signal masts per cable).

I am totally married to Digitrax. It is true love; boosters, block detectors. stationary decoders, signal control cards, the works. Digitrax is a great little company, with great service, and great products at a fair price. For me, loconet or something similar is a critical aspect of a good DDC system. With loconet, I can use the digital signals carried on track power solely for communication with locomotives. Computer communication with everything else is handled by via loconet. Things just seem to work better that way. Also, just like all of my N scale trains and structures, I already have a full compliment of Digitrax DCC control devices, and I see no reason to change. The exceptions to my Digitrax loyalty might be mobile decoders and the computer USB interface, where I really like the TCS decoder line and the way they handle back EMF and the RR CirKits LocoBuffer.

Likewise I am in love with The Freiwald TrainController Software. I think Juergen Freiwald is something of a genius, and he takes a very hands-on approach to dealing with customer service, technical questions, and problems. The TrainController Software is so deep and so flexible! I remain in total awe of this product.

The RR CirKits Loco Buffer Serial interface works fine, and RR Cirkits is a nice little mom-and-pop company.

As for sound, I currently use the Friewald 4DSound software to drive a little "desktop" 7.1 Surround Sound system.
For me, low volume levels seem to work best in N scale, so no need for high fidelity. This means I can use a very inexpensive playback system, Indeed, my little "desktop" system works pretty well in my current 20' x 12' train room. The software's "train tracking" feature is really pretty remarkable, and I am reasonably happy with the moving sound scape produced by this setup. 4D works pretty well in a regularly shaped rectangular room, although there are some problems with  surround and the way it distributes sounds to speakers, especially when locos are in the center of the room. 

Also, 4D is somewhat tedious to set up and program. It seems I could spend the rest of my life tweaking sounds and programming events and volumes and sound placements etc. Right now, the current AB&R is, for the most part, constructed against the walls of the train room without much track extending into the center of the room. This is ideal for 4D sound.
For more details and a discussion of 4DSound on my current setup see the "Sound" and the "Power & Control" sections of the "Trains and Equipment" section of this web site and also the discussions that follow in this blog.

On the contemplated new A&BR2, I am concerned about surround's credibility in the center of the room where I will have a lot going on. I have programmed a few middle-of-the-room scenarios on my current 4D system to evaluate how trains traversing  the large center section of the new room up might sound. It very well may be that in-train decoder sound will work better in the new setup, but I am
and concerned about the difficulty of installation of N scale sound decoders and speakers and unclear as to how much the lack of space in N scale models will effect the sound quality.  I plan to do some experimenting with this before I begin the new layout, but at first glance getting speakers into a fleet of 12 or 15 N scale locos looks like it might be a handful. I am also exploring the idea of using more than one 4DSound playback system. (See the discussions that follow later in this blog.)

HO vs N Scale

The current N scale A&BR is my third model railroad. The first two were HO. So I have experience in both scales. As I contemplate the possibility of an all new A&BR2, the question of scale seems to be paramount. Which scale should I choose? 

There has been a great deal written on this subject, and generally both camps remain entrenched. Both scales have advantages and disadvantages, and there are a number of areas to compare. 

HO just feels good, and operationally it seems to me a bit more robust and stable in almost all regards. Trackwork, loco mechanics, wheels, trucks, and couplers are all less delicate, easier to work on, more sturdy, and a bit more forgiving. Still, under ideal circumstances, today's N scale models can be set up to run quite reliably. However, as we all know, circumstances are not always ideal, and it takes perfect trackwork and constant, jeweler-like precision to keep a large N Scale operation perfectly tweaked and operating without incident. For me, its more stable mechanical performance is the largest asset in HO's time-honored bank.

Another point in HO's favor is the way it behaves with DDC computer control. HO locomotives can be calibrated more precisely than N units, and this calibration will remain more stable in HO units. This is not a decoder or a software issue. N scale decodes are excellent, and most control software is essentially the same for both scales. The difference in performance is a function of the size and mechanics of the locos themselves and the stability of the ballistics of their motors. As we all know, speed calibrations are not absolute and they can change with temperature, lubrication, wear, and other friction issues. N scale models are simply much more sensitive to such changes, and,  although they can be tweaked to stop within a few inches of a measured distance into a block, their stopping accuracy will probably slide more than HO models with time and changes in circumstance. A perfectly calibrated HO model might consistently stop within an inch or so of a designated point and, all things being equal, it might hit that mark for months and months. Overtime, N scale models will probably remain pretty close, but they will be more likely to slide off the mark as things subtly change. Still  in most cases, they remain close enough for me.

If your are married to sound decoders and in-train sound chips and speakers, then, of course, HO is the clear choice. The modern N scale sound decoders are fine, but the on-board space limitations in N scale still present problems in my opinion. If I go with N scale, I'll continue to use Friewald Software's Traincontroller and the associated Freiwlad 4D surround sound. It may not be quite as versatile or believable as the decoder sound, but, hen set up and tweaked properly, it is pretty impressive. For me, it's the olny way to go for sound in N scale.

On the visual side, when modeling in HO, one needs a great deal of space to achieve any kind of over-all panoramic realism. Because of its size, individual "scenes" in HO can be much more detailed, however only a gargantuan train room allows enough space to get these little 3 dimensional compositions far enough apart to achieve anything believable in overview. In HO, in the average, largish train room (say, 20' x 12'), one tends to create crowded, pressed-together strings of perfect little vignettes, which, although individually realistic in close-up, somehow fail when viewed as whole. With such a forced compression of scenes, HO layouts seldom achieve a consistent, realistic, natural flow of the over-all landscape. What is more, if one limits the against-the-wall bench work to, say, 36 inches,  as I do, in HO it becomes very difficult to create the kind of false perspective needed to effectively hide the marriage of the backdrop to the bench top.

All of the above-discussed visual issues pretty much disappear in N scale. In fact for me, modeling in N has a completely different focus than modeling in HO. In HO, one tends to focus on individual scenes, whereas in N, one's principle focus tends to be the panoramic sweep of the landscape as a whole. None of this should be construed to imply that one can't create detailed scenes in N. As this web site I hope proves, one certainly can do detailed work in N, while at the same time, the diminutive size of N scale structures, trees, vehicles, figures etc. tends to discourage the kind of "rivet-counting" analism one sometimes encounters in larger scale modeling. In my mind, the ability to create convincing natural, sweeping, panoramic landscapes containing entire towns, large yards, and sprawling industrial developments, coupled with the ability to create the illusion of receding  distances near the backdrop make up by far the largest deposit in the N scale account. In fact, for me this is the game-breaker in N's favor. 

A 34 inch deep section of the current A&RB modeling a mountain river crossing. Notice in N Scale there is ample room for a roadway, a double track  behind. Notice the forced perspectivcrossing along side a single track crossing, and a distant "high line" crossinge achieved by using progressively smaller trees beyond the high line track to the left and the z scale house on the hilltop to the right. Achieving this kind of panoramic result in a front-to-back space of only 34 inches would be difficult if not impossible in HO. 

To be sure the cost should be considered. At first blush, N scale would appear quite a bit less expensive. The cost of N scale structures, track, turnouts, and locomotives can be as little as half that of corresponding HO products. But remember, working in N, you will have four times the in-scale space to fill with track, structures and locos. So the cost issue is something of a wash. In my case, a large point in N's favor is the fact that I already have 50+ turnouts, 15 Southern Railway and L&N locomotives, 80+ units of rolling stock, 150+ fully detailed structures and so on. I suppose I could sell this stuff, but I doubt I could get anything near 50 cents on the dollar for it. So, this too sways me in the direction of staying with N scale. Frankly, I am really looking forward to building the train room and bench work, laying the track, wiring, and creating the landscape. I am not looking forward the assembling, painting, weathering, and detailing 80 new box cars, and 150 new structures.

Not too many years ago, one of the main arguments against N was the notion that there was just not as much stuff available for N scale as there was for HO - not as many loco types and models and roadnames, not a full array of structures, not enough other toys and whistles and bells, so to speak. As we all know, those days are long-gone. N scale catalogs overflow with choices, and a number of decoder manufacturers scramble to create full function, easy-to-install, drop-in boards for myriad of loco types. Perhaps N still suffers from standardization issues between Europe and the US, and problems that arise from lack of interchangeability among several different axle lengths and wheel sizes, track codes etc., but these issues are addressable. Indeed,
N scale has come of age, and it may very well represent the future of the hobby.

At this point, I have not completely settled on either scale for the A&BR2, but I am strongly leaning toward staying with N. I have completed two layout designs for the proposed 40' x 20' space, one N and one HO, and my leanings still favor N. In these preliminary designs, 
I have a series of connected benches the longest of which is over 35 feet in length. That's over a mile in N scale!

Ndesign N Scale Preliminary Design:
Double track mainline all around with one reversing loop, large yard, passenger terminal, 3 depots - very similar to my current track plan, but the large train room allows for behind-the-backdrop access isles for access to tunnels on 3 sides, much longer side-wall runs, larger radii, and, of course, the addition of the massive 5.5 foot wide 25 foot long center section. 

Further Thoughts Regarding Sound and Lighting.
As I mentioned above, I am concerned about the use of surround sound for the large section of the layout that will be in the center of the room. Surround Sound works well for the sections of the layout that hug the walls of the room and for listeners in the center of the room, but for sound for sections of the layout in the center of the room, it brings a mix of many speakers into play, and for listeners who do not happen to be in the exact center of the room this may cause some problems.


One solution would be to have a second 4D Sound system set up to feed a different set of speakers mounted beneath the bench work of the center section. Then I could create a separate soundscape for the center section that would not bring any peripheral speakers into play. Whether or not this within is the  capabilities of the Friewald 4DSound software is a question I have put to Jurgen Friewald. I am awaiting his reply. I know that TrainController can work with a network of multiple computers each of which can be set up to simultaneously control the entire layout or different parts of the layout; so perhaps this will be do-able.

I have also given some further thought to the lighting above this large center section. I plan to use two banks of daylight flourescent fixutes, one running along each side of the sofit above the 5 1/2 foot wide center section. The fixtures cloesest to a viewer will be hidden from his or her view by the sofit just as they are in the sofits above the sections of the bench that are attached to the backdrop wall, but unless I lower the bottom of the soffit considerably, the light fixutres on the side away from the viewer will be visible. I could add a center two-sided backdrop, but I don't think I want to do that because it would effectivly divide the trainroom in half, and I feel it is more spectacular to be able to see the entire layout all at once. Still, I am keeping and open mind. The effect might be better with a central backdrop. I'll probably only be able to get a good feel for this when the room is near completion; in the meantime, I think the best solution here is going to be an additional central sofit wall to hide the opposing fixtures. (See Drawing Below)

centersofit The additon of a center soffit wall will effectivly hide the  opposing flourescent light fixture.

Giving Some Thought to the Backdrop Wall Design

As discussed above, in the new 40' x 20' space I am envisioning a kind of "room within a room" design, which will feature a 2.5 foot wide access asile around three sides of the layout.
The design of the stud wall that will separate these access aisles from the train room and will support one side the the bench and carry sheet rock for the backdrop is on my mind today.


I am seeing a floor to cieling backdrop wall, with 2x4 studs 32" on center. It will feature a horizontal 2x4 benchtop support all the way around the room, the top of which will be 37 3/4" above the floor. Thus, when I add 3 1/2 inches for the 1x4 bench frame member, 3/4'' plywood and 1/2" hotosote, I will have my 42 1/2" bench height. (I may decide to go an inch or so higher.) Just behind this bench support, there will be horizontal stingers to support additonal vertical 2x4s, which will be added between the floor-to-cieling 2x4s, thus making the sheetrock wall structure above the bench top 16' on center, and the open wall structure below the bench 32" on center. This will allow easy access underneather the bench from the access aisles behind the backdrop wall. Finally at designated places in the walls (especially in then 48'' radius rounded sheetrock corners) the additional vertical stringers will be iinstalled 10'' above the bench height to allow me to cut away sheet rock and have access to the inside of tunnels through portals that can be as large as 32" x 10".


August 2015

New Layout Design

I got a call from my architect earlier this week, and he asked me if, for various reason having to do with  the functioning and look of the overall building, he could put the entrance door in the center of the long wall of the new train room as opposed to putting it at the end of the room the way I had originally drawn it. He also said that in this new configuration, he thought the stairs could be constructed outside of the train room space thus allowing use of the full 40' x 20' space. I told him that I though this would work, but that I would have to redo the preliminary layout design. The drawing below represents my first thoughts on this redesign. I have kept a lot of the ideas I used in the original design, including double track mainline all the way around, a large yard with dntrances on both sides, and a large multi-spur passenger terminal. Also I have kept one of the things I like best about my current design, which I have copied  in this design: the city of Altamont will occupy the entire 20-foot wall of the trainroom with one track running above through the city in the foreground and all mainline track flanked by a few "holding" sidings running underneath out of sight.

I have redesigned the yard to allow access from mainline trackage on both sides thus creating a second reversing loop, and I rearranged the behind-the-backdrop access aisles and
added the two wide areas to turn trains around on either side of the doorway. This last required me to narrow most of the center section by a foot, but I was able to flare it out at the end to allow a full 30" radius for the turnaround. Again my minimum radius overall is still 30", my maximum grade is 1" in 4' or just slightly over 2%, and with a few exceptions, the maximum reach across the benchtop to any spot on the layout is about 34".

After living with this new design for a few days, I think I like it better. The narrower center section seems to work better without potentially needing a central backdrop wall, the entrance to the room will be more spectacular, and there is a lot more aisle space for viewers as well as ample room for a control desk. (Note: 3/3/2016 The one serious problem here is the 2 1/2 foot wide aisles. Three feet aisles is in my mind the absolute minimum. Fortunately, as things progressed, the room turned out to be larger, and as you will see, all the interior aisles and the behind-the backdrop access aisles are at least three feet wide in the final plan.)

Reworking the design to accommodate the center entrance and the outside stairway.

Planning the Flow of the Landscape: Tweaking the Track Plan

As previously discussed, when modeling in N Scale, especially on a large layout, one can create remarkably convincing panoramic landscapes. Indeed, I contend that the central focus of modeling in N is not so much individual detailed "scenes," but rather the overall flow of the terrain through which your railroad travels. Creating convincing scenery in this regard takes considerable planning.

Now that I have a preliminary layout design that suggests the configuration of the railroad and general locations for yards, towns, and so on, I can begin to think about not only details like roads and fields and locations for structures, but also about hills and mountains and the overall topography in general. As my thoughts begin to gel on the size, the form, and the placement of these features, I may need to tweak my track plan to achieve realistic effects. The central focus of this kind of thinking is this notion: even though on a model railroad one builds the roadbed and track first and then adds the topographical features later, in the real world it is the other way around: the topography came first, and the railroad later altered it to suit its needs. This is a critical principle: when designing scenery, one should continually ask, "What did the landscape look like before the railroad was built?" Continually trying to answer this question should result in an interactive design process in which roadbed routing and elevations, roads and towns, and topographical forms ... indeed all the model's elements ... interact. Mountains are not simply placed beside the track, but rather the track slices through the hills in "cuts" and traverses the valleys on fill. In this kind of design process, nothing is ever "engraved in stone." Everything is always changeable and in flux. With this kind of approach one allows for both a well-thought-out initial plan AND a plan that can change as it is built.

The first consideration I made regarding the new design followed a simple maxim: "Try to avoid running too much track parallel to the edge of the benchwork." If all of the track is parallel to the edge of the narrow benchs, one quickly gets the feeling that the layout was designed to conform to the bench,  and the result is therefore somewhat artificial looking. As you can see, from the below track plan, I have altered the original plan to set the ladder at Altamont Terminal and the highroad track at East River at angles to the edge of the bench. I have also created a sweeping curve in the ladder at Atlamont Yard to break it away from the predictability of paralleling the bench edge. It doesn't take much, but all this will really help in the end. 

The second consideration I have made regarding the new design has to do with curves and curve radii. In N scale one can run most trains around a curve as tight as 18 inches in radius. This is highly unrealistic. Real train mainline curve radii are generally something on the order of  750 to 1500 feet. This translates to roughly on the order of from 50 to 100 inch radii in N scale. So, even my generous 30 inch mainline curves are much, much tighter than their prototypes .... and so they will never look exactly right, especially when carrying 70-foot-long passenger cars. So what can be done short of building a 50 x 100 foot rain room? The answer is, of course, hide them in tunnels. Fortunately, the offending curves are all in the corners of the room and in the "turn-arounds" at the dead ends of benchwork ... natural places for mountains. Thus, I can begin to design my topography on the A&BR2 by placing six mountains as depicted below, although I will not worry too much about the form of these mountains at this early stage.

At this point, things are beginning to take shape topographically in a very general sort of way. You may have noticed that when I added the two mountains at the ends of the bench to cover two of the large circular turnarounds, I put the high ground right over the towns of Fitzhugh and Westridge. This is OK for now, because what I'm doing is establishing what the topography looked like before the railroad and towns were built. Later I will create cuts into tunnel entrances and some stepped terracing on these hillsides to accommodate the streets and buildings of the towns, effectively excavating the terrain just the way it might have happened in the real world. The end result will be a great deal more real looking than it would have been had I simply built the town on the flat and placed a mountain next to it.

Similarly, along the length of the wall where the city of Altamont is to rise, I will create terraces stepping down from the high ground along the backdrop wall to the single traversing track near the edge of the benchwork. I have also added a small mountain in the center in order to brake up the regularity of the future terracing. The single traversing high-road track will cut through this small hill and the streets will tunnel through it as suggested in the plan below.  

roads Tweaking the track plan to set the ladders of Altamont Terminal and Altamont Yard and the highroad track at East River at angles to the bench edge, and roughing in mountain locations to hide un-prototypically tight curves in tunnels.

Roads, indicated in gray with bridges in orange, are roughed in without elevation notations just to get a general feel for how things might look as we progress. Crossing signals are indicated by two red dots.

Road Grading, Structure Pads, and Town Terracing: Preliminary Considerations  

Now that I have rough idea for a track plan and a general notion of the lay-of-the-land through which my track will pass, I can give a little consideration to how this landscape might need to be altered to accommodate roads, structures, and towns. At this early stage this is planning of the most general sort - just a rough suggestion of where roads may need to go to service trackside structures and where mountains may need to be "graded" to create terraces for mountain towns. This will all undoubtedly change and become much more defined as I progress into the construction phase, but for now I am only considering functionality so I can begin to visualize things from a macro point-of-view.

Notice in the above schematic that, at this point, the numbers +0, +2, +5 etc. are rough indications of track elevations at various key points. I have purposely omitted indicating road or town elevations as they will be determined by the topography and by nearby track elevations  (although a few bridge overpasses are indicated  in orange). Roads are much more flexible with regard to turn radii and elevation changes than track, so I usually follow a rough plan and tweak it as I build, laying in 1/2" homosote pads and terraces (designed to accommodate the footprints of the selected structures) and homosote strips for highway and street roadbed all of which will be built right into the homosote framework of the future topography. (Much more on this later, also see the tutorial "Shaping the Landscape" in the "Tutorials" section of this website.)

Lastly, now that I have an idea of where the roadways will go, I have indicated the location of crossing signals and gates using double red dots. I will defer decisions regarding signal placement until I finalize the location of all detection blocks.

The Wiring Scheme: Thinking About Detection Blocks 

Over the years I have read Allan Gartner's "Wiring for DDC" several times. I recommend it most enthusiastically.
(You can link to it at: http://www.wiringfordcc.com/wirefordcc_toc.htm) It has been my experience that following Allan's methods results in really "bullet-proof" DCC wiring. I'll get into the details of applying the methods when and if I actually start wiring the new layout. For now, I need only point out the Allan recommends that every single piece of rail on the layout be solder-connected to the power buses using droppers and feeders, with no soldered rail joints, and that every turnout be modified so that power to all rail components is likewise supplied directly from the power bus using soldered droppers or soldered jumpers. It is a lot of work, but it is worth it. Much more on this later.

Today I am thinking about detection blocks and power districts. Where will I put my blocks and how will they function in computer-controlled operations when the layout is complete. This is a critical part of layout planning, so I'll not rush this thought process. Rather I'll  think about it a bit, design a plan, think about it some more, and then revise the plan. The schematic below is the third revision of the original. There will surely be many more, but this one is getting close.

Instead of just putting detection blocks in just yards and at stations and in other places where I will want trains to stop, I have opted to place them pretty much everywhere. This is to say that any lengthy section of track between turnouts will be wired as at least one detection block. Very long track sections, might get two or more blocks. The turnouts themselves and any very short sections of track in between will be wired directly to the power district power bus and will constitute undetected routes between detection blocks. The plan is critical, so I take my time, paying attention to the logical flow of traffic through blocks as the system tracks the train around the layout and to the logical groupings of blocks into protected power districts.

In the scheme below I have 72 detection blocks divided among 4 discrete short circuit protected power districts (Power Districts Green, Purple, Brown and Blue) each contains 16 detection blocks. I also have 8 more detection blocks (Power District Yellow) and the 2 power reversing districts (Power District Red), each of which contains only a single block. These 2 blocks are the single sections of track that are the reversing loops on the layout. You may notice that I have drawn the wire routing on the schematic, but in this image the resolution is not adequate for you to examine it in detail. We will get to that later.

wire 4 Power Districts (Purple, Green, Blue, Brown)each with 16 detection blocks, 1 District (Yellow) with 8 detection blocks and 2 single line Districts with auto-reversing for the reversing loops (Red).

The Digitrax system I use is pretty straightforward in this regard. It employs 2 PM42 power managers, each with four outputs. Each output can be configured to power a short circuit protection power district that shuts down when it encounters a short without shutting down other power districts, or to power a district that reverses polarity upon encountering a short.  To create the detection blocks, Digitrax systems use the BDL168 card. This device has four inputs that take power from a PM42 district. Each input on the BDL168 has 4 outputs, so the unit will support 16 detection blocks. In a normal situation a BDL168 will have all 4 inputs fed by the same power district output from the PM42, resulting in a power district with 16 detection blocks. This is the case in 4 of the power districts above (Green, Purple, Brown, and Blue). To create power district Yellow, I feed inputs #1 an #2 of  a BDL168 from district #1 of  a PM42 to create 8 detection blocks. To create the two power districts Red, I feed input #3 and #4 of the same BDL168 from the same PM42 using the outputs for power districts #2 and #3, which have been configured as auto-revesing districts thus creating 2 auto-reversing power districts each with a potential of up to 4 detection blocks.  In this case I have only a single block per reversing district, each powering and providing detection for a single polarity revesing loop.  (Note: I later scrapped the idea of using the BDL for auto-reversing because it uses four detection blocks for each auto-reverisng circuit and, as I revised the track plan, I found I need these extra eight detection blocks. I therefore decided to use a separate auto reversng device on the two yard crossover blocks that require auto-reversing. (See post of 2-6-2016)

The Naming of Power Districts, Detection Blocks and Turnouts

The TrainController software I use allows you to name detection blocks whatever you like, but it will require you to create an internal digital address for each block according to a specific convention. Most model railroaders like to name blocks descriptively - names that suggest the block's location on the layout. For example if I Number tracks sequentially in a section beginning with the track closest to the edge of the backdrop, I might choose to call the block that is the section of track closest to the backdrop at East River "East River#5" or ER5. Or I might number the tracks in the Altamont Yard ladder such that the top or Northern most track would be "AY#1;" the next track down would be "AY#2" etc.

Of course, what you name the block does not matter to the digital control system. It has its own set of addresses. For example the track that we called "ER#5" will, according to the Digitrax rules as applied to my current wiring scheme, have a digital address "97-4." That is, it is output #4 from the DBL168 which I have designated and programed to have Board Address #97. If I name the Power District using the board numer of the BDL168 that they supply, this block name will also tell me the powerdistrict that the block in. In this case Power District #97, which is PM42 #90, output #3.

So the approach I think I will take to naming my blocks will include location and digital the block address. In the above example, the North Bound Outside Line block at East River will be called "ER5/97-4." A bit cumbersome perhaps, but it covers the all bases. 

Likewise, in TrainController turnouts can be called whatever your like, but each turnout must have a unique digital address (any number from 1 to 2048)  programmed into the DS64, a stationary decoder that can control four Tortoise switch machines. Following the same convention, I will name my turnouts to reflect location, destination, secondary location, DS64 board number and its output number, and the digital address.  I plan to use the digital addresses 1100 to 1600 for turnout address according to the following scheme:
1100 - Fitzhugh
1200 - Westridge
1300 - East River
1400 - Little River
1500 - Altamont Yard
1600 - Altamont Terminal
So the turnout at Fitzhugh that shunts trains to the AltamontCity High Line will be Fitz/AltamontCity/57-3/1105 assuming it is power by DS64 #57, output #3 and assuming I assign the digital address 1105 to this output. In another example, the crossover from one mainline to the other to allow access to and from the northern entrance to the depot sidings a East River will be ER/Crossover/Side/North/xx-x/1301A&B (where xx-x is a holdong place for the DS64 board numer and its output number (1-4), and "A&B" because this DS64 one output will operate both turnouts of the crossover simultaneously.

Note: I later ame to find that four digit DS64 address (that is addresses above 1000) can only be programmed and accessed using a computer program like TrainController and are not accessable via the Digitrax throttle, So I changed the address ranges above by subtraccting 1000. Thus I will use address 100 to 600 not 1100 to 1600. 

If I use these block and turnout names in all wire and BDL168 terminal block and DS64 output labeling and in under-the-bench Tortoise and dropper labeling, it will greatly facilitate troubleshooting and other servicing.

What Do I Need? 
8-12-2015 (updated 9-20-2015)

A key part of the model railroad planning process involves selecting what kind of components one wishes ot use on a new project, calculating the quantities needed, finding the best sources, and preparing some kind of budget. As the foregoing text details, I have pretty much decided what components to use and where to get them. I need now determine the quanties required and what the cost will be.

I'll start with track. The CAD layout design software I use has a neat little inventory feature. It totals up exactly what your layout will require in the way of track and turnouts. According to the current design I will need the following to build the A&BR2 if I am able to salvage all to the track and turnouts from the current layout: 
   27 | Large Left
   26 | Large Right
   15 | Medium Left
    7 | Medium Right
   9334 inches N Flex Track

A&BR current layout

   15 | Large Left
   13 | Large Right
   13 | Medium Left
   10 | Medium Right
   4467.956 inches N Flex Track


   12 | Large Left
   13 Large Right
   2 | Medium Left
   0| Medium Right
   3866 inches N Flex Track

I see no reason why I can't reuse all of the the Peco Code 55 turnouts from the current layout, and certainly a large portion of the track will be reusable, perhaps 90%. So I will need to purchase only 7 large left, 9 large right, and 8 medium left turnouts (~$540),
about 24 more Tortoise switch machines (~$388), and about 3866 inches of N Flex track (that's  107 3 foot sections or about 3 boxes of 30  or about ~$500). Total =  $1428. 

As far as Digitrax control devises go, I am pretty well set. I have the 5 BDL 168 block detection cards (4 on the layout and 1 spare), 14 DS64 stationary decoders, 3 SE8C signal control cards, 2 PM42 Power Manager cards, 1 remote loconet plug in port, and a DCS 100 Command Station. I have not done any planning so far regarding signal placement, but, if I put any signals at all in the yard and the main passsenger terminal, I am pretty sure I will need at least 1 more SE8C signal card (~$100) to go with the 3 I already have, and maybe 15 more masts (~$50), also one more remote plugin station (~$13) and quite possibly a BD150 booster (~$150), plus maybe 6 more DS64s ($150). Total =  $463. 

Given the large size of the new trainroom this project is going to take alot of wire, and I seriously doubt that I can salvage too much from the current setup - perhaps 300 feet of  red stranded 12 AGW and a like amount of red and black stranded 16 AGW. Maybe a little more, but I do not want to get into trying to splice together shorter pieces. It is better just to buy new, and it is best to get it all up front to take advantage of quantity discounts, which, when it comes to wire are sizable. So, how much will I need? Surely it is best to order it all up front in bulk to take advantage large spool pricing and probably avoid shipping charges. 

For power buses I use stranded 12AGW building wire. Assume the average 12AGW mian line red bus run in a trough is 2/3 of the length of the room, that is to say 30 feet. There are 44 main line blocks, so that's 1320 feet. The remaining 28 blocks are for ladders and the turntable fan. These will require much shorter runs, say 10 feet on average, so that's another 280 feet. Add another 200 feet or so for PM42 to BDL168 lines, for the main undetected power bus, and for PM42 to main booster lines, and we get a total of 1800 feet, less the 300 salvaged from the current layout gives us about 1500 feet or 3 500 foot spools (~$160). Turnouts and undetected routes will get power from the undetected power bus which is part of the additional 200 feet I added, so I need no additional heavy wire for them. 

For black 12AGW, I'll need on average 30 feet for each power zone or reversing loop, that is 180 feet plus the 200 feet for the main etc. that is 380 feet. So better get 1 500 foot spool of black 12 AGW (~$53).

For feeders I use stranded 16AGW automotive wire. I'll need about 3 feet on average for each power block connection and 3 feet for each turnout. I will connect to the power blocks every 3 feet or so. Mainline power block might average 40 feet (10 connections  each times 44 mainline blocks times 3 feet each = 1320 feet). Ladder power blocks will probably average about 8 feet (2 connections each time 28 ladder blocks times 3 feet = 168 feet). There are 61 turnouts, so time 3 that's another 183 feet. Plus 200 feet for short run connections to BDL168 and PM42 etc. In all, I estimate I need 1663 feet of red and of black. So let say 6 500 spools all together, 3 red and 3 black ($330).

For lights I'll also use 16 AGW automotive wire in green and blue. I don't really need wire this heavy to run my LEDs, but automotive wire or "primary" is actually cheaper than lower gauge "hookup" wire. At this stage I can't estimate the amounts I'll need but I'm sure it will be over 500 feet. So 1 each blue and green 500 foot spool at (~$110).

For power droppers I'll use 20 agw insulated solid copper. 2 100 foot rolls should do it (1 each red and black for track power, and 2 rolls of 22agw solid copper insulated green and blue for lights droppers) (~$55). Total ~$700.

I'll also need some cable ties.

I'll also need more loconet cable and more RJ12 connectors. 100 feet 6 conductor phone wire (~$10) and 50 RJ12 Connectors (~$15).

61 Edge connectors for the Tortoise Switch Machines.(~$180).

Barrier blocks pending and inventory of the current layout. Need 63 8 position, 5 20 position.

For lighting structures I'll need about 200 each 5mm warm white LEDs (~$10). I will also purchase a 5 volt power supply. The power suppl I hav emy eye on right no has both 12 volt 6 amp output as well as a 5 vlt 24 amp output. $40.

For everything except loconet and the Tortoises I will use hard solder connections insulated with electrical tape or wire nuts. The Digitrax control cards of course have edge connectors, and I will wire short lengths of 16 AGW to the pins of these edge connectors and then marry these short "pigtails" to the heavier wiring of the rest of the layout using the Barrier Blocks.  I will install individual block on/off switches in these "pigtails." I have appropriate toggle switchs on hand from an old non-DCC layout, so toggles are not on my list. (Note: I later decider not to add these switches.)

I'll have contractor erect the interior walls with their 48 inch radius corners and install all the room lighting, so none of that is part of my shopping list. The next item to estimate then, is lumber for the benchwork: 8 8' 2x4s , 18 8' 1x6s, 21 8' 1x4s, 8 finish grade 3/4" plywood 4x 8 sheets  and 8 1/2" homosote 4 x8 sheets. Black formica laminate 2 each 12x 30" sheets = $140. A quick look at the Lowes web site for prices, some quick calculation and, voila: ~$790.  

I'll also need some more black skirts with velcro strips~$300, and 100' feet of 1" stick-back velcro ($30). 

I do not think I will try to reuse rail joiners, so with 64 turnouts I'll need 400 plastic insulated Peco joiners plus maybe 30 more for other block isolation, and perhaps an equal number of Peco metal joiners. Let's say 16 packages of 24 metal joiners @ &2.50 and 32 packages of 12 plastic @ $2.50 or a total of $112.

Likewise I don't think it is a good idea to try to reuse cork roadbed, so I'll need 10,500 inches of the stuff; that is about 900 feet or  300 strips or 6 boxes of 25 strips at $16 per box = $96. Plus 19 large left and 19 large right and 20 medium left and 8 medium right cork turnout pads @ about $2 each  for total of about $112. (Note: I later decided not to use the precut raodbed and switch pads. I'll use 1/8" corl sheet andcut my own.)

Glues and adhesives (3M Adhesive spray, CA, 1 gal Elemrs white glue, 1 quart water based contact cement for glueing down roadway and paving styrene sheets, and a large squeeze bottle of yellow carperters glue for homostoe mountan frames), nails and screws (long and short sheet rock screws, small brads with flat head for screen, small wood scres for securing circuit boards and tortoise machines), plastic screen, .40 sheet styrene,  (about $100)

In the computer area, I need several things. First, I need to upgrade my TrainController software from, Silver to Gold, the highend wham-o, expert version that I will need to take full advantage of the +Net software I need to buy in order to employ two computers  as previously alluded to in this blog. The upgrade and the +Net software will cost me about $600. I plan to do this soon so I can get things running on the current layout. That way I will be fully conversant with this new technology by the time I start wiring the A&BR2.

I also will need a second computer. My thinking here is to get two inexpensive matching 4 gig refurbished HP desktops running windows 7 or 8 (about $200 each) so I will have matching computers to sync up via +Net when I get to that. This along with two small new monitors will make a slick control desk for the new layout. Software, computers, monitors, additional loconet interface box, keyboard and mouse will probably total in the area of $1400. 

If my sound scheme works I'll need another 7.1 sourround USB Box ($32) and 3 more desktop Speakers ($25) and pretty good deal of some speaker wire ($100).

4 gal white primer, 3 gal sky blue, 1 gal white, 1 gal black = 9 gal @ ~$35 = $315. This will cover the materials I need to paint the graded blue sky backdrop with some nice clouds dry brushed onto it. I already have the brushes and multiple rollers and pans I need to do the graded shading of the sky (slowly getting lighter as it approaches the horizon). Dry brushing very convincing cloud forms is easy once you get the hang of it, but there is a trick to it. More on that later. The backdrop will also get a few rows of distant mountains, but I generally just lightly pencil their outline on the wall to begin with. I will  paint them later, one section at time, when I know exactly what will go in front of them. (see "Marrying the Backdrop to the Layout" elsewhere on this site.)

So about $6400 in all to paint the backdrop, build the bench work, lay the track, wire, and computer and sound equip a 40 x 20 foot layout with 51 turnounts, 61 blocks, and over 850 feet of track. This does not include the interior backdrop walls and the room lighting and sofits, the HVAC and AC wiring, which the contractor will see to. It also assumes that I have salvaged from my current layout almost all of the turnouts I need along about 40% of the track, almost all the DCC devises and software, and a little wire. 

Selecting Vendors

I've done a little on line shopping this morning to check prices and inventories. I'll probably use the following vendors. most of whom I am already familiar with:

M. B. Klein ("Model Train Stuff" on line).
Peco track and turnouts, the rail joiners, and the Digitrax devices, cork roadbed and turnout pads, Tortoise switch machines and Tortoise edge connectors. I have dealt with this company for years. Their prices are good, and their service is excellent. With such a large order, I hope to negotiate a small discount and get free shipping. I also need to check their prices against Amazon who has listed some incredible deals on model railroad products lately.

ActivePowerSports online. This is new online vendor, and thier prices on man items are a bit better than even Klein, but the sevice and selection is not as good.

Lumber, 12 agw building wire, large cable ties, glue, nails, screws, and paint

Appalachian Supply (my local lumber dealer)
Homosote and Plywood

Dell City.org 
Wire, speaker wire, hookup wire etc. (I will probably get the 12 agw from Lowe's on line and have it shipped to the store near me, and I'll get the 16 agw from Amazon.)

US Plastic Corp
Styrene sheets

5mm LEDs, desktop speakers, and Encore USB Surround Sound Box. I generally first check Amazon for anything I need. With Prime I get free shipping, their prices are generally the best, and they have a lot of electronic stuff.

Jameco Electronics
loconet cable, RJ12 connectors,  and barrier blocks, small cable ties

Best Buys (on line)
Refurbished computers and monitors etc.

Further Tweeking of the Track Plan


I sketched in locations for the depots at East River, Westridge, and Fitzhugh this morning. The depot at Fitzugh is an on-the-mainline stop. It has no siding, so no track changes were needed other than to separate the to mainline tracks enough to allow a long passenger paltform to fit in between. However, the depots at East River and Westridge do have sidings, and so I added cross-overs to accommodate trains approaching from and departing to the mainline track away from the sidings. I also added a crossover near Little River to allow the EB highline to access the Altamont City line. Finally, I am thinking about adding a streetcar line down the center of the main street in Altamont. I adjusted block locations accordingly and also adjusted my shopping list and the budgets above to reflect these additions.

This is the kind of tweaking that comes from thinking about the plan and tracing imaginary train routes. The added crossovers allow access to sidings located on the opposite side of the other parallel mainline track.

Tweaking the Wiring Schematic


I think the track plan is now close to the way I want it, but I'm still going over the positioning of  the blocks on the A&BR2. I've done a little more tweaking, and shuffled the power districts a little.  One of the the things that worries me a little is the length of some of the wire runs. A few of these runs are over 40 feet, and most experts including Allan Gartner, say 30 feet is about the max one should go without adding more boosters. Are my blocks so long that I am asking for inductance and termination problems? Will I need an additional booster?

I think I have hit on a solution. If I place the booster or boosters in the center of the room under the central section of benchwork and centrally locate the two  PM42s along the north and south edge of the central bench section, and if I install under-the-floor troughs from the center of the central  bench to both the north wall bench and the south wall bench, I can then place all BDL168 block detection devices in the center of their respective benches while, at the same time, holding the wire runs from the PM$2s to the various BDL168s to under ten feet as recommended buy Digitrax. Feeding buses like this out from the center will hold almost all of my runs to under 30 feet. Most will be considerably shorter.

This way, if I still happen to need a second booster, I can install it easily right next to the first one. I need only wire them together and then have one feed each of the two PM42s. This will split the overall load almost equally. 

The revised schematic below reflects all these changes and color-codes the 5 power districts
(Green, Blue, Purple, Brown, and Yellow) and the two reversing tracks (Red). It is still not of high enough resolution to allow a detailed analysis of the layout wiring nor does it detail the turnout decoders or power buses or or signal logic cards and feeders or all of the track connections. All of this will be detailed later, when I create separate wiring schematics for each power district.

More Work with Block and Turnout Names


I've spent more time honing the naming of detection blocks and turnouts and labeling these blocks and turnouts out on my schematic. In overview, things look pretty messy with all of this text on the schematic; but when I get down to making a block list and doing individual schematics for each power district, things will look much more orderly. I will use these individual schematics as a guide when I start wiring, and the block lists to create computerized labels for wires, barrier blocks, and digital devices. Having a comprehensive plan and labeling as you go are essential when it comes to keeping things straight once wiring begins.
Block Names and Locations Overlaid on Track and Wiring Schematic. When rendered in overview like this, the scale is too small to be of much practical use, but this gives a general idea of what the schematic looks like. 


Turnout Names and Locations Schematic Overview. Notice I have used the notation xx-x where the DS64 board number and output number will go. I suppose I could assign these number now, but I think it is best to wait and assign them as I install and hook up the DS64 swithch machne control decoders.
With such long and descriptive turnout names, using the full name to name the turnouts in TrainController will result in something of a mess. So I'll just use the 4 digit digital addresses to name the turnouts in TrainController. I'll use these longer descriptive names, which include the digital address, for all wire and devise labeling.


Ever since I started this blog, I have been having nagging doubts about this project. It is not so much that I am unwilling to demolish the A&BR1. I can dealt with that. No,  my concern rises from the notion that the A&BR2 might be bitting off a little too much. Simply put, I fear that it might just be to big, When the interior backdrop walls are complete, I will have an empty trainroom that is 35' x 17.5' - over 600 square feet - a room that will accommodate over 125 ruining feet of bench front - over 3 times the size of the A&BR1 - almost 4 miles in N scale - enormous. On the face of it, this actually sounds pretty good, but when I stand in my current 20' x 12' trainroom, which is actually pretty big, and try to imagine an empty room almost twice as long and twice as wide, I feel a little overwhelmed. It is too much?

Too much for what? Too much to finish before I die? Too large to actually work? So much track to lay, wire to run, and scenery to build that I will grow tied of it?

Well, perhaps, but there is another side of me that says, I need this project to keep me engaged in something, to inspire me and drive me on, to maintain my vitality, happiness, and sence of purpose - even to add years to my life.

It's a conundrum.

Upon refection, I think it comes down to this: build the A&BR2 or find another passion to occupy my thoughts and actions in retirement. I have been pretty good at finding new passions ever since I retired in 1996. For years I wrote books, but the publishing industry changed so drastically that this became sort of pointless. For a while, I spent time learning C# and programming multi-relational databases using SQL, but I am continually searching for applications, and I have reached the point where I can't really progress in any meaningful way without some help, ie. going back to school, which I really don't want to do. For a short while I dabbled in robotics and did I/O programming to control various devices, especially servo motors, but again the lack of applications seemed to be a road block.

So, in the absence of anything else, I might just need this project. Whatever the case, I plan to follow it along and just kind of see where it leads me. The real decision will come when my architect and I complete the plans for the new building. At that point, I will have to do some real soul-searching. What will it cost? Should I build it? I can make all kinds of rationales for the the new structure, but the real truth is: if I build it, my central, underlying motivation will be to create the A&BR2.

Block Lists


I am now starting to get pretty comfortable with my track plan and my block placements. Certainly there will be a few changes as I continue to think about the schematic, and probably I will make a few more changes when I begin to trace the trackplan onto the layout benchwork. But what I have now is pretty close, and so I can go ahead and create a list detailing the configuration of each of the 5 BDL168 16 block detection cards.









































































AT/Cross/99/9 Rev




AY/Cross/99-13 Rev




In all, 76 detection blocks, on 5 regular short circuit protected power districts and 2 auto reversing power districts, and one BD4. On a large layout, I cannot overemphasize the importance to this kind of carefully considered pre-planning when it comes to block naming, block powering,  and especially block placement. There is a lot going on here, and this kind of planning and documentation will really help keep things straight once wiring and labeling begins. What is more, these kind of schematics and lists along with thorough wire and device labeling will prove invaluable during troubleshooting and service once the layout is complete.

Turnout Lists
Similarly I have prepared a list of Turnouts. there are 68 in all, so I'll need 17 4 output DS64 stationary decoders, plus I'll need probably three more DS64s to control lighting relays, crossing flashers. So 20 all together. Again, notice I have entered these names using "xx-x" placeholders where the DS64 board number and output will go. I will number DS64's sequentially around the layout as I install them beginning with the board number 50. So I'll have numbers  50 through 69 each with outputs 1-4. Note: The listing below is an early iteration does not exactly reflect the finished plan, nor does it incorporate the subsequent three digit numbering scheme, but this at least gives a good idea of my method.

A&BR2 Turnout list (68)

Trunouts at Fitzhugh (7)










Turnouts at Westridge (6)


WR/Crossover /Side/North/xx-x/1202-A&B




Turnouts at East River (8)








Turnouts at Little River(7)








Turnouts at Altamont Yard (19)


















Turnouts at Altamont Terminal (21)






AT SouthSide/West/xx-x/1606














 Again this list will aid in keeping things in order when wiring begins and when the numbers are filled it will also be helpful in layout servicing and troubleshooting. 

Creating a Switchboard in TrainController

Now that I have a track plan and all of my block names, I can begin creating a switchboard in the new TrainController Gold software that I have just downloaded from Freiwald Software. I have not yet purchased the license for Gold, but Freiwald allows you to use the software in an off-line, demo mode for 30 days. You can create all of the objects needed to run your railroad, and even run trains on the screen in a simulated mode. The only thing you cannot do in this demo mode is connect to your layout. You need a license for that. What is more, the 30 limit is calculated to include only the days you make changes to a file, so it is really quite a liberal trail period.

I am currently licensed to use the Silver version of TrainController, and designing the switchboard in Gold is pretty much the same as Silver, so there was no real learning curve to get beyond. In Gold there are some much more sophisticated methods for handling turn tables, but trying to learn all that now with no real layout to test it on seemed risky, so I used the a turntable bridge block connected to a small ladder, just the way I have been doing in Silver. It may look a little clunky, but it works. Anyway, the below switchboard only took a few hours to create and a good portion of that was spent pasting in the block names. I have not yet put in my turnout names, but the software assigns a working name to each object at the time it is created. So I can go back and put in my own unique turnout names any time I like.

Still More Track Plan Tweaking

Unlike the renderings from my XCAD model railroad design software, which are quite detailed and strictly scaled, the rendering of the TrainController switchboard is a highly simplified block diagram of the track plan. This stylized format makes it much easier to evaluate the flow of the layout.

Looking at the above switchboard, I see two areas that still need attention. Both the North and South entrances to Altamont Yard are not quite right in the way they tie into the auto reversing crossing track. Designing a yard between two double track mainlines with two AD (arrival and departure) tracks and an auto reversing crossover is tricky. Below are before and after renderings detailing these changes using the TrainController switchboard to present the simplified overview, and closeup of the detail using XCAD.  

Notice that this change required the creation of an additional block, the deletion of 3 turnouts and the addition of four new turnouts. All my BDLs are full, so I did away with the block Alt6/Center.98-1 and used that position to create the new block for the new block I created when I split AYSouth/CrossAD into two blocks: AY/SouthAD and AY/Cross, the reversing section. The track section that was Alt6Center will now be an undetected route.




In additon I have added a runaround for the North AD Track. I will update both the block list and the turnout list, although I see no need to post the updated lists or the revised XCAD rendering here.

Below is a rendering of the revised TranController switchboard reflecting all of the changes.

Thinking About the Yard

I have been busy entering turnout names and labels into the TrainController Switchboard. I have also been contemplating Altamont Yard to make sure my tweaking is done. I reproduced a closeup of the yard switchboard in a separate jeg file and pasted some labels over the top in order to define yard functionality. This helps to evaluate the way traffic will flow. Here is the completed jeg showing the yard between the two double track mainlines, with two separate arrival/departure tracks, a yard crossover for correspondence between the North and the South mainline tracks, 2 run arounds, a staging track, 4 classification tracks, a caboose track and an auxiliary engine service track.

Remember, TrainController Switchboard schematics are not to scale, so don't worry if, for example, the South AD track looks a little short. On the scaled XCAD schematic it is not. It is actually about quite a bit longer than my longerst train. Ideally, these AD tracks ought to be something on the order of twice as long as the longest train on the layout, and any reversing crossover should be at least as long as the longest train.

Signal Placement and Terminal Track Plan


Over the last few days, I have been fooling with signal placement. This will require a great deal of thought, but I think it is best to just sketch out something to begin and tweak it as I go along. Accordingly, I have entered my initial mainline signal plan onto the TrainController switchboard. Although stylized and not to scale, this switchboard offers the most clear and readable representation of my layout. Each little green signal icon represents one three aspect signal head. As you can see, I will be using double headed masts with 3 aspect heads exclusively (the top head always shows the mainline ahead and the bottom head the siding ahead). Conditions for each of these signal heads will be programmed into TrainController later. The resulting logic will determine whether each head is red, yellow, or green depending on block occupancy and on turnout positions ahead of the signal. When wiring and programming are complete, these icons will change color in sync with the signal heads on the layout. (Note: for a more complete discussion of signals and signal placement see entry of 2/24/2016)

I also made a few changes to the track plan at Altamont Terminal, altering the entrance to the ladder and the positioning of the crossing reversing track to streamline entry into the ladder and to more exactly reflect the plan in the XCAD  schematic.


Another Design

Further discussions with my architect have resulted in an altered design for the new building that will house the train room. Trainroom dimensions have shifted, and I have completely re-done the XCAD schematic to shoehorn my trackplan into a different space. There are some compromises regarding the access aisles, but in general it all fit pretty well in what has now become effectively a 31' x 19' space.

Since it is a not-to-scale, stylized representation of the layout, I was able to alter my TrainController switchboard to reflect the new plan without much effort, and with a few exceptions, my turnout and power block lists remain the same. All of this serves to remind me that, at this point, my designs are only preliminary renderings; and until I get finished architectural drawings, I will not be able to create a "down-to-the-inch" track plan.

Nonetheless, every re-working of the track plan, yields positive results. In altering the plan to fit into a smaller space, I made some changes, which I will incorporate into my original design should I go back to the original room dimensions. Notably, I have cut back on the number of ladder tracks at Altamont Terminal, changed the access to one of the hidden sidings under Altamont City, added a few spurs, and moved one siding. All of these changes are  imporvements, regardless of final room size.

Below is the XCAD rendering of the revised trainroom.


September 2015

Further Considerations Arising from a Smaller Trainroom

The proposed new trainroom is about 33' x 19, that is 627 square feet. The original design totaled almost 595 square feet. So I have roughly the same, but with a little less length and a little more more width. This new plan cuts the amount of track needed by about 1300 inches (about 15%), and it will probably reduce the amount of red bus wire I need by about 300 feet or so, so I can get 5 boxes of track instead of 6 and perhaps 2 500 foot rolls of 12 agw wire instead of 3. The change in the amount of black bus wire is minimal and the requirements for red and black 16 agw feeder wire will not change much. Likewise the requirements for loconet cable and dropper wire are inconsequential. 

In the original design, I was right out on the edge when it came to the length of the longest bus runs. With this new plan, I have probably reduced the risk of creating any of the problems that can arise from overly long bus runs.

On the whole, I actually feel pretty good about the altered the trainroom. I think the concept of the trackplan is still fine, but whether or not cramming all of the features of the original layout into a slightly narrower space will appear over-crowded is a question that can only be answered when track laying begins. I suspect that there is still plenty of space for everything, but should things look cluttered when I begin to lay track, I'll re-assess the trackplan.

There is always give and take between the track plan and track laying, so this kind of flexibility is a good thing. For example, when I begin to lay track, I get a feel for the scale and the look of of things that cannot be gleaned from an on-paper plan. At this point, I can begin to make some fundamental decisions about the exact size and shape of topographical features and about the placement of key structures. This will require sketching the topo of mountains and roads, etc. onto the benchwork and sometimes tweaking track routing to make things work exactly right. It will also require looking up the footprint dimensions of key structures and altering track spacing as needed.

Still MoreTweaking of the Benchwork and Track Plan

The change in trainroom size has precipitated numerous changes to the track plan. Although in general layout and function remains pretty much the same, many details had to be tweaked to accommodate the reconfigured space. Along with these changes have came improvements. For example, I moved the entrances to Altamont Terminal from the hidden yard under Altamont City and placed them on either side near the entrances to the hidden yard. This facilitated a consolidation of blocks using only three, where originally I had used five. Likewise, I added a crossover to access the hidden yard siding entrance at Fitzhugh, and consolidated the block placement in that area thus creating two more free blocks. I used the four extra blocks for the streetcar line (2) and for the leads on the southern runaround at Altamont Yard ladder tracks (2). All this switching of blocks necessitated a reworking of the BDL168 wiring and a more logical re-grouping of blocks within the various power districts. Here is the revised switchboard for the new trainroom configuration.


Turning my attention to the benchwork, I removed the walls at the entrance to the trainroom, thus eliminating two sheetrock backdrop radial walls. I think I will achieve a cleaner look with the mountains at Fitzhugh and at Westridge nestled up against a higher, contoured faisure board instead of having a large radial wall protruding out into the room. I will still have some access inside the tunnel from behind the backdrop wall, although this will be more limited than in the previous design. I will probably have to make part of the mountains removable as I have done in the the corners of the A&BR1. The sketch below illustrates my general intent for the revised entrance.


Finally, I reworked the benchwork to employ standardized one foot radii in all the bench corners along the cental asiles. I also  located the four under-the-floor wire conduits. Below is the revised trackplan and benchwork schematic. (Note 3-3-2016: The problem here is still the 2 1/2 foot wide aisles. As noted in the final, I as able to expand these out to 3 feet, hich is in my mond the minimum acceptable width.)


Accessing the Hidden Yard

One of things I most want to incorporate into my new A&BR2 is good access to all track in tunnels and hidden yards. The large removable homosote framed mountain forms I have used on the A&BR1 work well enough, but they are cumbersome to remove and replace. I plan to use a similar scheme to assure full access to the corner tunnels of the new room and in the tunnels in the turnarounds by the door. There will be pretty good access to these areas from behind the backdrop, and so there should be fewer occasions when I will need to remove the mountains. Still, in the case of a sever track maintenance problem, I will want full access to all areas, and so I will add the removable mountains in some of the corners just to be sure I can really open everything up should the need arise.

As for the hidden yard, I plan to build Altamont City on a series of hinged covers that will lift from behind the backdrop to fully expose the under-the-city mainlines and holding tracks. The schematic below details this plan. The scale reads in feet, but here it is to be interperted in inches, so it details the 5" elevation of the siding through the city, and the street plan with spacing dimensions for all lateral streets as well as the grade angles of the cross streets. This hinged "lid" with the city atop will traverse the portion of east wall of the new train room that is accessable from behind the backdrop. The same design but with  the hinges  on the other side will travese the short section fot he east wall that is against the block outside block wall. This shorter section will not have the tree lines attached and will hinge up at he wall for acess to the track undeneath from the trainroom side. (Note: I later altered the backdrop wall to bench edge from 32 inches to 36 inches, the below drawing dowes not reflet this change.) (Note: Finished working drawings are shown and discussed in the post of 3-25-2017)
alt streets

Checking Elevations and Grades using XTrackCAD


While creating my track plan, I have paid rather careful attention to keeping grades under 2% and turn radii over 30 inches. XTrackCAD has some very cool features to aid in me in this. I can set the overall minimum radius and maximum grade when I begin creating the plan, and the program will warn me if I create anything that does not conform to these limits. With regard to the minimum radius, this checking takes place automatically, and any offending section of track will appear in red. With regard to maximum grades, in order to have the program check grades, I must manually enter elevations at key points on the layout. Once a length of track is marked with an elevation at the ends of the two outer-most sections, the program will assume a uniform incline and compute the grade percentage. Again non-conforming sections will appear in red.

I did have to tweak my standards just slightly, reducing the minimum radius to 26" and the maximum grade to 2.1%. This does not represent a significant compromise, and only three curves are under 30 inches. Most are well over that. Likewise only one grade exceeds 2% and that just barely.

XTrackCAD also allows one to set the spacing between parallel tracks. I have used 1.25 inches from center line to center line on both mainlines and in yards, with a slightly wider spacing for double track radii in the corners or  in the turnarounds. I may just go to 1 3/8 inches everywhere.

More Track Tweaks


I plan to cover the 180 degree turnaround loop to the west of Altamont Terminal with a low removable mountain in order to hide the tighter than prototypical circular radius of track. As already noted, a 30" radius, while generously larger than the N scale 18" minimum, is much smaller than any radius on the prototype, and I have found that it is best to hide at least a portion of all long turns in tunnels. This loop and its cloaking mountain cover about 32 square feet of bench top. This is an enormous area with not much going on at this point, so I decided to add a siding going up to a small industrial area on the mountainside.

The Luxury of Time

As I continue to tweak the track plan, I become increasingly aware of the value of taking one's time to fully think through all the details before track-laying begins. In my case I have probably a year or sobefore the new building that will house the new train room is completed. So I have a wealth of time to consider exactly what I will do. In all likelihood, the final building plan will be a bit different from the preliminary sketches I now have, so I will probably have to re-do the plan at some point. Still, the current tedious evaluation and tweaking is not a waste of time. All of these considerations will be incorporated into the final plan. The more tweaks I make, the more aware I am that if I did not have this kind of time to fully think things through, I would be building the flawed layout from a flawed plan.

Evaluating a track plan is tricky. One can come at from so many angles. At first I was concerned primarily with traffic flow, but as I progressed I began to consider other issues: access, block placement, signal placement, structure placement, wiring, lighting, topography, scenery, an so on. For example, today I made a few changes in order to move certain track sections away from the backdrop. It is important to leave few inches (at least 4 inches) between the track and backdrop in order to have enough space to create a believable transition from bench-top to backdrop. The farther away track is from the backdrop the more convincing this transition can be, so this is always something of a compromise.  Surely 6 to 12 inches would be better, but a lot can be done in a 2.5 inches space in this regard. Illustrations will follow once I start building scenery. For now I am just trying to place the track is such a way as to optimize the realism of my future scenery.

A year is a long time, and surely tweaks to the plan will become less frequent and finally end. In the meantime, I will continue to  discuss them in this blog as a way of making notes to myself. I'll not document the construction of the new building here except as it applies to the train room, so these blog posts will certainly become less frequent for a time.

Rethinking the Removable Corner Mountains  and Resizing the Corner Radii

One of the problems in lifting out the movable mountain forms in the corners of the A&BR1 is that they fit snuggly against the backdrop wall. Going in and out, they have a tendancy to scape off foliage and texturing material affixed to the mountain silhouettes that are glued to the backdrop wall. An easy fix for this would be to design then to sit on the bench work about an inch out from the back drop wall, and to cover the gap with a curved first tree line silhouette. This would avoid any tight fit, and it would give more depth to the scene at the place where the illusion of depth is most needed, that is, right where the 3 dimensional form meets the 2 dimensional backdrop.


Bridge, Viaduct, Over/Underpass,
and Culvert

Planning ahead for topography, roadways, and structures is critical to the success of any new layout. Generally in the initial stages of planning one need have only some general sense of where mountains, streams, cuts, roads, and structures will be. Initially this can be as loose as simply knowing where, say, a row of warehouses will be and how large the largest of these warehouses might be. This will naturally lead to thinking about how the roadway and street plan will look, where there will be crossings and where there will be bridges and overpasses, etc.  In the case of most structures, the building selection and exact placements is best worked out in detail interactively as roadbed centerlines are drawn on the layout. Likewise, in the beginning only a general sense of where mountains and streams will be is required in order to effectively tweak track planning. Again the exact forms and elevation etc. are best worked out as roadbed is layed out. This may seem imprecise, but I have found that too much prior on-paper, two-dimensional planning with regard to landscape and structures is often a waste of time. It can even get in the way. The fact is, I'll almost always make changes as I see things begin to come together in the real three-dimensional space. I'll go into detail regarding the interactive process of roadbed creation and topo-roadway-structure planning as things progress.

One notable exception to this kind of loosy-goose initial approach to planning for structures, roads, and topography is bridges. It is critical that an initial plan include the description, dimensions, span, height, clearance, and location of all bridges, overpasses, culverts, viaducts, and crossings in order to ensure that track plan elevations correctly conform to the requirements of these structures.  I like to purchase, assemble, and weather all my bridges before the roadbed is laid down. I then install them along with any abutments and piers or pilings as the plywood roadbed runners are constructed.  As this is done, I get a good feel for exactly how the scenery and topography will work around and under each bridge and the look of the landscape where each bridge is attached to the layout.

In the present case of the A&BR2, I have made a list that details the specifications of the14 bridges that I  have on hand from the old A&BR1, the descriptions of the any bridges I plan to purchase, and a track plan detailing where each will go in the A&BR2.

ON HAND from A&BR1 (green on drawing below)
2 cornerstone double track truss/
10 X 2-3/4 X 2-3/4"
1 cornerstone double track contrete overpass  13 7/8x11x5 1/12 .
2 kibri single track arched  34.8 x 5 x 4.7 cm
2 single track volmer arched /6-1/4" x 1-3/8"
1 single track cornerstone tressel
2 Micro Engineering N 75-150 80' Open Deck Girder Bridge Kit / span 6 inches
1 Monroe Models N 9005 Concrete and Girder Bridge Kit (High Density Hydrocal)/ 1-7/8"+ expansion/ H x 6" L x 5" W
3 plate girder (these are code80 NFG for ABR2)

Need New (red on drawing below)
1 long arched double track steel bridge - Faller N 222580 Bietschtal Bridge -MTS 71.99  594x69x131mm
2 double track code 55 plate girder (plate girder bridges on A&BR1 are code 80 - need code 55)  Atlas N 2081 Code 55 Double          Track Through Plate Girder Bridge MTS 14.99 Double Track - 6" x 2-5/8"
1 country road wooden overpass - Monroe Models N 9007 Country Road Bridge MTS  29.19    9-1/2" L x 1-3/4" W x 2-1/4" H
2 Kibri straight stone viaduct (Note 2-16-2016 :This will not work on curve need - substitute double track palte girder ) 7-1/2 x              1-1/2" 

Pier type risers for the 2 curved crossovers of the Altamont City single track overpasses. How many?


Road, Terrain, and Topography on the Revised Track Plan


Now that I have the revised track plan well tweaked, it is time to revisit the issue of roads and mountains. Clearly, marking the locations and dimensions of all bridges and over/underpasses and viaducts suggests a great deal with regard to adjacent elevations. As I did before, I can now roughly color in the topography, not in any great detail, but enough to locate and position the main features - valleys, mountains, streams,  etc.


November 2015

Tortoise Interface

I have been giving some thought to a clean, easy to service Tortoise Switch Machine interface.

First, I plan to change the way I modify Peco turnouts for DCC operation. On the current A&BR, I have attached a dropper to each element of the turnout (both main rails, both stationary closure rails, both pivoting closure rails, and the the frog.) Using seven droppers makes for a bit of a mess under the table. On the A&BR2 I plan to use very fine wire to make jumpers from the main rail, both red and black, to the associated closure rails, both stationary and pivoting, and then attach my droppers only to the main rails and the frog. So I will reduce the number of droppers per turnout from seven to three. (Note: I subsequently decided not to use Peco track and turnouts. See discussion of Atlas Code 55 1-18-2016).

Next I plan to use both a Tortoise edge connector with a bundled pigtail and a 9 position barrier block wired as indicated in the below schematic. This plan includes the addition of two dwarf signals to indicate turnout position operated off the spare Tortoise single pole switch. More on the design of these dwarf signals later.

Tor inter
So the finished interface will look nice and clean and easy to service - like this. Much mo better!
(Note:  I subsequently made a small modification to allow the resistors needed for the Dwarf signal LEDs to be mounted on the block between the input and  the output of the green wire (5 volt negative). Thus, I'll use, not 9, but 11 positions on the 12 position blocks I plan to use. See the February 29, 2016 entry.)


I will build 86 of these interfaces while I am waiting for the new buiding to be completed. Each will consit of a labeled barrier strip attached to a 4.4" x 4.5" mounting board with a Tortoise pigtail attached. This way as track is laid and turnouts are installed, I can measure the length or wire required for the inpit bundle, cut the required wire, and attach it to the barrier strip on the bench. That way the only work I will have to do under the bench is mount the mounting board next to the turnout droppers, attach the droppers from the A and B rails and the frog and from the dwarf signals, and secure the input bundle to the underside of the bench with wire clips.
PM42 and BDL68 Interfaces

Likewise, I have been thinking about clean, easy to service wiring from the PM42 and BDL168 Edge Connectors. This is a very short run, so I plan to use 16 AGW wire to run from the edge connector to the input side of a 22 position Barrier strip, and then use 12AGW wire on the output of the barrier strips.  Something like this, except t
o keep it clean I'll probably use shrink wrap to cover the solder joint at the edge connector terminals.

bdl edge wire

Also, since I have pretty much completed the track plan and named all the blocks and assigned addresses to them all, I can create schematics for each PM42 and BDL 168 interface. Notice the on/off switch on each track of the output BDL168 barrier blocks. This will be a great aid when troubleshooting shorts. (Note: I later decised not inclue this switch in the initial installation)

These two PM42 and six BDL168 interfaces can the prefabricated on the bench on mounting boards - something else to do while I await the completion of the new building.

Moving Ahead

I have made the decision to move ahead with the construction of the new building that will house the new train room for the A&BR2. The architect will approve the siting tomorrow and move ahead with finished plans, which should take a couple of months. We will probably break ground in January or February, so it will be late next year before I am ready to begin work on the new layout. I've done pretty much all the planning I can for now, so this blog may not be too active for the next 8 or 10 months. I will, however, post any new thoughts I may have. Also, I have ordered edge connectors and barrier blocks to begin the construction of the interfaces described above for the Tortoise Switch Machines, the PM42s and the BDL168s. I'll post some photos when I have prototypes complete.

Dwarf Signals

I have completed a prototype for the ground level dwarf signals I plan to use to indicate turnout position of the AB&R2. There will be two of these for each turnout - one for the mainline and one for the diverging line. I am using three legged, 3mm, red/green. LEDs, and I am controlling them using the extra single-pole switch on the Tortoise Switch Machine so that both sides of the dwarf on the mainline will be green when the turnout is in the mainline or "closed" position, while both sides of the dwarf on the diverging line will be red. When the turnout is thrown to the diverging line, the colors will reverse. (See schematic of Tortoise Interface - posted to this blog 11/4/2015). Using my 5 volt lighting power bus and a 200 ohm
resistor (I may play with this value before installation) for each LED, I get good signal light intensity, that is not unrealistically bright. 

I am using .040 inch styrene for the 3/8" square concrete pad/base, a 3/16"' length of 3/16" Plastruct styrene "Square Tubing" for the body of the signal, and a 3/16' square .010 styrene top. The inside of the square tubing body is just large enough to accommodate the head of the bulb of a 3mm led, and I have drilled a hole in the center of the .040 base just large enough to accommodate the flared base of the 3mm bulb. I finished this off with an N scale single aspect signal target from Shapeways who markets a line of products created with 3D printer technology.

The cost of a single dwarf signal is about $1.40 (about $.60 each for the two targets, about $.12 for the led, and $.08 for two 200 ohm resistors). The cost of the styrene is negligible.

I'll need 172 of these if I use them on all 86 turnouts on the AB&R2. Of course there is no neeed to install dwarfs signal on turnouts in the hidden yard, so I'll need 158 all together. All of these can be manufactured ahead of time while I am waiting of the new building to be completed. (Note: recalcuated the number of dwarf signal needed - 4/14/2016. 87 turnouts less hidden yard = 160 less one for each of the 19 crossovers = 141)

(Note 2-26-2016: To date, I have assembled 36 of these. After assembly I tested them all on the little breadboard setup pictured below. One thing I found was that on some there were small areas where the black paint was not heavy enough or on others small gaps between the signal target and the plastic box. In both cases the interior light showed through in unwanted spots. This was easily fixed with a few dabs of thick flat black paint. On the next batch, I'll try doing two coats of the black spray paint and file the plastic face smooth around the to holes here the targets are attahed, then I'll take more care when cementing the targets in place.)
(Note:  2-28-2016: I subsequently made a small modification to the Tortoise interface barrier blocks to allow the resistors needed for the Dwarf signal LEDs to be mounted on the block between the input and and output of the green wires (5volt negative). Thus, I'll use not 9 but 11 positions on the 12 position blocks. See the February 29, 2016 entry.)

Dwarf Signal prototype on the test bread board. The target board on this prototype  is a cut away section of a Shapeways 3 aspect N scale target. I plan to use the round single aspect Shapeways target for the rest of these. 


I've created label forms in WORD for bus wires, feeder wires, and all barrier blocks (the 22 position blocks for use with the PM42s and the BDL168s and the 10 position blocks for the Tortoise interfaces.) (Note I subsequently decided to use the 12 postion Chinese barrier blocks - one for a Tortoise inteface and 2 for PM42s and BDL168s)

Below is a sample of a Tortoise barrier block label. To the left are associated bus and feeder wire wrap labels. To the right are labels for the sides of the Tortoise machine and labels for the bundle pass-through hole in the trough.  Notice the label background the color-coding. Green is for Power District 90-1, that is, the 16 blocks serviced by BDL 95 and the adjacent routes.  (District 90-2/BDL 96 is brown, 90-3/BDL 97 is blue, 91-1/BDL 98 is yellow, and 91-2/BDL 99 is purple as per the schematic above.) (Note: I subsequently decided that becuse of the wire color coding the wire wrap labels and the pin-out for the barrier block are not necessary. So I'll just need the large green label to label the barrier block and the labels on the right to label the Tortoise and the pass-through hole for the feeders.)

tort label

Below is a sample of a BDL168 barrier block label. This one is for BDL168 # 95 monitoring blocks 95-1 through 95-16.
(Note: I subsequently devised a better scheme involving labeling the plastic barrier block covers. See post of 2-11-2016.)

ndl label

Below is a sample of a label sheet for Track Power Bundles. This one is for outputs 1-4 on BDL168#95. The small labels at the top are wrap-around wire labels that will be secured to individual wires with transparent tape. The large labels are for under-bench labeling for the pass-through holes where the feeder bundles will branch off the man bus.  I plan to run and label all the wires and buses, and then twist the main bus bundle in the trough as well as the individual tortoise feeder bundles, before I attach the feeders to the buses. (Note: After rethinking the need to twist bus wires, I later decided not to use the wire troughs in leiu of a pre lebled flat board, spaced wire, mounting scheme. This will be easier to service and install, and do away with the need for the wire wrap labels, and thereby cut out a lot of work. Since this flat mounting board will be labeled, there will be no need for the wire wrap labels.)


Reconsidering Twisting and Bundling of Power Buses - The Bus Wire Mounting Board.

I've have recently had several discussions in forums online regarding the twisting of power buses in order to cut down on impedance and interference problems. According to Alan Gartland's excellent "Wiring for DCC" any run over 30 feet constitutes a potential problem and should be twisted. Digitrax, on the other hand, does not recommend twisting power bus wires, and maintains that the maximum bus run can be up to 50 feet. The NRA holds that the best practice is not to twist after a detection devise, but to maintain 1/2" spacing between all bus wires. To further complicate matters, most of the literature I found regarding the use of twisted wire to cut down on impedance, noise, cross talk, etc. relates to twisted pairs. I have found nothing discussing the benefits, if any, of twisting multiple A wire feeds into a bundle with a single common B wire return.

So what to do?
In the case of the A&BR2, all but 8 of my 80 detected power buses are under 30 feet. Certainly I should twist all 10 AGW pairs between the DCS100 and the PM42 and between the PM42s and the BDL168s. After the BDL168s I could also run redundant B common returns for these 8 runs that are over 30 feet and twist each of those pairs or twist them all into a single bundle, but I suspect this is unnecessary, and it may cause leakage problems after the detection device.  In short, it appears to be the conventional wisdom to twist before a detection device, but not after a detection device. After the a detection device, the best pratice is to try to keep all buses and feeders as spatially separate as possible.

My problem, I think (if I have one at all) is the proximity of wires in the trough. At this point, I think the best solution is to do away with the troughs, and use a single 12 inch-wide plywood  mounting board recessed back and below the edge-of-the-layout fascia board and ruining all the away around the layout. This would, of course, lower the bridge, of the duck-under to get up under the bench by 12 inches, but then, the 4 inch troughs with their associated 3 inch access from above and their 4 inch mounting board below lowered the bridge by 11 inches, so there is not real difference. 

This would allow room for the mounting of the 16 12 AGW Track A power buses from a BDL168, along with one common Track B bus, and a single undetected A power bus for trunouts and routes, as well as the DS64 feeder bundle, a 12 volt DS64 power bus pair,  and a 5 volt lighting bus pair, the flat signal cable, and finally the loconet cable - all spaced at least 1/2 inch apart. I would then strategically place 12 x 6 inch mounting boards mounted on spacers out from this wire mounting board to accommodate the mounting of DS64s and slightly larger mounting boards for other devices. In cross section this plan might look like this:

There are a number of advantages to this system. First, it seems the best way to head off noise and interference problems. Second it allows better access than the trough to all wiring and makes it much easier to wire feeders.  Third, with this scheme, I will not have to individually wrap-label the individual bus wires (I can simply place label on the mounting board.) And forth, it is much easier to build.

December 2015

Wiring Color Scheme


The A&BR2 will conform to the following color wiring color scheme:

Track A Power - red
Track B Power - black
Frog - yellow

5 volt lights + blue

5 volt lights - green
12 Volt +/-DS64 output1 - brown 
12 volt -/+ DS64 output2 - white 

Some Thoughts Regarding Droppers

I have ordered 7 100 foot rolls of 20 agw stranded primary wire (one in each of the above colors) for droppers and for the Tortoise interfaces. I have also ordered 6 rolls of 16 agw primary wire in the above colors (excepting yellow) for feeders.  Certainly it is good idea to use 16 agw wire for the 1 to 3 foot track feeders between the track droppers and the bus. However, it is not necessary to use this kind of heavy wire for feeders for lights and for DS64 outputs. Nonetheless, as an automotive standard, the 16 agw primary wire is considerably cheaper than the 20 agw wire so I'll use it for all my feeders including lights and DS64 outputs. (Note: I later subsiquently decided to use 20 agw for the DS64 connections to the Tortoise machines since the 16 agw will not fit inot the DS64 output connectors. See entry of 2/26/2016. Note: Also I am considering using 30 agw wrappng wire for LED lights since they are to all be in parallel and an individual bulb ith its associated 200 ohm resistor will only draw about 10ma. See entry of 2-29-2016)

On the current A&BR, I used 4 inch droppers made from 20 agw, solid copper, pre-tinned uninsulated bus wire soldered to the underside of the rails and passed  through individual holes in the bench top and then soldered directly to the insulated 16 agw feeders. I then covered the feeder-to-dropper solder joint with a plastic wire nut. This gave me a good, solid solder connection to the rails and generally worked fine, but it was a bit messy with short lengths of uninsulated bus wire exposed under the bench. 

On the A&BR2 on strainght sections of flex track and on turnouts, I plan to use short, 3/4 inch lengths of 22 agw bus wire soldered to the underside of the rail or to the leads of LEDs. These short droppers will be then soldered to 4 to 8 inch lengths of insulated 20 agw primary wire, which will go through small holes in the bench to the Tortoise interface barrier blocks or simply be soldered to the 16 agw feeder buses in the case of  track or LEDs. Using this scheme the solder connection between the short uninsulated bus wire droppers and the insulated stranded wire dropper extensions will be inside the individual pass-through holes in the bench. These hole will have to be slightly larger than usual dropper holes to accomodate solder joints and the insulated 16 agw wire, so I'll fill any gaps around the 20 agw bus wire as it leaves the bench top with modeling clay to avoid ballast leakage.

I place a short service loop below the bench where each dropper passes through to facilated repairs if needed. In the rare case that a rail dropper's or LED's solder connection fails (either at the rail or at the connection to the insulated dropper extension inside the benchtop pass-through hole), I can then pull up the turnout or track or LED to make the repair; or the connection can be restored by drilling a new hole next to the rail and feeding a new bus wire dropper soldered to the old dropper extension up though the new hole and soldering the new dropper to the side of the rail. With a little paint and ballast, it is easy to hide this connection, and this is much better than pulling up the track or the turnout to access the underside of the rail.

Exactly how I will handle droppers on curved flex track is, at this point, up in the air. Soldering to the underside of the rail may be problematic, and it may be best to solder to the sides of the rails after track is in place or to solder curved flex rails together at the joiners.

Tortoise Interface Prototype


I have been experimenting with cheap 10 or 12 position barrier blocks from China, which I got from Amazon. The 12 position blocks are actually cheaper (5 for $6), and they seem to work fine. They are, however, made from very brittle plastic and will not stand much in the way of screw torquing, so hand tightening is the way to go. Here is a photo of the prototype of my Tortoise interface using the dropper technique described above, the Tortoise edge connector with a 20 agw color-coded pigtail, a 10 position Chinese barrier block, and the associated 8 inch 20 agw colored droppers, and 3 foot 16 agw colored feeder wires. Given the color coding and the new bus wire mounting board scheme, the wrap-labeling of individual bus, feeder or dropper wires will not be necessary.

Note: When wiring to the pins on the Tortoise edge connector, notice that on some connectors there are actually two flat pins pressed together,one for each side of the connection, instead of the usual single pin attached ot both sides. In this case, be sure that each wire is attached to BOTH pins, or that the pins themselves are soldered together. If the solder connection is not made properly to both pins, when the unit is stressed or the wires bent, it could cause a faulty connection on one side. To be sure, only one side will be used, but which side is used will depend on the orientation of the turnout, so ensure that both sides are hard soldered together. The first batch of these I made tested fine in the bench, but I experienced several failures when the connector was flexed or placed under a little stress during installation.


Tortoise Interface Cost = $3.45 each (wire ~$.03/foot x 20 feet = $.60 [including feeder wires]
, barrier block = $1.15, edge connector = $1.75). When I begin to mass produce these, I will use 9 positions of a 12 position Chinese block because it is cheaper than the ten position block. Also, I can use 2 of these 12 position blocks to replace the expensive 22 position barrier blocks required for the BDL168 and PM42 interfaces. Lastly, I have ditched the idea of mounting the Tortoise barrier blocks on 4.5'" x 4.5" mounting blocks. There will not always be enough space for these mounting blocks especially underneath the yards, so I'll just attach the Tortoise edge connector pigtail and the feeder bundle to the barrier blocks on the bench and then screw the blocks to the underside of the bench or to the 3/4" plywood roadbed runners. This will allow much more flexibility with regard to the placement of these interface barrier blocks. (Note:  I subsequently made a small modification to allow the resistor needed for the Dwarf signal LEDs to be mounted on the block between the input and and output of the green wire (5 volt negative). Thus, I'll use not 9 but 11 positions on the 12 position blocks. See the February 29, 2016 entry.)

Rethinking the Train Room Lighting Scheme

As I promised back in July, I will now revisit the room lighting question. On the A&BR1, I currently use a Lutron system of dimable florescent tubes. I know that when I say "florescent", there are those who will wince, but really, you have to see this particular installation. It looks great!
This system has a number of things in it favor:
1. The quality and warm color of the light  produced by the Sylvania F32T8-835 T8 tubes is excellent - quite realistic indeed.
2. These tubes are economical to operate, and they put out A LOT! of light - 2850 lumens each. With the lights all the away up, it is indeed broad daylight on the A&BR, and the light is quite diffuse. It saturates the whole layout evenly with virtually no shadows.
3. The dimable ballasts are a bit pricey (about $174 each - one ballast can run three single-bulb fixtures. ). Still, lumen for lumen this is a very economical system, and t
he Sylvania T8s last 24000 hours! I have 11 fixtures on the A&BR1. I'll need 7 on the A&BR2.
4. One can employ reasonably inexpensive (about $10 each) tube covers to filter out the ultraviolet part of the spectrum, and thus avoid the color-fading that comes with fluorescents.
On the negative side:
1. In the lowest position, dimmers do not fully turn off tubes, but they go down to 1%, which is acceptable.
2. These florescent dimmers are not programmable, and I have found no way to interface with the internal electronics to make them so. It is just beyond my depth.
3. The installation of the ballasts is on the complex side, and it takes a good electrician. These can be a little costly to install compared the convention fixtures.

On my current layout, I have overcome the above-mentioned lack of programmability obstacle with a system employing servos controlled by a little program I wrote for the Arduino to move the dimmer faders mechanically. The different dimming routines in this program in the Arduino are triggered individually from the Freiwald TrainControler software via the outputs of a Digitrax DS64 decoder.
As I mentioned before, mechanically this sytem is a real Rube Goldberg, but it works, and so  my complex array of automated TC schedules are regularly complimented by sunrises and sunsets, and even a thunder storm with sound provided by the synchronous 4D Sound surround.

I have been working on a design for a mechanical interface that is a bit more streamlined and elegant.  Below you can see a schematic of the new mechanical dimmer-slider interface.


Likewise I am designing a more direct control interface with TrainController. Electronically, the new scheme is quite simple really.
Servos will take commands directly from programs I wrote for my PC that are triggered in the TrainController software in operations or actions windows of pushbuttons using the "Execute" option. These programs communicate with the servos directly, sending serial strings using the USB and bluetooth. The use of this direct serial interface eliminates the Arduino and the DS64s.

These programs are written using C#, and the code is disarmingly simple. Here is the core code from one of my programs. This is a sunset routine, moving the dimmer faders slowly from their present state to the off (1%) position over period of 15 seconds. The code string addresses a single servo connected to a SSC32 Servo Control Board using the SSC32 serial string protocol. (The code to attach the necessary System namespaces and to create and setup the serial port and the timer is not included here.)

namespace A&BRSunSet
    public partial class Form1 : Form
        public Form1()

        private void Form1_Load(object sender, EventArgs e)
             String codeString = ("#0 P1300 #1 P1300 #2 P1300 T15000 \r");
             SerialPort serialPort1 = new SerialPort("COM3", 115200, Parity.None, 8, StopBits.One);

             timer1.Enabled = true;


        private void timer1_Tick(object sender, EventArgs e)


Here is a photo of the SSC32 Servo Controller Board with one small servo attached. I use this setup to test my servo programs. I just open TC, operate the prescribed pushbutton to call the program, and the servo moves to the prescribed position at the prescribed speed. I have also written a little servo setup program that allows me to address each servo individually after it is connected to the dimmer slider and move it incrementally, using an on-screen slider control, to any location.  This allows me to calibrate each servo to the precise rotation needed. I can then enter the corresponding command into the individual operational program. (The servos are powered by an external power supply that plugs into the jack next to the two switches on the lower left of the card in the photo. In this setup, the SSC32 is powered by a 9 volt battery. When the SSC32 is installed permanently on the A&BR2, the battery will be replaced by external 9 volt supply.)


I recently spent some time looking at alternative lightng optins including the new LED lighting control technologies, especially HUE by Phillips. This stuff is impressive indeed, for it gives wireless computer control (via WIFI) down to individual fixtures, not only in matters of brightness, but in matters of color. The possibilities are endless. I have not looked too deeply into the control software, which appears to be Apple - Android oriented and tailored to cell phone and home lighting needs. Still, this stuff uses ZigBee protocol, and it appears that there are (or can be) windows applications either existent or createable that will do what I want.

The bad news is: the HUE system is really expensive. I did a little quick math. If on the A&BR2, I install a dimmable florescent system like the one I am currently using on the A&BR1, the cost
(not including installation) of the 19 fixtures with dimmers and ballast etc will be about $3000. If I reuse what I have on the A&BR1 on the A&BR2, then my cost will be only about $1750.The cost of bulbs alone (to produce the same number of lumens) on the Phillips HUE system comes to over $4200.

To be sure the Phillips system will be cheaper to install since it uses conventional incandescent fixtures and no dimmers; still the installation cost difference can't be much more than $500 or so.

Is the HUE system better? Yes, it most probably is - even though what I have going now looks pretty darn good. Does Hue give better control? For sure, especially when the notion of color comes into play. Of course this will depend on the software I use. Is it worth it? For me, well ...maybe. I may purchase a starter kit ($199 for 3 bulbs and a hub) and play with it. That way I can figure out what I need in the way of software and see just how bright it gets and how many bulbs I will really need to do the whole layout. I am many months away from needing to decide, and if I wait a bit the cost of the bulbs will likely come down.

More about the Phillips HUE lighting system

The more I think about the new WIFI controlled HUE LED bulbs from Phillips, the more convinced I become that this is the way to go. The ability to slowly change brightness AND color is indeed intriguing. I have spent some time researching the possibilities for controlling this system from the Windows desktop, aiming at applications that can be called from Traincntroller operations windows. Although the software I need doesn't appear to be currently available off-the-shelf, there is a C# compatible API for developers called Q42.HueApi that will run within the Microsoft Visual Studio environment. This API will enable me to write my own custom software to control HUE lights. I have downloaded Q42.HueApi and installed it within a new C# Visual Studio Project. Although I have yet to find a comprehensive code resource, I have nonetheless found enough code samples online to write code that will find the hub on my WIFI network, register as a "client," and address a designated single bulb (or a select array of bulbs or all bulbs) to turn on, dim, change color, and/or blink.  I have not attempted to change the timing parameters of these functions, but that is the next step. Granted, this new program is not finding the HUE hub because I have not purchased the system yet, so whether or not this code actually works is still very much in question. Nonetheless, it does run and compile without error, and that is a very good sign. Bolstered by this success, I ordered a starter kit today (3 bulbs and a hub that will run up to 50 bulbs).

Beyond getting my code to work and fashioning workable sunrise, sunset and thunder storm sequences, there are still other questions to be answered. What is the system delay time for changes in brightness and color change? What will be the best spacing and positioning for these bulbs?
Is the light saturation more directional or diffuse? What kind of shadows will they produce?   How much heat do they produce? What are the unforeseen problems? They can put out 600 to 800 lumens each depending on the color, but can I get by with 50 or less on the new layout? If so, it will greatly simplify the programming. If not, then I'll have to divide each move into small increments and address both hubs before I go to the next move. Otherwise, the hubs will respond sequentially. Long sequences of very small moves will create the illusion of a simultaneous response, but this will entail a lot more code, and long system delays could make this approach unworkable altogether. To be sure, these incremental movements can be handled neatly using nested "for" loops, but that kind of thing can get messy if one is not careful. 

If my little program works, then it will not take me long, once my kit arrives, to answer most of these questions. Since the bulbs screw into any old conventional bulb socket, it won't take long to install a test setup. But my experience tells me that getting my program to run realistic sequences won't be that easy, especially without a good code reference guide for the Q42.HueApi API.

More on this later.

Preliminary Testing of Phillips HUE System

My HUE Starter Kit from Phillips arrived a few days ago. It was a snap to get it running using various free ware apps. I put an app on my cell phone and a different app on my PC. Both allow for easy, grouping of bulbs,  creation of stored "lighting scenes", and easy manipulation of brightness, color, and intensity right down to the individual bulb level. Impressive!

The LEDs, of course, run very cool so heat is not a problem. The max brightness is about what I expected (600 to 800 lumens depending on the color and saturation - about the same as a 65 watt incandescent flood. It appears that to light a layout from an overhead lighting soffit, I will need to space these bulbs no more than three feet apart in order to get the kind of  bright, full daylight I want. With the 3 foot spacing, I do get some unwanted shadows. An object on the layout below, midway between two lights will cast a faint shadow both left and right, and (
if it is not directly below the center line of the light array) a dark shadow in the middle. Therefore a two foot spacing would be better, but this starts getting really expensive! Even then, the light diffusion and overall light quality is not as good a the T8 dimmable fluorescents I am currently using.

My conclusion is that I should stick with the current T8 dimmable florescent system, with the florescent fixtures mounted in the corner of the soffit away from the backdrop wall and angled down on the layout and slightly back toward the backdrop and install fixtures for hue bulbs at six foot intervals in the soffit next to the backdrop wall and angled slightly toward the aisle edge of the bench work. This would avoid any shadows on the backdrop, give me the quantity and quality of the daylight fluorescents, as well as add the color changing capability of the HUE lights for effects. Using this approach cuts way back of the cost of the HUE bulbs required, and allows the fluorescents to "fill" any shadows cast by the more direction HUE bulbs (except, of course when the fluorescents are very low or off - in which case - like at sunset - the shadows would be desirable.

With this approach I will need no more than 20 HUE bulbs for the entire A&BR2 (about $1200).  This along with the $1750 I need for additional florescent fixtures, dimmable ballasts, and bulbs comes to just under $3000. Is the extra $1200 for the HUE bulbs wroth it? For sure! It is really cool, and the possibilities are endless!

How exactly I'll control all of this is still very much up in the air. To be sure I will write custom stand-alone windows .exe applications that can be called from TrainController. Whether or not I write multiple separate programs - one group for the servos that will control the florescent dimmers and one for the HUE bulbs, or whether I will attempt to write one program to synchronously control both systems is a big question, which I will not be able to answer until I fully figure out how to write using the Q42.HUEApi package in the Visual Studio NET 4.5 environment. I've made some progress, but it is slow, because using the Q42 package is taking me to some places I have never been as a programmer. There is precious little in the way of help on the Internet, and learning curve is steep. So, as they say, stay tuned.  

Phillips HUE Program Up and Running

Well, yesterday afternoon all of the stars finally aligned, and I got all the right references added, and all the right addresses and codes and usernames in all the right lines of the Q42.HueApi code, all in the right order and in all the right places in my Visual Studio C# Windows Forms Application, and Voila! the lights came on both literally and figuratively. In the end, it only took seven lines of code to connect to the HUE bridge and issue a command to my three new Phillips HUE light bulbs.

From there on out, it was a snap. I set up a little application to test all the Q42 commands for brightness, color, saturation, on/off, etc., and I plan to use this to fool around with bulb groupings later. I then created a stunning sunset application that can be called from TC. As the main  florescent room lights go down, my new application will slowly fade up the HUE lights while slowing changing their color to a deep red/orange, then when the room lights are completely faded out and everything is aglow in red, the  red/orange HUE lights slowly fade away and morph to a soft and very dim blue/violet starlight. It's awesome! (In the old and more powerful sense of a much abused word.)

There are a few issues.  1) I have not been able to use the new bulbs on ABR1 layout in tandem with the florescent room lights because the HUE bulbs have a very limited range. In the train room, my phone gets wifi from my router just fine, but the HUE bulbs don't, so I'll need some kind of router repeater. The published range of these bulbs is 30 meters from the router, but if there are any walls etc. they don't come close to that. 2) When commanded to go to a brightness of "0", the HUE bulbs, get pretty low but do not go all the way off. This is a problem with most dimmer systems for CFL and LEDs, but given the ability to change colors, changing the color of a "0" brightness bulb to a a soft blue/violet, gives a lovely very dim nighttime effect that is actually more dramatic than complete darkness. I could, of course, program the system to simply fade to "0" and then turn the lights off, but the transition from "0" to off is a little jarring. I say "a little" because when the HUE lights are commanded to turn off, they do not just turn off, they actually execute a nice ~ one second fade.  3) The way this system handles color, is pretty messy. One can program color changes using RGB based commands which are six digit strings incorporating 3 pairs of two digit hexadecimal numbers. UGH this is cumbersome in "for" loops! Or one can use a somewhat baffling x,y coordinate system. Either way,  fades across large portions of the color spectrum are tricky, because not all of the RGB or xy colors are producible by the HUE bulbs, and if you enter coordinates that the bulb can't produce, it will simply ignore the command. This is going to take some more study and some getting used to. 4) The configuration of "bridge locator" routines in Q42.HueApi is still something of a mystery to me. I was able to get things running by using a crude software tool provided by Phillips to register with the bridge and then get the randomly assigned username back for the the bridge. I then typed this username and my ip address into to my code to connect to the bridge. Since this will be a single user device, I really don't see this as a problem going forward. 5) There might be timing issues with some future routines. Phillips recommends that commands be at least 500 ms apart and 1 second apart for commands to groups of lights. This does not present a problem for the kind of slow fade ins and outs I require. I  just use a loop to move from a brightness of 255 (all the away up) to a brightness of 0 (minimum) decreasing the brightness in increments of 5 or 10 with a 1 second delay after each. So a fade lasts either 51  or 25.5 seconds - smooth as glass. To produce a faster fade, I would have to make the delays less, which could cause problems, or make the increments larger, which might make them visible and kind of jittery looking.

In all, the HUE system and the Q42 API together are nothing short of wonderful - very flexible and powerful (although not totally transparent) systems that are every bit of what I expected and more! The only problem is the timming issue which limits my ability to accomplish anything except very slow fades. The error trapping in the Q42.HueApi is excellent, and it usually not only defines errors, but suggests a fix.

More About Marrying the Layout to the Backdrop.

As you may know I have written extensively about how to make a realistic transition from the layout scenery to the backdrop, or more accurately how to hide this transition (see the "Marrying the Backdrop..." tutorial on this web site). I have just returned from New York, where I visited the American Museum of Natural History, and I am here to tell you that, if you are interested in how to hide the transition from a three-dimensional space to a painted, two-dimensional backdrop, you should visit the African Mammals exhibit and study the dioramas. All of the principals I discuss come into play, and even though these dioramas are life size and employ expertly painted backdrops, there is a lot one can learn from studying them.


January 2016

Installing and Programming a Phillips Hue Lighting System on the AB&R1

I have installed a little WIFI repeater to strengthen the WIFI signal out to the train room, and I now have my Phillips Hue bulbs installed the track lighting fixtures in the trainroom and they are all responding to the program I wrote in C# employing the Q42.HueApi reference library. Although the primary lighting source for the A&BR1 is an array of dimmible florescent T8 fixtures, I also have a ring of conventional Halo lighting track with 10 aimible "can" fixtures pointed down at the layout. This dimmible incandescent track lighting is designed to augment and add effects to the main fluorescents. I have been using conventional colored bulbs and a mechanically automated  dimmers of the same servo-powered type I use for the florescent dimmers. It is now a simple matter the lock the track light dimmers to the full-on position (Hue bulbs do not work with dimmers) and simply screw Hue bulbs into the cans. With the addition of a WIFI repeater to get a good strong signal from the house to the Hue bulbs in the train room, I can now call my custom Hue programs from TrainController Action and Operations windows. This all works great with the three bulbs that came with the Hue Starter Kit, and I now plan to add 7 more in order to cover the entire layout.

Along the way, I have figured out the Q42.HueApi bridge locator and start up protocol and syntax, and devised ways to use short (1 to 2 second) pauses in my dimming and color-change programs to avoid potential problems created by the slow-response limitations on the system. I am currently changing color using six-character strings incorporating three pairs of two place hex digits to address color via RGB number. I am certain that there is a more elegant way to address color change, but for my purposes, this works fine.

To get my sunrises and sunsets running perfectly, I am now going through a series of program adjustments to tweak timings between the Arduino controlled florescent room lights and the Windows controlled Hue accent lights and the action on the layout. This has entailed tweaks to both my custom Hue software and to my custom Arduino firmware, as well as tweaks to the delay timings in the TC Action windows of the pushbuttons I use to call both systems. Its a bit of a balancing act, but in the end it works very well and the effects are stunning.

With this "on-the-layout" testing of the system I am convinced that using a similar setup on the A&BR2 is the way to go. I will not need the dimmible lighting tracks, but rather a series of simple incandescent sockets mounted in the soffit about 5 feet apart.

Architectural Drawings for the New A&BR2 Train Room

I met with my architect last week to review he drawings for the new building that will house the new train room for the A&BR2. The one story building will feature two bedrooms on the ground floor with the train room in the basement. It will be built into the side of a rather steep grade so the basement train room will be half under ground and will have its entrance at ground level on the back of the building.


The train room will be 38' x 22'' with a ceiling height of 8 '9".


Below is my drawing detailing the plan for walls (heavy black lines) and soffits (fine black lines). Notice that with the 22 foot wide I have been able to expand all asiles both in fornt of and behind the backdrop to at least 3 feet in width.


Below are the details of the wall sections. Notice that Type B and Type C walls allow a 12" high behind-the-backdrop access to track in tunnels, and that walls adjacent to the behind-the-backdrop aisles have studs below bench level that are 32' on center to allow for easier under-the-bench access.

Below is the basic electrical scheme showing 18 35 watt T8 florescent fixtures and 20 Hue bulb sockets as wells as 8 four-gang outlets. There will be more outlets built into the bench work. These are not detailed here. I expect to break ground within the next two months.


Atlas or Peco Track?

My original plan was to use Peco Code 55 flex track and turnouts on the A&BR2. I have had a good experience with Peco on the A&BR1 and continuing to use it on the A&BR2 will allow me to reuse all the track and turnouts from the A&BR1. This could save me some money if I stay with Peco. Lately I have been looking into the possibility of switching to Atlas Code 55, which not only looks much better, but is less expensive. So switching to Atlas will cost me about $1250 more.

The rub with Atlas is that I may have to re-gauge the wheels on some of my locos and I will definitely have to replace a lot of cookie cutter MicroTrains wheel sets with Atlas-friendly alternatives. I have ordered 5 lengths of Atlas flex track and 2 turnouts, in order to run some tests and see how it looks, feels, and how my locos and rolling stock perform on it. More later.

If I do switch to Atlas and buy all new track, I can keep the A&BR1 up and running at least until I have the bench work complete and all of the track laid and wired (without any Tortoise Switch Machines installed) on the A&BR2. All I have to do is buy a new Digitrax DCS100, a universal panel, and a throttle ~$275 (on a layout this size, I will need 2 universal panels and 2 throttles and I should probably have a spare booster on hand). I can use my spare BDL168 for testing as I lay track on the A&BR2. If the Atlas track tests well, I think this will be the way to go.

Designing and Locating the Main Lighting and Electrical Panel


I plan 4 15 amp circuits for the new train room and 2  20 amp circuits. Power will be supplied from a sub panel localed on the east block wall as indicated in the below electrical plan. Power will be run from the sub panel to the Lighting and Electrical Switch Panel through the backdrop wall just below the bench supports.

The Lighting and Electrical Switch Panel will have a dimmer and 2 switches:
1. 1000 watt dimmer for 15 amp circuit # 1 for the 19 - 35watt florescent fixtures (7 dimmible ballasts)  
2. Switch for 15 amp circuit #2 for 19 Sockets for the 9 watt Phillips Hue bulbs 
3. Switch for 15 am circuit #3 for 8 2 gang recepticals under the bench. These are for all layout powered devices including the DCC booster, and other DCC components and for layout LED lights and other on-the-layout power needs. This will be the Master LAyout On/Off Switch.

The 8 4 gang wall recepticals on the peripheral block walls will not be switched but rather be wired directly to the breaker panel. These are for  the main computers and monitors, work lights, power tools, vacuums etc.

In addition, the little LG HVAC unit will require 2 20 amp circuits.

This is all pretty straightforward except for one consideration: the SSC32 board that I will use to control the servo motor connected to the main room florescent dimmer slider will have to be connected to the main computer via a serial interface. The length of this serial cable run along the bench, down under the floor, and back up to the panel will be about 20 feet. This is near the maximum recommended length for a serial cable. However, I have run them much farther without any problems, so this should work fine. If not, I can always mount the SSC32 half way in between the SSC32 and the computer, and thus shorten the serial cable and then extend the power and control wires to the servos; or I could use a bluetooth connection to the SSC32.

Switching to Atlas Code 55

The Atlas Code 55 flex track and the #7 turnout I ordered arrived a few day ago, and I must say they look GREAT! A significant improvement over the Peco product. My, the little turnout is delicate compared to the Peco turnouts! After testing a number of my locos, I find they all seem to run well, and will probably not require much if anything in the way of wheel re-guaging. I will however have to replace a lot of MicroTrains "cookie cutter" wheels, but this is no big problem. Based on these tests and on the fact that switching to Atlas, allows me to keep the old A&BR1 up and running until I am finished laying track of the A&BR2, I have decided to switch to the Atlas track.

The question now is are the internal connections, in the Atlas turnouts reliable enough for me to forgo any soldered bonds to the point, closure, and frog rails. If so, this will save a lot of work. After querying a few forums, and consulting with a few experts, I am inclined to just connect the outside rails to the power buses, and hook up the frog without additional bonding. I will deal with any subsequent failures as they occur by dropping new wires to the bus if needed. I just hope I don't develop a lot of intermittent connections. According to my research, I don't think I will.

I have ordered 3 more turnouts, 1 more #7 and 2 #10s, so I will have one each of the turnouts I need for the A&BR2. I am anxious to paint and weather a few of these to see how that goes, and to make my templates for the actual laying out of the track, More on that later.

So, for 10,000 inches of Atals code 55 flex track and 86 turnouts (51 #10s and 23 #7s.) I'll need about $2100, that is $850 more than if I reuse my Peco track and turnouts. With the additon of a second DCS 100 booster, a second PM42 (About $400), I will be ready to lay track, and the A&BR1 can continue operations until I am ready to begn wiring and install Tortoise switch machines, BDL168s, and DS64s.

Weathering Track

I weathered three sections of the Atlas Code 55 track today, and it looks very good indeed. I'll weather a few turnouts soon and set up a more thorough test to elavauate how all my loco types run on the Atlas weathered track and turnouts just to be sure before ordering my 86 turnouts and 10,000 inches of track. When the new Atlas track and turnouts arrive, I can then weather all the turnouts and all the track that will not be in tunnels (about a quarter [2500 inches] of A&BR2 track will be in tunnels) while I am waiting for the new trainroom to be ready.

I have a number of projects like this that can be done ahead , including continuing work on the construction and wiring of 86 interfaces for the Tortoise Switch Machines, and 5 BDL168 interfaces , and two PM42 interfaces (projects discussed earlier in this blog and already underway) and the fabrication switch tempaltes and cork road bed and switch pads from 1/8 inch cork (more on this later).

A word about my techniques for weathering track. I am not a fan of rail blackening solutions, because, for one thing, I do not like to use anything more abrasive than Bright Boy eraser to clean my track. Smooth, unscratched rails make for a much better electrical connections and are much easier to keep clean. So I weather track using spray paint. Lay a track section to be weathered on a flat surface with plenty of light. Shake well and have ready 1 can of flat, dark dark brown spray paint (I like the camouflage colors, but any flat spray paint will work - DO NOT use gloss or semi gloss or even satin finish), 1 can of medium flat gray, 1 can of flat light brown, and 1 can of flat white. Begin with the dark brown - 2 nice smooth steady passes at a low angle on one side of the rails and two on the other making sure you coat the sides of the rails, and then a smooth steady pass from directly above. You want to
just lightly coat everything  but not end up with a heavy wet coating. Immediately, while the paint is still wet, use your finger covered with a soft clean rag to wipe the top of one rail clean. Then recover your finger with a clean section of rag and clean the other rail. Again move your finger to a clean section the rag and wipe each rail top again angling your finger this time so you clean a little of the inside top edge of the rail. Let the paint dry throughly.  Then do smooth quicker passes from above, not to cover but just to dust a mist of gray, light brown, and finally white, cleaning the rail tops and letting the paint dry thoroughly after each pass. The amount of gray, light brown, and/or white you use is a matter of preference as is the order in which you apply these colors. You will need to experiment a little. Go look a some real world tracks. They were certainly  dark brown when they were new, but over the years they have become lighter, and bleached even to a grayish tone, and they are often covered the a fine powder of white dust and spoted with dark grease stains. This is the effect you are looking for: weathered looking ties, and shiny rail tops with dark rail sides sides. 

After the painting, I take a black magic marker with not too fine a tip and lightly draw a black line right down the exact center of the ties to simulate a center line of oil leak stain. This is fairly typical on real tracks. After you draw your line, if it looks too heavy you can rub it with your finger to dull it or your can give everything a final misting of gray or white to tone it down. This is not a science, it is an art - or at least a craft. Finally. I use a very fine brush and some rust colored paint to paint the nail heads. Below is an example of weathered track next to an unweathed piece. You may notice that the side of the rail is not totally covered by the dark paint. This is because I forgot to wash this section of track with soap and water and a soft brush before painting to remove any oil residual from manufacturing. It is always a good idea to soap wash and then dry plastic or metal modeling components before painting.

wt A new section of Atlas Code 55 track above and a weathered section below. You can see how the flat dark brown flattens out the look of the ties and gives a realistic look to the side of the rails, and how the gray misting dulls the color a bit ( I did not use any light brown on this section of track. Finally notice that when I sprayed the final dusting of white, that the spray can did not put out a fine mist. Rrather, it  produced a very fine spatter. I kind of like this effect, and I wish I could get it to do that every time.

I don't do any more than this until after the track is laid. Then, after I ballast, I generally go through and add spills of coal and lime etc. and patches of dry grass and dirt here and there among the ties. At this point I also use paint and light colored washes to simulate all manner of accents, and spills, and dirty spots on the ties. Not too much of this, but a little bit here and there adds a lot of realism. 

Testing the Atlas Code 55 Turnouts 

I set up a test layout consisting of one Atlas Code 55 #7 Turnout with one length of Atlas Code 55 flex track connected with conductive metal joiners to each end of the main line and one connected to the siding. I then powered up the rails using test leads from my DCS 100. I then ran all my locos at slow speed and then at a faster speed through the turnout in both directions, first in the closed and then in the thrown position. Almost all of my switchers and road switcher are Atlas locos, most Alco RS units. I also have a number of Intermountain F Units (3s and 7s both A and B types.) Finally, I have 2 identical  Kato E8/9 units. None of these locos is over 6 or 7 years old, and they all preformed fine across the new turnout.  All, that is, except one of the Kato E units, which consistently derailed when traversing the thrown turnout in the forward direction (backing the thrown turnout from either direction worked fine.) My first reaction was that if my entire roster ran on the new turnout, and only one loco derailed, then the problem must be with the one loco. But after making careful measurements with a track gauge and a micrometer on the offending unit, and carefully comparing it's wheels and trucks to its identical sister unit (which ran the turnout fine), I could find no difference.

Studying the derailment at slow speed, I could see the offending unit was riding up over the curved closure rail instead of making turn. A careful examination of the turnout revealed that there was a slight misalignment at the pivot connecting the point rail to the curved closure rail. If I ran my finger along the point rail and across the tiny gap between the point and closure rail, I could feel the sharp protrusion of the exposed end of the closure rail sticking out beyond the point rail. I took a very small file and made a few light passes across the exposed end of the closure rail (more to kind-of round the end, than to narrow the rail width, and viola, the loco ran the run out fine.

I then ran a few locos pulling  a few freight cars and then a few passenger cars though the same tests. Everything worked pretty well, but I did get a few derailments with some of the passenger cars. But then I have had trouble with these same cars on my Peco turnouts. These all have Microtrains trucks, but over the years they they seem to have acqired a variety of wheel types, so I think simply replacing all the wheels with new Microtrain wheel sets will solve most of these problems. In general, I do not see a problem with the Atlas Code 55 turnouts, other than to say that, like all N scale trackage and all N scale rolling stock, pretty much everything has to be prefect. What else is new?

Revisiting the Track Plan

It has been over three months since I gave the track plan any thought, but
occasioned by the Atlas code 55 turnout decision, I have lately been revisting the design. Since the Atlas turnouts have been in somewhat short supply for the past few years, I reasoned that, even though I won't be needing them for many months, it is probably a good idea to go ahead and place a track and turnout order. So I began an inventory of what I will need based on the track plan I last modified back in August. This in turn occasioned a thorough recheck of the list of block names and the block wiring list as well as the turnout names list, just to make sure all these lists incorporated all the changes I made back in August, as well as any new changes I currently might make.

After so much time away from the track plan, several flaws immediately jumped out at me, and I saw a better way to layout several sections. Specifically, I added two more sidings to the hidden yard underneath Altamont City, thus creating a dedicated siding for each of the four mainline tracks that pass through the hidden area under AltamontCity, and thus negating the need for eastbound and westbound crossovers to access hidden sidings. (These sidings are very long indeed because I am reluctant to place turnouts in the area underneath Altamont City. However, once the bench work is complete and hinged top that will support the city of Altamont and cover the hidden yard is in place, I hope to find that there is sufficient access overhead to allow the turnouts to be moved into the hidden yard under the hinged lid, and thus shorten the siding lengths significantly.) I also completely reworked the reverseable cross track and added a couple of crossovers at Altamont Terminal using a scheme similar to the one I used at Altamont yard. I also moved a few crossovers closer to the sidings they are to service, thus making the associated blocks longer and connecting routes shorter. The below switchboard from TrainController will give an idea of these changes. 


All of these represent significant operational, electrical, and conceptual improvenments, and the entire episode has made me accutely aware of what a luxury it is to have over a year to plan a layout as complex as this one. Certainly, putting it aside for a few months and then comming back to it has been benificial. It would not have been wrong to build it the way it was, but it is definatly better this way.

Reworking and Double-Checking the Block and Turnout List Against the TC Switchboard

I spent quite a bit of time today rearranging, double checking and updating the TrainController switchboard to reflect recent changes and checking my block and turnout lists as wells as the BDL interface schematics to ensure that they all agree in every way with the updated TC Switchboard. In doing this, I made more changes, swiching blocks around among the 5 BDL168s in order to keep all blocks served by any one BDL168 as close together as possible and thus keeping the wire runs from the BDL168 to the track as short as possible.  This is tedious business involving close attention to exact block names, locations, and addresses and to exact turnout names and addresses. These lists and the switchboard graphic will be very important when I begin to lay track.  I may not update the XTrack CAD file further because I will not use to create full-scale templates when I lay track as some do. Replacing the Peco turnouts with the Atlas code 55 turnouts
in the XTrack CAD will  involve pretty much starting over on the XCAD plan, and I really don't need the overview anymore. The overview Xtrack CAD program was useful in the initial design for making sure the layout conformed to track radius minimums and grade percentage maximums, but now that I am down to the fine tuning and I am sure everything will fit together within specification, I don't have to worry about updateing the overvierw CAD renderings. One reason I don't like to use the CAD rendering  for the final layout, is because all the curve radii (after the easements) are perfect  circular arcs. I have found that making more sweeping elliptical arcs is much more realistic. Much more on this technique when I begin laying  track laying. Still, before I begin any tracklaying I will make an new fully detailed XCAD rendering working section by section to be sure everything fits as expected with the new turnout lenghts. I'll then use this final rendering as the definative referance for turn out placement when I begin to lay track. 

For the record, below are the finished block and turnout lists. From these block lists I will now update the BDL Interface Schematics and from the Turnout Lists I can organized the DS64 addresses and unit locations when I begin to lay track and install the DS64s.
b95 b96
b97 b98

Turnout Lists: Note: Since I have not yet assigned DS-64 outputs to the turnouts, the turnout names below do not yet include the DS-64 postion information or the turnout digital address. eg. WB Hidden Yard Siding/57-1/1100.
tf tw
te TL

Turnout Templates

Now that I have one each of the Atlas Code 55 turnouts I plan to use, I can make templates for later use in laying out the track plan onto the bench work. These templates are simple 2 1/4 inch-wide strips cut from .040 styrene plastic sheets. Their length is the exact length of the turnout, and their ends are cut and notched to allow me to line up the turnout with the existing rails or center lines of mainline and siding tracks, and to mark drilling hole centers for the frog wire, both A and B track power wires, and both the left and right holes in the throw bar when the turnout is exactly centered between thrown and closed.

To construct a template I first cut away two plastic rail ends on the turnout, one on either side, adjacent to the guard rails to expose the under side of the stock rails where I will attach my droppers. I then place the turnout, rails-up, on the scanner glass, close the cover, and make a paper copy of the turnout, making sure that the printer is in the 100% size mode with no "fit to size" or other algorithms that will effect the size of the image. I check the printed  image to ensure it is exactly the same size as the turnout, and then carefully cut it out to create a paper template 2 1/4 inches wide that is trimmed at the ends to exactly match the rail ends of the turnout depicted. I then cut a piece of styrene to exactly match this paper template and glue the paper, image up,  to the styrene template using rubber cement. Since I scanned the #7 right hand turnout rails up, the side of the plastic template with paper graphic copy showing will be the bottom, so I write "TOP" on the other side. Using the glued-on paper copy as a guide, I then cut narrow 1/16 inch-wide slots about 1/4 inch long to match all six rail-ends and mark the three center lines. I then drill 3/32 inch holes at the points of the frog ring, at the two droppers beneath the outside rails at the point where I cut away the rail-ends, and at the two holes in the centered throw bar. Lastly, I label the top of the template "RH #7."

After the track centerlines are drawn on the bench work, I can use this template to precisely layout the position of all RH #7 turnouts. Then, I will use it again later when I lay track to mark on the cork switch pads the drilling hole centers for the A and B power droppers and the frog droppers and for the Tortoise spring wire, using either the left or the right hole in the throwbar. (More on cork roadbed and on this final line up and on track laying later).  Below are photos of the bottom and the top of the template.

tb .040 styrene plastic tempalte with the scanner copy image of the botom of a #7 RH turnout glued to it.
tt #7 RH turnout above and below, the top of the #7 RH tempalte labeld and with center guides drilled. 

Tweaks to Hue Lighting Sunrise and Sunsets Completed

I have finally finished tweaking the sunrise and sunset routines on the A&BR1. This has been a bit tedious owing to the fact that it involves incremental timing changes to three interactive elements at once: the .exe applications I wrote to control the Phillips HUE lights, the Arduino firmware I wrote to control the fluorescent room lights, and the TrainController Operations/Action Lists that trigger these programs. On the A&BR2, I will be able to address the fluorescent room lights via a serial interface, so I will be able to write an .exe application that can address both the HUE Lights and the fluorescent dimmer servos, thus avoiding much of the tedious interactive tweaking process. Most of the tweaking will be done in the .exe code, and all TC has to do is start a single program using a pushbutton operated either manually or via an operation/action list. 

Some Thoughts Regarding Cork Roadbed


Having just spent a little time discussing turnout templates and reflecting on future track laying,  I  now find myself think about roadbed. Far and way the most popular material is cork. It is inexpensive, it looks great, it takes ballast well, it is super easy to install, and it very effectively deadens the sound of the trains that run on it. Most people use a product manufactured by Midwestern Products along with various types of cork switch pads. The Midwestern Products roadbed 36" strips come 25 to the box and cost between $18 and $20 per box for N scale. The switch pads run about from $2 or $4 apiece. This doesn't sound like much, but on the A&BR2 I'll need about $240 worth of roadbed, and another $200 or $300 worth of pads.  So $500!

Luckily, for reasons that go beyond the cost, I like to use 1/8" sheet cork in 36" wide 30 foot rolls at $75 per roll. One roll will be more than enough for the A&BR2. Granted, it is little extra work to slit and bevel the stuff, but it goes pretty fast when you get the hang of it. Best of all, using this material allows you to fashion switch pads that exactly conform to your turnouts and to the turnout templates you made. This saves some time at track laying time, and it ensures that all the holes for Tortoise spring rods, and wire droppers are in the exact right place while allowing enough play in the positioning of the turnout to get a good, snug rail joint. More on this later when track laying begins.

(Note: added 2/202016. I have treid to set up a system for sliting the sheet cork into 1 1/8 in wide, two-piece, beveled-edge strips and I find it very dificult to maintain any kind of reliable uniformity. The deminsions and perscision of the cuts are critical in order to be able the install the cork exactly along a center line drawn on the homosote base, and any variation in the width of the strips might cause a problem. You can not be too percise when it come to laying N scale track, so I will use the Midwestern Products roadbed for the track base after all, but I will custom cut the turnout pads from 1/8 inch sheet cork to conform to my templates to aid in positioning turnouts. By the way, I have also found that the Midwestern cork roadbed is about .040 of an inch thinner than the sheet cork I have. This is most likely not  a problem, but I'll probably shim the transition with small pieces of sheet styrene just to be as precise as possible. Then I'll bevel trim the edges of the pads to size after the turnouts are installed. At that point no critical cutting persision will be required. More on this later.)

February 2016

An Easement Template

With turnout drilling templates so recently on my mind, my thoughts now turn to another kind of template I use in track laying: easement templates. An easement is a curve that starts out very gradually at the point of beginning where it joins with a straight section of track and decreases its radius (ie. becomes sharper) as it progresses out and away from that point of beginning until it matches and joins the radius of a fixed radius curve. Using easements allows a train to kind of ease into a curve as opposed to going directly from straight track to any fixed radius track. There are a number of ways to incorporate easements into a track plan. I like to use an easement template like the one pictured below, which I downloaded from the Internet.
It is important to get the scale exactly right when you copy any downloaded easement template image. I like the one pictured below because it has the scale marked on it in one inch increments so I can be sure no unwanted or inacurate scaling has taken place in the downloading or copying process.


Any number of algorithms can be employed to produce any number of different easement templates, depending on how radically the algorithm changes the radius from one point to the next. The curve above is just one that seems most useful to me - not too steep, not too long etc. There is a lot of math to this and much discussion on the Internet, but for an N scale model railroad, one need not get too bogged down in all of this. The fact is, unless you have some impossibly tight curves, you probably don't need easements at all. Sure, theoretically your trains will run better with easements, but you are unlikely to experience derailments unless you run  into a very tight curve at very high speed. On the A&BR 2, my minimum fixed radius will be something on the order of 28", so I could easily get by without easements, but I use them on any curve with a radius below 40 inches, not because I am a rivet-counting, prototype fanatic, but because easements make for more realistic looking track work.

Once I have a properly scaled paper copy of my easement template, I glue it to a sheet of .040 styrene, copy the radii arrows and mark each (in this case, 40'' through 22" radii) below the centerline curve with a pencil, and then cut the plastic and the glued-on paper along this centerline curve. I use this curve to scribe the easement's center line on the layout.

When I lay track, I lightly draw a, say, 32 inch diameter curve and the adjoining straight section just the way I would if I were not using an easement. Then I place the template so that the straight end is parallel to the straight section I just drew and I then place the arrowhead marking a 32 inch curve on the centerline curve of the template touching the 32 inch radius I just drew. I can then scribe the final easement section from the 32 inch curve to the new, final position of the adjoining, parallel, straight track which will be slightly offset from the original section I drew. You have to kind of plan ahead for this offset, but it is quite small and its all really pretty easy when you get the hang if it. More on this when track laying begins.

Track Plan Revision at Fitzhugh


I have been tinkering with the track plan again, and I have made a few more changes.

When the Atlas Code 55 #10 turnouts arrived, I realized that they were significantly longer than the Peco Medium turnouts that I had used in the original XCAD track plan, and I knew that this might create a tight squeeze at both Fitzhugh and Westridge. Accordingly, I created a new XCAD file detailing just the plan at Fitzhugh using the #10 turnouts. To make everything fit, I moved the South entrance to the four hidden yard siding turnouts to the area under the City of Altamont. I had planned to do this anyway in order to make the holding sidings shorter, and once I determined that the design of the hinged top to be such that I will have plenty of room to service these turnouts should the need arise, I went ahead and made the change, but this made the holding siding too short, and so I had to place the North entrances to these sidings back at where I originally had them North of Altamont Yard. This opened up enough space to fit in the two #10 crossovers into the rather short strainghtaway at Fitzhugh.

I also added a depot siding at Fitzhugh and two mainline crossovers to access it from all directions. All of this required the addition of five additon detection blocks, so I decided not to use the two districts of the PM42 #91 for auto-reversing, thus opening up six more detection block positions on BDL168 #99. I'll use two separate devices to handle auto-reversing of the crossover tracks at Altamont Yard and Altamont Terminal. I have updated and rearranged the block and turnout lists and TrainController Switchboard to reflect these changes. Then I used a similar aproach creating a new schematic to evalute the situation a Westridge vis a vis the #10 turnouts and found everything fit without much alteration. These more detaled XCAD schematics will be essential for prefabricating certain parts of the these two sections, and working section by section, I will eventually recreate an XCAD rendering of the entire layout to use as a reference to determine all trunout loactions with pinpoint accuracy on each of the many prefabricated sections I will create. I will then use these rendering again to position each prefab section on the benchowork right down to a fraction of an inch.


Do-ahead Projects

It now appears that we will break ground on the structure that will house the new train room in about a month or so, and it is my guess that construction will take no less that 6 months. In the meantime, in addition to continuing to pursue the careful planning of the A&BR2, there are a number of things I can do ahead in preparation.

I have already completed about 30 of the 86 new Tortoise Interfaces I'll need,  4 of the 5 BDL168 Interfaces, as well as about 20 of the 172 new dwarf signals I'll need. So I need to construct 54 more Tortoise Interfaces, 1 BDL168 Interface and 2 PM42 Interfaces, along with 152 more dwarf signals.
(Note: recalcuated the total number of dwarf signal needed - 4/14/2016. 87 turnouts less hidden yard = 160 less one for each of the 19 crossovers = 141)


Owing to the supply problem Atlas has been experiencing over the last few years with regard to the code 55 track, I plan to order pretty much all of the flex track and turnouts I need as they become available so I don't run into any delays once I get started. This means that, as this stuff starts to come in, I'll have a gob of track and tons of turnouts to paint and weather. I can also cut cork switch pads and get that chore out of the way.

I can go ahead and order all of the new bridges and over or under passes I have planned, and assemble, paint, and weather them so that they are ready once I start to lay track.

Perhaps most importantly, I want to re-inventory what I have on hand and then revise and update my complete shopping list to reflect all of the changes I have made. I can then identify the sources I want to use for each item and assign future order dates for all of the materials I'll need. I have already done this for the lumber and the track and turnouts, but I now need to work back through all my requirements.

Redesigning the PM42 and BDL168 Barrier Block Labeling Scheme.

One of the things I like about the 12 position "cheapo" Chinese barrier blocks, is that they have a plastic cover. This removable flat cover is the perfect place to attach the fan-out wire labels. So I redesigned the barrier block labels for all the BDL168s and PM42s. Notice the color coding for the various power districts.

Relocating PM42s and BDL 168 on the Layout.


Now that I have had a week or so to live with the revised block list, I can go ahead assign locations to the two PM 42 and the five BDL168s on the layout. I want the two PM42s pretty close to the Booster
and I'll use twisted pair 10AGW for the connection to each. Then I want the BDL168s each in the center of the the area of the track they service. Digitrax recommends that a BDL168s be no further than 12 feet from the Booster using 12 AGW for the connect, but on the A&BR1 I have one case where a BDL 168 is roughly 20 feet from the booster if one includes the run from the booster to its associated PM42. I have experienced no problems with this setup.

On the A&BR2, I plan to have one PM42 right next to the booster and the second about 10 feet away. The aggregate length of the runs from the booster through the PM42s and on to the BDL168 will vary, three are well under 12 feet, one is about 15 feet, and the longest is about 20 feet. On all of runs I'll use twisted pair 10AGW to connect the BDL168s to the PM42s, and I'll take care to keep these runs well away from any source of electrical interference. So two of these booster-to-BDL168 runs are a good bit over the Digitrax recommend length, but in N Scale with 10 AGW, I suspect it will work fine. I ran this all by Digitrax technical support service, and they also think it will work fine, but of course there is no guarantee.


Mounting PM42s and BDL168s

The current plan for the bench work on the A&BR2 is to have a one foot-wide wire mounting board
running all the way around the layout mounted just below and slightly behind the fascia board. To this board I will attach all of my track power bus runs from the BDL168s to their farthest junction point to the individual blocks. These 12 AGW runs will be stapled to the mounting board 1/2 inch apart. This leaves ample room for parallel bundles of DS64 feeders, and a scenery lighting bus, as well as the loconet cable, all of which will be spaced well apart and well away from the track power buses. The plan is to also locate all control devices (DS64s, PM42s BDL168s and SC8s etc.) on individual mounting boards that will be mounted at various place along the wire mounting board 3/4" above the wire runs and separated from the wires on the main mounting board below by 3/4" plastic spacers.

Since I have competed the wiring for the two PM42 interfaces and all five of the BDL168 interfaces, I went ahead and mounted these devices on 12" x 12" mounting board as pictured below
(or on 12" x 24" mounting boards in the case of the two BDL168s that will be located right next to the PM42s). I have yet to drill pass-through holes for input and output wires, as I am not sure whether or not the 10 AGW track power bus wires will fit into the 12-position barrier strips that I am using. If not, I'll have to mount larger two-position barrier blocks to handle the transition. I will build a prototype DS64 mounting board as soon as the two new DS64 arrive.

dmb Top row: 12x12 NDL168 Mounitng Bords
Bottom row: 2 overlapping 12x24 PM42/BDL168 combo Mountng Boards

Thinking About the Bench Work

Bench top Height: Since I began this blog I have been considering raising the height of the top of the bench work from the 42 1/2 inches that I used on the A&BR1 to something a bit higher. The idea is to force viewers to look more across and through the scenery than down on it. This kind of perspective makes things look  more realistic, but it must be balanced with other considerations. There must, of course, be a consideration for children. And there is, I suppose a point at which the bench just simply looks awkward and just "too high."  I suspect, for example that a 52 inch height might just look wrong. So what is the ideal height? Something tells me I can get away with more than 42 1/2 inches. Looking at various forums online, I find that this is much debated and that many layout benchtops are over fifty inches, while many others are under forty. If there is a consensus, it is probably about 48 inches, but this, it is warned, is too tall for small kids, and requires some kind of risers to accommodate younger children. 

Today I was trying to envision the 12 inch wide wire mounting board that will run all the way around the bench just beneath and slightly behind the fascia board. In this envisioning, I had an eye to the "crawl-under-the-bench" clearance. I built a little mock-up on the A&BR1 and found that, with a 42 1/2 inch bench top, the crawl-under clearance beneath a 12 inch high board affixed below the fascia board was 25 inches, workable but pretty darned tight. The addition of another 2 or 3 inches to the overall bench height raised this clearance by 2 or 3 inches and thus made the under-the-bench crawl much easier. This added the little push I needed to induce me to go ahead and raise the bench height on the proposed A&BR2. Accordingly, I have adjusted the drawings to make the finished bench top 44 1/2 inches. I may even go to 45/12inches.

On the A&BR1, I have about 70 running feet of velcro backed black fabric panels that attach to the bottom 1 inch of the bench fascia and hang fown to hide the space beneath the bench. With the raised benchtop and the narrower fascia board on the A&BR2, these panels will hang about 4 inhes off the floor. This isn't a disaster, but it will have to be addressed. I 'll either have to replace these panels, add material to them, or devise a way to mount them below the fascia board and/or use a wider fascia board, which might get in the way of the mounting panels. This will require a little experimenting once the first section of the bench is built.

Plywood: When I was making my shopping list last week, one of the items I listed was the 3/4 inch plywood used to cover the bench work in large flat areas, and also used in the construction of roadbed "runners" in places where the bench work frame is open on top. Two questions came to mind. Is common pine plywood sheathing sufficient, or do I need cabinet grade birch plywood like I used on the A&BR1? And i
s 3/4 inch plywood an overkill? Wouldn't 1/2 inch do as well?

Since all the plywood I plan to use will be covered with homosote, the more expensive grades of plywood are not really necessary for smoothness. To be sure the cabinet grade product has more layers and is thus probably less susceptible to warping, but if I stay with 3/4, this should not be a factor, so the rougher pine product should do fine. I suggest a 5 ply exterior garde with pluggred and and machine sanded surfaces.

But do I really need 3/4 inch plywood for N scale, or could I use 1/2 inch? Despite the  difference in cost, and the fact that the 1/2 inch will surely be easier to work with, in my mind, the question turns on the issue of warping. With bench work cross-members 12 or even 16 inches on  center, I suspect 1/2 inch plywood would be fine. But I do not want to use 12 inch centers or even 16 inch centers. My cross members will be 24 inches apart! I want to be able to comfortably stand between cross-members during tack laying, and also I find that cross-members have an annoying way of constantly getting in the way of switch machines and other under-the-bench installations, so the fewer of them I can use, the better. Still, even with 24 inch centers Murphy's Law will surely dictate that there will be cross-members in the way of many switch machines, so I will just tack in the cross-member supports inplace until the modules containing the switch machines are all installed. This will make it easy to move a support member should the need arise. After all the under-the-bench installation is complete, I will go back and secure all cross members to the frame and snug down all 3/4 inch plywood surfaces to the frame and to the cross members at no less than 12 inch intervals. Too flimsy, you say. Well, really it is not, given the fact that most of my bench work will be attached to the wall. This adds a huge amount of stability to the entire structure, and makes 24 inch centers quite workable. Some folks even go with 32 inch centers. (Note: I may use 16 inch centers on the center section, which is not attached to the wall.) Whatever the case,
the upshot of all of this is this: because my 3 to 6 inch wide roadbed runners will only be supported every 24 inches, I think that it is best to fashion them from 3/4" plywood (probably a sanded pine mid-grade, not the rough sheathing, but not cabinet grade birch either). With a support only every 24 inches, 1/2" plywood  runners may sag of their own weight or buckle up, I don't really know;  but I know 3/4 inch plywood runners will not.  I know because I used 3/4 inch plywood runners with supports from the cross members every 24 inches on the A&BR1, and I have had no warping problems, and the bench work is sturdy enough to climb on. (Note: I had originally planned to build all of the bench work myself, but watching professional carpenter at work day after day has convinced me that it will be better to get the pros to build the basic 1x4 bench framing and attach the legs and the wire mounting boards. They can do better in a day or so what it would take me weeks to build.)

Updates to the Train Room Floor Plan Including Under-the-slab Conduit Locations

I met with my contractor last night, and we reviewed the train room floor plans and the wall sections that I completed a few weeks ago. I have added some supports for shelving in the behind-the-backdrop access areas and a place for two small work benches in the corners. I also located all the electrical outlets, and I have indicated the exact location of the under-the-slab conduit and tweaked the wall sections a bit to reflect the new bench height and to adjust things so that the reach-through access openings along the east wall conform exactly to the drawings I did for the hinged covers above the hidden yard. The completed drawings are below. For clearity, I have omitted the electrical and lighting deatil from this plan as this is not part of the contractor's commission.


A Power Bus and a Prototype Mounting Board for DS64s

Now that I have the mounting boards for the BDL168s and the PM42s completed, I turn my attention to a simple mounting board for the DS64s that will control my 86 Tortoise Switch Machines and my crossing gate flashers. I'll need 22 DS64s all together, and they will be mounted all around the the layout on mounting boards similar to the BDL168 and PM42 boards, that is, they will be mounted on the 12 inch-wide wire mounting boards that will run all the way around the layout, positioned 3/4 of an inch above the wire runs using plastic spacers on the mounting screws. Most of these will have only one DS64, but some, like those associated with large yards will have several. For the prototype, I have chosen to create a board with two DS64s.

Whereas I think it best to power BDL168, PM42s, and SE8c boards individually using the little, 14 volt, 300ma PS14 sold by Digitrax, I will employ a fused, 16 AGW, 12 volt, 6 amp power bus (similar to the bus I will use to power LED layout lights) to power the DS64s. This will keep things nice and neat and avoid a tangle of power supply wires. Note that with a 6 amp bus like this, it is a very good idea to make sure the circuit is fused to avoid any overheating.

I had originally planned to use 16AGW feeders from the DS64 outputs to the Tortoise interfaces because as the automotive standard, it is actually cheaper than 20 AGW. But as I began this mounting board prototype, I found that the DS64 output wire connectors are too small to accommodate 16AGW wire. So I soldered short 20 AGW wires to the ends of the feeders as pictured below.

dsm DS64 Mounting Board with 16AGW transitioning to 20 AGW. 

This worked fine, but upon reflection, it occurred to me that with 22 of these, each with 8 connections, this scheme called for a lot of unnecessary soldering, which I felt could look bit messy and might be subject to the occasional failure, so I considered using 12 position barrier blocks to make the transition. This, however, seemed unnecessary. So in the end, I decided to use 20AGW feeders instead, and I found a source for 20AGW primary wire in 500 foot rolls that was only a couple of dollars more than the 500 ft 16AGW rolls I had originally planned to use. The finished prototype looks like this without the loconet cables, which of course run out of the opposite side of the device from the feeders and will pass though their own, separate exit hole, which will be drilled in the upper left corner of the board as pictured:

Notice the color-coded label (Green for Power District 90-1 BDL168 #95, with the device board number to the left and the list of the 4 outputs including the associated addresses and descriptive turnout names to the right.
Modifications to the Tortoise Interface

One lingering details with the the Tortoise interfaces I am now building is how to handle the two little resistors needed for the LEDs in the Dwarf signals. With the few LED layout lights I have on the A&BR1, I have simply attached resistors to the LED droppers right at the bulb. This leaves very the small, delicate resistive components just kind of hanging unsecured under the bench at random spots. They can be hard to find, and they are definitely in harm's way. Since I am using only 9 positions of the 12 position barrier blocks in the Tortoise interfaces, why not mount them on the block between the green (5 volt negative lighting bus) feeder and where the green droppers come into the block from the LEDs like this?


This makes them easily accessible should I need to change the values, and it keeps things much neater and keeps the delicate resistors less exposed to damage. Remember I am using the 12 position barrier blocks instead on 9 position because they are cheaper - like the 16 gauge primary wire, they are the automotive standard. I'll have to change the 40 or so Interfaces I have already completed, but that is no big deal. 

Layout Lighting Power Bus and Lighting Distribution Blocks

Dealing with the Dwarf Signal resistors set me to thinking about the resistors in the layout lighting scheme. For layout structure lighting, I plan wire 5mm warm white LEDs and a resistor in parallel circuits. The layout lighting bus will be powered by its own 5 volt supply. The LEDs don't draw much current - maybe 10ma each, and I'll probably end up with 200 or so, so anything over three amps will probably do it. To turn the layout lighting bus on and off, I'll employ a relay capable of handling the 3 amp circuit, with a 12 volt relay control circuit so I can use a DS64 output with a diode on the output to turn layout lights on and off from TrainController or from my Digitrax throttle. The use of the diode, of course, transforms the DS64 reversing polarity output into a simple current/on - current/off output to control the relay.

The problem seems to me to be that this kind of wiring can get really messy if you let it, so I think I'll employ small circuit board-moutned bused pins to feed
the LEDs. These boards can  contain the individual resistor at each pin output and be mounted out from the 12" wire mounting board on 3"x12"mounting boards with labels like the one below, just like the DS64s. The multi-circuit pairs can then run out in feeder bundles along the wire mounting board until each wire is adjacent to the area it will service. This way I keep all the connections and labeling on the mounting board and avoid additional connections and labels under the bench. Likewise, blue and green wires can be run out from the mounting board to dwarf signal inputs on the tortoise interfaces. I'll use 30 guage insulated blue and green wrapping wire (very small stuff, but a single LEDs will only draw a little over 10 ma so it should do fine.) There are several advantages to using the 30 gauge wrapping wire. First, it is really  inexpensive, so I can fashion extra long droppers that run all the way to the distribution block and just trim off the surplus. Second, it is designed to wrap around the pins (and I'll add a drop of solder to each) and it is very thin and holds its shape well, so it lends itself to making neat orderly wiring. (See Prototype Layout Lighting Distribution Block entry of June 1 2016)

Note: After experimenting with the 30 awg wrapping wire, I found it too delicate and difficult to work with, so I reverted to a more conventional plan using 20 agw sub-buses that feed of the 16 agw layout lighting feeders and daisy chain to the legs of LEDS in a lighting group of 5 to 20 bulbs. (See entry of 6-24-2016)


March 2016

Labeling the Wire Mounitng Board

One way to avoid having to do wire-wrap labels on all the bus wiring is to label the mounting board to which the bus wires will be attached. The plan is for a 12 inch mounting board, with 24 latteral wire runs spaced 1/2 inch apart. I can a place a 12 inch long label like one the labels pictured below every 6 feet or so, and that sould do it. Of course the labels will be different for each power district.

mbl 24 position Wire  Mountting board labels will be 12 inches long with 24 1/2 inch divisions. They will attach to the board every six feet or so.

Installing Turnout Interfaces


There are a lot of moving parts (or should I say moving polarities) involved in the Tortoise interface. I have been thinking about the best and most efficient way to get the Tortoise wiring right without having to crawl under the bench a half a dozen times. Of course, I will prefab a lot of the the most complex trackwork in modular sections on work bench, so this is only an issue with on-the-layout istallations.

After all holes on the turnout template are drilled to size and the turnout is in place and attached to all adjacent track sections
with appropriatly long red, black and yellow droppers attached,  following these steps will, I think, minimize the under-the-bench crawls.

1. Measure the 6 feeders in the input bundle (2 - 16agw red and black track power, 2 - 20 agw brown and white tortoise power, and  2 - green and blue 20 agw dwarf signal light power.)
2. Install the Tortoise machine with no wires attached and test it by manually moving the throw from closed to thrown to check that the rail closures are snug and perfect. (More on this later.)
3. Solder 1 inch 22 agw bus wire exrensions to the legs of the Dwarf Signal LEDs.

4. Drill three holes for each drawf signal LED and install the two dwarf signals with 1 inch 22 agw extensions soldered in place. 

5. Go under the bench and solder four blue 20 agw wires to the red and green outside legs and two green 20 agw wires to the center legs. Mark the red legs with a red marker on the blue 20 agw wire about 3 inchs down from the each LED and twist the blue 20 agw- red leg from LED #1 together with the blue 20 agw-green leg from LED #2.  Then twist together the other two blue droppers together.
6. Plug the 8 pin female edge connector into the Tortoise male edge connector. Manually move the tortoise to the mainline (closed) position. Use your multimeter to check to see that the frog is connected to the black (common rail) input. If it is connected to red, reverse the edge connector.
7. Attach the track droppers, the frog dropper, and the green LED droppers to the barrier block .
8. Attach the blue twisted pair droppers that connect to the green leg of the dwarf signal on the mainline and the red leg from the dwarf signal on the siding to position #10 of the barrier block and the other blue twisted pair droppers to position # 11. Attach the two green 20 agw wires to positions #6 and #7. Attach a 200 ohm resistor between positions #5 and #6 and one between positions #5 and #7.

9. Secure the feeders to the underside of the bench (the brown and white and blue and green can be bundled together - the red and the black should be run separately each spaced 1/2 inch from any other input wire). Drill the mounting board pass-through holes and attach feeders to the proper bus wires on the mounting board and to the DS64 (I like to leave a little slack on the backside of the mounting board - about 6 inches or so - as a service loop.) Then affix an appropriate wire label to the mounting board at each pass-through hole.

10. Now power up the track and the DS64 bus and the lighting bus power supplies, and check to see if the dwarf signals' LEDs are correct (i.e. the mainline should be green and the siding red with the Tortoise in the mainline [closed] position.)
11. If the colors are not correct, go back under the bench and switch the twisted blue pairs (position #10 to #11 and position #11 to #10).
 Screw the barrier block to the plywood next to the Tortoise machine out of the way of any future installations, and neaten up the dropper wiring using small cable ties and a staple gun as necessary and trimming any unnecessarily long wires to the proper length.
13. Operate the turnout with a DCC throttle and re-check the dwarf signal LEDs and the frog polarity.  Also make a final check of the turnout to insure that you have power to all the rails. Finally run some trains through it - fast, slow, forward, backwards, freight, passenger etc.

Looking Ahead

It does not look like we will break ground on the new building until sometime in April, and exactly how long it will take to complete the structure, I can not guess. In the meantime, I have done about all I can do. I have completed 60 of  the 86 Tortoise Interfaces I will need, and about 60 on the of the 172 dwarf signals required. I have the material to complete 40 more dwarf signals, and then I'll stop. I've also completed the 2 PM42 Interfaces, and the 5 BDL168 Interfaces, and I've designed labels for the main bus wire mounting boards, the Tortoise interfaces, the PM42 and BDL168 interfaces, the DS64, and the lighting distribution blocks. Most importantly I have completed the track plan (with roughed in locations for signals, crossing gates, roads, towns, mountains, bridges and viaducts, and track elevations) at least to the point where I need do no more until I build the bench work and sketch in the track plan in three dimensional space. At that point, I will begin to think in more detail about topography, elevations, and roads, and town terracing etc; and at that point I will probably make a few more changes to the track plan.

I have on hand most of the new Altas code 55 track and over half the required  turnouts, as well as a good portion of the wire needed, and a new 12 volt and 5 volt power supply for DS64 and LED lights.  I'll order a couple dozen Tortoise Switch Machines soon, as well the needed DCS 100 and an extra PM42, so I can begin wiring without having to dismantle the A&BR1 right away. Other than that, there is not much to do but wait, and I suspect this blog will not be too active for the next few months. I plan to upgrade from TrainController Silver to Gold in the next week or so, and I will doubtless be busy getting the A&BR1 running under the new software.

And I am constantly thinking about the A&BR2, and as new ideas and topics evolve, I'll post them here for any who may be interested. I have had a few emails from folks who have been following along
(you can get my email on the home page of this site), and several have offered thanks for the overview and help. I welcome any feedback. A lot of this blog is just me making notes to myself, but to the extent that it is of interest to anyone else, I am gratified. More soon.

Aux Power Supply

The little multi-output power supply I ordered from Amazon has arrived. It is a Meanwell RT-125 (132 watts with three outputs: 5 volts -15 amps, 14 volts - 6 amps, and 12 volts - 1 amp) cost about $38. I plan to use the 5 volt output for LEDs (both dwarf signals, which will be powered by a fixed leg of the 5 volt supply, and structure lighting, which will be powered by a switched leg of the 5 volt supply (individual bulbs wired in parallel each with a 200 ohm resistor- each will draw about 10ma) and the 12volt - 6 amp output to power 22 DS64 stationary decoders (each will draw no more than 300ma). I plan to run these  power buses all the way around the layout on the wire mounting board and pick off of them with feeders as need be. I will also require a switched leg off the 12 volt bus for LEDs in series like the one used in statin platform lighting. I will accomplish the switching  switching using a relay switched by a DS64 polarity reversing output along with a zenor diode.

When the unit arrived, I first fashioned a robust 3 prong AC plug-in wire and attached it to the live, neutral and ground inputs. I then opened the unit up and set the input voltage to 115 volts. It came with a label indicating that the power input switch must be set before use - there were no other instructions. It took me a bit to figure out how to get cover off, but once inside, the switch setting was obvious. With the cover back on, I proceeded to bench test the unit and found the outputs and the green output-LED unstable - sort of blinking on and off. I found no useful documentation for this unit on line, but after reading up on power supplies in general, I discovered that some need a load to function properly. So I attached a 12 volt auto light bulb to the 12 volt 6 amp output and Viola! - all outputs functioned correctly.

I then began to think about where to mount the supply and how to route the aux power buses. I concluded that a mounting board with two 8 position barrier blocks mounted on the wire mounting board under he city of Altamont and adjacent to the center of the long middle section would work best. From this point, I can plug the supply into one of the under bench switched layout power outlets, and I can run 4 fixed buses and 4 switched buses out from the 5 and from the 12 volt output barrier blocks: a pair for the long run to Atlamont Yard, Little River, East River, and all the way around to Westridge; a pair down each side the the center section, and a pair to Fitzhugh.

I made a little schematic to document the hook up:


Here is photo of the finished Aux Power Supply Mounting Board without the relay and zenor diode for On/Off control via a DS64 output. I'll probably add fuses to each output once I measure the normal current draw with everything hooked up.


T Splicing Wire

On most DCC layouts, feeders wires are attached to bus wires using some kind of T splice. One of the things I am unhappy about on the A&BR1 is my soldered T splices.  Sure, they are strong and they work fine, but they are pretty ratty looking. So in my continuing effort to make the wiring on the A&BR2 as neat and as bullet proof as possible, I decided to look into the subject of T splices, because I'll be making a hell of a lot of them.

I first considered using so-called suitcase connectors, but I decided against this because they are designed to join two pieces of wire that are similar in size, and on the A&BR2 I will be joining 12agw to both 16 agw and 20agw, and 16agw to 20 agw. There is a lot of controversy regarding the reliability of these connectors, and those who claim to get reliable performance all say that one must stay within the specifications of the connector to get good results. As far as I can tell they can not accommodate wires of vastly different gauges. So they are not an option, and I would rather solder anyway, that is if I can learn to make a neater job of it.

As it turns out, there are two problems with T joints on the A&BR1. First, the bus wires follow the track around under the bench, which makes for a bit of a tangle, and often renders the T joint difficult to access and create. On the A&BR2, the use of the bus wire mounting boards will insure that all of the bus wires are laid out in a neat rows, and all the the T joints are easy to access. Second, my soldering ttechnique was not the best. I did some research about how make good T splices, and indeed, I found that I had a lot to learn. I am using all strained wire and the reccomended method for joining stranded is a little different that the method for joining solid wire.

Here is what I learned:
1. Removing the insulation in the middle of a bus wire can be tricky, and one must take care not to cut the wire itself, so a little practice on some scraps is a good idea. I use my wire strippers, but for this kind of stripping, I have learned to use the notch 1 higher than the gauge I am scoring. If I am scoring 12 agw, I use the the 10agw notch on the stripping tool. This will not cut all the way through the insulation, but it will insure that you don't cut the copper wire. Then I use a box knife to cut a slice out of the notched area all the way down to the copper. I then do a little more cutting with the box knife and a little pulling with a pair on needle nose pliers will bring the insulation cleanly away along the score cut. For feeder-to-bus joints I strip a 1/2 inch section the bus wire, and then I strip the end on the feeder to expose 3/4 inch of wire.

2. I then tightly wind the bare end of the feeder around the exposed bus wire all the way up to the feeder insulation - about four turns.

3. I then paint the joint with flux paste, and using rosin core electrical solder (not acid core) and a good sized soldering iron, I heat the top of the joint well, holding the iron down on the joint tightly for 5 or even 10 seconds. When the wire is good and hot, I lightly touch the end of the solder to the place where the iron tip meets the copper wire, and I slowly feed the solder down into the the strands. Moving up and down the joint, I feed solder into to all the turns. It is done when the entire joint is lightly covered in a very thin coat of solder. It should be shinny, and the detail of the wire strands should still be visible through the solder. This takes a pretty good bit of solder, but be careful not to use too much. You want enough to get well down into the joint so you don't have to do both sides if you don't want to, but if you create globs of excess solder, it can increase electrical resistance in the joint.


4. Finally, I have devised a way to tape the joint with a small piece of electricians tape notched with scissors so it will wrap both joined wires. You can figure it out.


This may all seem a bit anal, but this is the kind of detail that separates the great model railroads for the good ones. Figuring out and practicing details like this ahead of time will save much time and heartache later and result in a much better installation. I am particularly looking forward to using these techniques, because the linear bus wire mounting boards will allow me to make all the T joints on bus wires mounted right in front of me on the well-lit board, and not hanging over my head in the dark under the bench. 

Signals and Signal Placement

Today, I have again been thinking about signals and signal placement. There are a lot of of ways to approach this, and a purist/prototypical approach is, in my opinion, not always best. For one thing, it difficult to define what is prototypical when so many rail systems in the US and worldwide use any so many different signal protocols and varried sets of operating rules. On a DCC model railroad, as on many real railroads, block occupancy and turnout position are generally the keys
. In this regard, signal placement is of foremost importance, and it is necessary to first devise a complete protocol and a full set of operating rules in order to place and program signals in a way that consistently indicates the status of the track ahead.

I am generally pretty happy with the set up I have on the A&BR1. I use Digitrax SE8c cards to control 3 aspect, double headed signal masts. I can control 16 masts (32 - 
red/yellow/green heads) with one SE8c card. Signal programming in TC can be complex, but both the wiring and the installation of the masts is very straightforward thanks to Digitrax's use of 10 conductor ribbon cables between the SE8c outputs and the individual masts.

In general, the system protocol I use on the A&BR1 and will continue to use on the A&BR2 is this:
1. The upper signal head on each double headed mast is for the main line; the lower head is for the diverging route.
2. If the diverging route goes off to the right the signal will be placed on the right of the mainline track; if the diverginh route is on the left, the signal will be place  to the left of the mainline.
3. Red means either the bock ahead is occupied and or a turnout between the signal and the next block is thrown in a way that will derail the train.
4. Yellow means the next block is free but the block beyond that it is occupied and/or a turnout after the next block but before the subsequent block is thrown in a way that will derail the train.
5. Green means both of the next two blocks are free and all turnouts before the second block are in safe pass-though positions.
6. Other than the dwarf signals, which reflect trunout position only, I will not use signals in yards, so these rules apply only to main lines.

So with a double tracked mainline, like the one of on the A&BR2, signals will generally be placed just after the stop marker at the end of a block 
in the direction of travel only. In certain cases, where there are crossover tracks, there will be signals on a single track in both directions. If there are a lot of turun outs or crossovers in the route ahead before the next block or in the route beyond the next block, this can get a little hairy to program, but programmmed, it all works nicely with TrainController. If the stop markers are set near the end of a block as they generally are, one can observe a train on a automated schedule stopping to observe a red signal, and then continuing on under yellow when the block ahead clears. To get this all right, signal placement is key. One of the best tools for visualizing where to place signal masts is the TrainController switchboard. This simplified, stylized, linear block and turnout diagram makes it easy assess all the posibilities of signal placement, traffic flow, block occupancy, and turnout position in any given section of the layout. Here is the Fitzhugh section with signal icons in place:


I prefer to place and wire and program all signals when I lay track. As I go along laying and wiring track and turnouts in a given section of the layout, I also wire up all associated dwarf signals and signal masts in that section. I addition, I program the section's DS64s, and as I progress, I create contacts in TrainController, enter all the installed block and turnout addresses, and set up the block and turnout properties. I also set the installed block max and min speeds, create the appropriate speed, stop, and brake markers, program station stops, and program the signal heads. I then preform rigorous testing including the running of trains on schedules arond the section under construction. It is easier for me to do all this block by block and turnout by turnout, section by section as I go along - while I am focused on the workings a single block or turnout or section.

A&BR2 Specifications

Elsewhere on this web site I have posted the complete specifications for the A&BR1. I am now far enough along with the planning of the A&BR2 to post a similar spec sheet for the new layout.

Altamont and Blue Ridge Railway2 - Layout, Power, and Control Specifications 
Number of Power Districts - 5 
Number of Power Blocks - 80
Number of Switches – 86 
Track - Atlas Code 55 flex 
Switches – Atlas Code 55
Switch machines – Tortoise 
Turntable - Walthers 130' with DDC 
Max Radius – 28” 
Max grade – 2.1%
Bench Height 44.5""
Depth 34"
Room Corner Radii 48"
Track Power Bus Wiring – 12AG stranded, insulated, max 25’ 
Track Power Feeders – 16AG stranded, insulated, max 3’ 
Track Power Dropper – 20AG tinned bus wire, max 4” 
Computer – (2) HP –4G ram, 2.6Ghz, Windows 10
Interface –RR Circuits Loco Buffer USB 
Train Control Software – Fieiwald Software - Traincontroller Gold F8/5
Power – Digitrax – CS100 DCC Control Station and Booster, 5 amps 
Power Management – (2)  Digitrax – PS42 Quad Power Managers 
Block Detection – (5) Digitrax  BDL168 Loconet Occupancy Detectors and 1 BD4 Occupancy Detector (for separate streetcar line) 
Signal Control – (3) Digitrax SE8C Signal Decoders  
Stationary Decoders – (22) Digitrax - DS 64 Quad Stationary Decoders  
Mobile Decoders – Digitrax, TCS, and Lenz - various 
Control Software - Freiwald Software 4DSound (2)
Computer Playback - (2)- ENCORE ENMAB-8CM 7.1 Channel USB Audio Box 
Speakers – (10) Gear Head Sp2600acb Powered 2.0 Desktop
Dimible Warm Daylight T8 Fluorescents (19) USB Servo Controled
HUE Controled Color Lights (20) Wifi Controled
Custom Windows Software

TrainController Gold

I finally purchased TrainController Gold, and I have spent the last couple of weeks getting the A&BR1 running under the new software. The initial installation and switch-over to the new Gold F8 software was actually remarkably easy. All of my settings came up fine, and everything worked just as it had under TC Silver F7. However, as it turns out, the upgrade for  F7 to F8 involved changes to the way the system calibrates Digitrax decoders, and so I had to re-profile all my locos. This was over-due to be done anyway, so no complaints. Still, it took quite a lot of time. After the profiling, I had to tweak some stop markers, but in the end, I got it all done, and I was then able to export all my loco train settings, including the new profiles, and import them into the new TrainController Gold files that will be used with the A&BR2.

Servo Controlled Dimmer for Train Room Fluorescent Lights

Back in December, I posted an entry to this blog detailing plans for an improved train room lighting dimmer control system. Since I have no way to electronically control the 600 watt Lutron dimmer I use for the main train room lights, I have had to control things mechanically. The current system on the A&BR1 uses a program loaded into an Arduino microprocessor to send commands to servos that mechanically move the dimmer sliders. The program is triggered from TrainController via inputs from DS64 stationary decoders. This works fine, but it is a bit cumbersome. The proposed new system will employ a servo control card (SSC32) that will send servo addresses directly via USB from custom C# desktop programs, which can be called from TrainController. This is a much cleaner approach as it takes the Arduino and the DS64s out of the loop.  Below is a my design for the mechanics of this system.


Yesterday I built this little module, using a lever on a block-mounted servo attached to an 8 inch 2-56 steel rod with a plastic ball link on one end and a plastic clevis on the other end. I hooked the servo up to the SSC32, fired up the computer, and with a few alterations, it all worked fine. The alterations involved the spring steel rod, which I had included in the design in order to apply tension to the slider and allow a little play with regard to servo settings. Unfortunately this arrangement tended to twist the tiny dimmer slider causing it bind. I ended up with the clevis directly below the slider attached to the slider with a small piece of wire. With this configuration, the clevis pushes the slider straight up and pulls it straight down. This works fine, but it requires rather exact servo-rotation commands, which were no real problem. I set the servo to the center position and mounted the device board so the dimmer was in the center position as well. I then used a calibration program I wrote to try various rotation commands until I got an all-the-way-up and an all-the-way down command that took the slider to the very end of its throw just as the servo cut off. Below is a photo of the completed installation. Now all of my lighting (both the fluorescents and the HUE bulbs) is controlled by custom desktop C# programs that can be called from TrainController.

ds Servo with lever,  rod, and clevis attached to dimmer slider & the SSC32 Servo Control Board

Programming the Lights on the A&BR

I have made numerous references in this blog to the custom C# programs I am using to control both the fluorescent and the HUE lights on the A&BR. Now that the complete new lighting scheme is fully implemented on the A&BR1 and it is acting as a prototype for the future A&BR2, I will discuss these C# programs in overview. To do this, I'll take as an example the Sunrise program.

Like all my lighting control programs, this is a Windows Forms Application written iC#
n using the Microsoft Visual Studio 2013. In addition to the usual Windows Forms system references, this program also employs System.Threading and System.Threading.Tasks in order to support the use of Thread.Sleep() program pauses. It also uses System.IO and System.IO.Ports to set up and communicate via the serial port, and Q42.HueApi and Q42.HueApi.Interfaces (downloaded and installed using the NuGet manager) to set up and communicate via WIFI with the HUE bridge.

using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Linq;
using System.Text;
using System.Threading;
using System.Threading.Tasks;
using System.Windows.Forms;
using Q42.HueApi;
using Q42.HueApi.Interfaces;
using System.Net;

namespace Sunrise1
    public partial class Form1 : Form

        public Form1()

I next created a string to hold the private ip address of the bridge, which changes from time to time so I will get it each time I load the Form. Since getting information from the HUE bridge requires an "await", it is necessary to use an asynchronous Form Load like this:

        string ip;

        private async void Form1_Load(object sender, EventArgs e)

            IBridgeLocator locator = new HttpBridgeLocator();
            IEnumerable<string> bridgeIPs = await locator.LocateBridgesAsync(TimeSpan.FromSeconds(5));

            ip = bridgeIPs.ElementAt(0);


The form loads but is not visible on the screen, since I have changed the Windows State property from "Normal" to "Minimize."

Now that I have the HUE bridge's current private ip address, I can begin my Sunrise() sequence by connecting to the bridge using my user name. This is a long character string that I got from the bridge using the little debug application on the Phillips HUE site. Unlike the private ip address, it does not change unless the bridge is reset, so here I just perminently plug it in programmatically instead of having the program get the information from the bridge each time at start up.

        private void SunRise1()
            //initialize connection to HUE bridge
            ILocalHueClient client = new LocalHueClient(ip);

Next, I set up a new light command,

            var command = new LightCommand();

Now I am ready to manipulate the lights. For my sunrise, I first turned on the lights at the lowest level of brightness and set the color to a nighttime deep-blue. There are several ways to specify color in Q42.HueApi. They all appear pretty cumbersome to me. I have elected to use Hex pairs. The Color below, 6F1BFF, represents three two place, HEX numbers, 6F, 1B, and FF, that is, the RGB for the deep-blue I selected 111, 27, 255.  Jesus!

            ////All lights ON/Dim
            command.Brightness = 0;
            //All Lights ON/blue           

Next I slowly turned up the brightness to a little over 1/3 intensity (90 out of a possible 255) in order to create a growing pre-dawn blueish glow (9 steps, lasting 1 second each).  Notice that the Brightness function requires a conversion to an 8 bit integer.

            //Fade Up to 90
            for (int i = 0; i < 90; i += 10)
                command.Brightness = Convert.ToByte(i);
I then slowly morphed the color from my deep-blue to the deep red/orange that I will use for the sunrise (10 steps - 2 seconds each). I have not devised a way to change color through the spectrum using a loop, so I selected the colors I want for each incremental step and coded in the sequence manually. (Note: I finally spent a little time investigating the programming of colors  using RGB and another system that HUE uses involving xy coordinates to generate color. See my post of 12-7-2016.)

            //morph color from blue to red-orange




            etc. to



and then slowly increamented the brightness to full intensity.

            //Fade Up to max
            for (int i = 90; i < 255; i += 13)
                command.Brightness = Convert.ToByte(i);

Then I morphed the color slowly from the deep red/orange to a warm white daylight, (in the HUE system, some colors are much brighter than others, so in this case, slowly changing the color from deep red/orange to a warm white, also causes the entire scene to slowly get brighter)

                 //morph color from hot red-orange to warm redish white - 8 moves ~ 16 sec
etc. to       
                                      //warm white

Next, I used SSC32 protocol (servo = #1, rotation = 1720 microseconds, duration = 15 seconds) to send a serial string to the servo telling it to slowly fade up the fluorescents to full daylight over a period of 15 seconds. Servo rotation is controlled by sending a PWM signal along the servo's control wire. In this case, I sent a pulse width, "P," of 1720 microseconds to servo #1, which will rotate the servo a bit beyond the center to my "all-the-way -up" position."  Most servos have a 180 degree throw with 1500 being the center at 90 degrees, 500 being zero degrees and 2500 being 180 degrees. It is generally recommended to work between 750 and 2250 and avoid sending the servo to the limit of its throw. In this protocol, the speed of the servo move is specified by "T," the number of milliseconds for the entire move. In this case I have specified 15000 milliseconds (15 seconds).

            String codeString = ("#1 P1720 T15000 \r");
            SerialPort serialPort1 = new SerialPort("COM3", 115200, Parity.None, 8, StopBits.One);

Once the powerful warm day light fluorescents are at full brightness, the HUE lights become surperfluous and can be slowly faded out and then turned off, and the program can be programatically terminated.

            //Wait for fluorescent room lights to fade up to full

            //Fade Down to Min ~ max speed ~16 sec for 8 moves

            for (int i = 255; i > -1; i -= 30)
                command.Brightness = Convert.ToByte(i);

            //Reset all Lights to Red-Orange and then turn off the HUE lights and close the program           



Layout structure lights and street lights etc. are controlled by using Traincontroller to address a DS64 stationary decoder.The polarity reversing characteristic of the DS64 output is conversed to a current ON/OFF circuit using a zenor diode.
The DS64 output can then operate a relay to turn layout lights on and off. The DS64 layout lights off command is in the same operation window as the execution of Sunrise program, but the layout-lights-off operation is timer-delayed until after the sunrise program has closed using an operations delay set in that window in Traincontroler. 

Some File Management Housekeeping for TrainController

As mentioned in an earlier post, the conversion from TC Silver F7 to Gold F8 went well. It did however point out some errors in my file structure. TrainController, I have come to find, is designed to operate with all associated files in a specified default directory that it creates under C:\\Users\\UserName\\Documents. One can, of course, place files where ever one wishes as long as one updates the links to these files in TC. I had placed all of my TC-associated sound files in a separate directory, and it was not until I installed TC Gold and imported my Engnes and Trains data that I found out that a new TC file would not automatically connect to these files. I did not think much of this at the time, I just updated the unsuccessful links (paths) in TC Gold and went on with a successful installation.

During this time I had also been fooling around with Microsoft OneDrive, at first syncing some WORD files to a backup in the cloud in order to have an ongoing, constantly undated backup. This involved placing the WORD file in the OneDrive folder, selecting it as a file to be synced to the cloud, and then working with that file in the OneDrive folder. This worked great, so I reasoned, why not place all the files I manipulate on a regular basis, like Quicken, Outlook, TrainController etc., in the OneDrive folder, work on them there, and sync them to the cloud so they update a cloud-backup with each change. I tried it with Quicken and it worked fine. But with TC the inability to automatically find the paths to files not in the default directory looked like a problem. It ran fine in the OneDrive folder and backed up fine, but moving it away from the default folder meant I would have to update the links every time I updated it back. 

Well, I reasoned, I'll leave it in the TC default folder, work on it from there, and write a little C# program using System.IO to copy it to the One drive folder automatically each time I close it. Some thing like this:

           string sourcePath = @"C:\Users\Wilber\Documents\Railroad & Co\TrainController";
           string targetPath = @"C:\Users\Wilber\OneDrive\TC";
           foreach (var srcPath in Directory.GetFiles(sourcePath))
               File.Copy(srcPath, srcPath.Replace(sourcePath, targetPath), true);

This worked fine for moving things around and over-writing files to any folder except the OneDrive folder. It turns out that the OneDrive folder can only be accessed by direct user action, and will not accept automated commands.  While I was researching this, I found that it is NOT recommended to use the OneDrive folder to sync Quicken or Outlook or other my other large files that are written to while running. These kinds of applications create just too much activity, and one runs the risk of data loss, file corruption, or worse. So it was back to the drawing board.

The good news is, I learned a lot. I now upload backups to the cloud for all my high-use, currently active files often (at least weekly), and then I do my monthly back up of all non-system files to my Seagate 1TB external hard drive as usual. In the process I have returned all TC-related files for both the A&BR1 and the A&BR2 to the default directory and updated all links, I hope for the last time. Whew!

May 2016

Troubleshooting Loconet After a Lightning Hit

I live on the top of a mountain in North Georgia, and lightning is an ongoing problem. I have taken three hits that rendered the A&BR1 inoperable since it was begun back in 2008, so I am getting pretty good at troubleshooting when my entire system goes down after after a hit. Here is how I go about finding the problem(s).

Generally, if lightning causes damage to any component on a loconet network, that component will corrupt the network signal and render the entire system inoperable. It is pretty easy to find the faulty component if there is just one component damaged. It gets a little trickier if there are several damaged components. 
On my Digitrax system, if a component is corrupting loconet, I get an error message in the form of nine audible beeps. 

As a first step, if you have a computer connected to loconet via a USB interface, unplug the interface and try the system again. I use an RR Circuits device as a USB interface. To my mind, this is the best loconet/USB interface, but it is very susceptible to lightning damage. (The newer units are quite bit more robust in this regard than the older ones.) If the system works after you disconnect the USB interface, replace the interface and you're done. If not, leave the USB interface unplugged, for it may very well still be bad.  

Next, completely unplug loconet from the booster. The first piece of information you want to know is is the booster working, so plug a throttle directly into the booster and check to see if you get track power, then see if you can run a loco from the throttle. If so, you know the booster and the throttle are good. If not, try another throttle, or check your throttle on a spare booster if your have one. If the booster is inoperable
from a know-good throttle and you can't get track power or run a know-good loco, the booster must be repaired before any further troubleshooting can be done.

Next start plugging in the remaining devices in the loco net chains(s). Work one chain at a time, and either unplug devices one at a time working back in from the end of a chain to the booster, testing the system as you go , or go out half way out and disconnect half the devices in groups until you find the culprit. If there is only one device in the chain that is bad, you will find it pretty easily this way. If there are several, it might be best to use a long loconet cable and check each device individually to see if it corrupts the booster or not. As you find the damaged device(s) using either method, take it (them) out (bypass) of the chain, and plug any other chains back in and test the system. If it works. go to the last step. If not, troubleshot he next loconet chain using the same method.

As a last step, once you have found all the bad devices and the system is functioning, plug the USB interface back in to test it.

Lightning is strange stuff. It often selectively takes out only a few compontents in a chain of many. So far, lightning damage on the A&BR1 has been limited to the USB interface unit, signal control cards, and once an occupancy sensor. I have never had more than two devicces wiped out at once by lightning, I have not experienced a booster failure due to lightening, nor have I had any kind of lightening-related decoder failures, although I suspose these are possible. 

DPM Kits

We've had delays starting the new building owing to technicalities with the building permit, County Land Use Board etc. Don't ask! We should break ground soon. So the new train room is still at least 6 months away, and I have done pretty much all the planning I can do. So what's next? Well, I'm going to need a lot more structures, so I guess I'll start building kits.

Anyone who models early to mid 20th century N scale towns and cities should be familiar with DPM (Design Preservation Models) kits. DPM makes a line of simple, nicely detailed kits that generally fall in three categories. Brick stores and shops, brick  low to mid-rise commercial buildings, and small to medium brick warehouses and factories. Typifying small town structures built in the first half of the 20th century
all across America, these kits are easy to build, relatively inexpensive and very versatile. They form the backbone of the town architecture on the A&BR1 where I have scores of them, and they will do the same on the A&BR2 where I'll need scores more before I am done.

DPM Structures on the A&BR1

I decided to begin by building, assembly line style, 16 small DPM stores and shops, so I ordered 16 kits (2 each of 6 different kits, and 4 kits of another design that I plan to kitbash into a single large-ish building) - all at a cost of about $220. What follows is a step by step account of  the mass construction of the 12 kits to create and detail 12 very different structures from 6 different kit designs. I'll deal with the kit bash of the remaining four structures in the next post. There are many ways to go about this, and there is nothing magical about my approach. It is just the way I do it.

The kits

Working on 12 very similar kits at once requires some organizational discipline to avoid getting the parts and pieces  mixed up. So I begin by getting a large board with 16 small open containers on it. I then open all the kits one at a time and
wash the wall sections in warm soapy water to remove any oils leftover from the manufacturing molds. I then rinse and dry the sections and I place all the contents of each kit along with the kit label in a container. I then go back and, one kit at a time, cut the residual plastic sprue material from the wall sections and trim and sand the edges,  returning each wall sections to its container as I go. As I paint, assemble, and detail, each structure, I will always return it and any associated parts to its container so it does not get mixed up with any other kits. This is very important, because the components of all of these kits look very much the same.

The containers

Paint It Black
The first thing I do is paint the inside of all wall sections black.  Working in carefully laid out groups of 4 kits, I lay out the wall sections on my painting board and spray a coat of flat black straight down on the the insides of the wall sections to ensure that they will be completely opaque when they are lit from within. I then return the 4 sets of wall sections to their respective boxes and go on the the next set of 4. This goes very quickly.

Painting the inside of the wall sections black.

Color Schemes
The next thing I do is to plan my color schemes - 1) the color of the brick, 2) the color of the window trim, and 3) the color of the other wood and masonry trim for each structure. I place a small label on each box to indicate the color scheme for each structure. Of the 16 little stores and shops, 8 will be unpainted brick (4 each in two different red-brown shades), 4 will be brick painted a medium gray, two will be brick painted ocre, one a cream color, and one will be brick painted white. I'll use 6 different colors for the window trim (white, black, light gray, dark gray, dark brown, and dark green.) Some of the bare brick structures will get a light colored wash to articulate the masonry between the bricks. The some of buildings will only get a dark, weathering wash, some very dark and dirty, some not so much. 
All of the window trim colors will be sprayed on first using regular canned spray paint (see below). Then after masking the window trim, I'll spray on the brick colors. I'll use 5 different colors for the wood and masonry trim (white, black, light gray, dark gray, dark green, and light green). They will be sprayed on or brushed on using acrylic model paint (see below).

Spray Paint
I don't generally use an airbrush. Although one can do amazing things with these tools, they are just too much trouble for me in most cases, and when it comes to painting structures I can get excellent results using regular canned spray paint and a myriad of non-airbrush weathering techniques. For plastic structures flat spray-on colors are best. I have had excellent results using primer, which, generally comes in three different very useful, very flat colors: red/brown, medium gray, and white. Also the spray-on camoflauge colors that are now available are excellent and very flat (dusty greens, tans, browns, and grays). But any old flat spray paint works. (If you need a color that you can't find in a flat finish, then you can use a satin finish or even a gloss, and then over paint with Testors Dull Coat.) Apply your spray paint from about a foot away using slow, steady, even swipes from different angles. Do not stop moving, and do not over-paint.  Spray lightly, just enough to get complete coverage. This stuff is very thin, so the paint tends to quickly get in all the cracks allowing all the detail to show through. The coverage is excellent, and it dries very quickly. Since these spray paints are not water soluble, they have the advantage of tending the hold fast when you later use water or alcohol-based washes or rubs to weather or to articulate the masonry in the brickwork.

I keep a variety of canned  flat spray paints, primers, and camoflague colors on the shelf  in my shop. On hand I generally have: black, white, dark green, olive green, light yellowish green, yellow, dark gray, medium gray, light gray, light tan, cream, medium tan, dark brown, 2 shades of brick, light blue, light pinkish purple, as well as metallic aluminum, several gloss coats, a matte finish, and Testors Dull Coat. 

Brush-on Paint
For years I used Testors Poly-Scale Railroad Colors, but they discontinued that line, and so I went to Testors Model Master acrylics. These give very good results if the consistency is just right, but they tend to thicken up after opening, and they are not completely water-soluble (you can clean up with sap and water, but you shouldn't try to thin these paints with water). Recently I've read a little about the water-soluble Vallejo acrylics from Spain, and so for this project, I decided to try them. They available in 100s of colors, and sets are available in a large variety of color mixes. The sets designed for painting plastic replicas of tanks and armor seem to fit the bill (nice, soft, grays and greens and tans and browns.) A number of my Model Master colors are dried up or about to go South, so I ordered a set of 16 WWII armor colors from Amazon at cost of about $45. We'll see.

My Paint Box with the new Vallajo WWII Armor Assortment in the lid.

Painting the Wall Sections
Not too long ago,
in order save the tedium of hand painting the window trim, I devised a technique that involved spraying on the window trim color first, then masking each window, and then over-painting with the final brick color. For years I did this by hand, thinking that it was not possible to successfully mask these tiny windows. But using careful application of regular canned spray paint over tiny window masks cut-in-place using matte finish Scotch Brand "Magic"TM  Tape cut with a very sharp hobby knife and lightly tamped down into place with small stick of balsa, yields much better results than the results I got painting the trim by hand, and it can be done in a fraction of the time.

Again working in groups of four, I first spray paint the brick side of all the wall sections using  the window trim color. Here the window trim will be a nice dark-chocolate brown. I do not spray straight down, but rather from angles to insure that inside of the window frames are covered. Four quick light passes, spraying at angles first from 12 o'clock, next from 3, 6, and finally from 9 o'clock ought to do it. I then let them dry completely.

Cutting the mask for the individual windows is a bit tedious to be sure, but it goes surprisingly  fast once you get the hang of it. The secret here is and very sharp hobby knife and the very thin, not too sticky, Scotch Brand tape used for the masking. It is easy to cut, it  stays in place, is easy to remove almost immediately after painting, and it does not bring any of the window trim color away with it. In cases where the window trim is to be the same color as the other trim, you can mask off the store front and the top cornice as well before painting the brick.

bc Over-paint the brick wall color. Here I've used a mustardy ocre color from my new Vellajo paints, which are great - good coverage, good consistancy, good container. - very nice. Again, if you use spray, just enough to cover. Don't soak it. When the paint is starting to dry but not fully dry, remove the window masks with tweezers. Then after the overcoat is completely dry, touch up with a brush dipped in a tiny bit of the trim color spray paint sprayed into a small cup. Generally, not much touch up is needed, and generally the line along the window trim is much cleaner than anything you could paint by hand.

pt Next paint the other trim, including the storefront, and the stone window headers,  capstones, and sills. This is generally easy, owing to the fact that most of this is painting raised surfaces, so nice, clean edges are easy to get. If the store front is to be the same color as the window trim, you can simply mask it off before you paint the brick. If it is a different color, you can spray paint it by masking off the brick, or you can hand do it. No big deal.

Paint-based Weathering and Dull Coat:
If you are using any weathering techniques that involve paint, or dry-brushed paint, go ahead and do that now.  

I generally use a thin white acrylic paint wash to articulate the masonry of unpainted brick. A few drops of acrylic white mixed with maybe 1/2 tsp water and rubbing alcohol to break the surface tension  - about a 50/50 mixture.  The idea is to get the wash to remain in the recesses of the masonry joints. It is best to use a very light color for window trim if you are going to use the masonry wash.

When any paint-based weathering is thoroughly dry, apply any dry  transfers or decals using plenty of setting and solvent solution to get the decals well down into the texture of the brick. Let the decals dry for at least a day. Then over-spray all your painted wall sections (with or withour decales) with Dull Coat.

Further Weathering:
For weathering painted brick. I use a light black wash of acrylic or ink and sometimes I use pastels or chalks rubbed on by hand to get dusty effects. I apply weathering washes and chalks and pastel after I spray with Dull Coat because Dull Coat or any matte or flat lacquer spray will tend to dissolve the washes, chalks, and weathering and greatly lessen their effect. This means you will have to exercise care in handling these structures because some of the weathering (especially the chalks and pastels) are not set, but getting the weathering effects right is a lot easier if you weather after the Dull Coat is dry and not before your spray it.on.

When everything is completely dry, assemble the wall sections of each structure using a good square to maintain perfect 90 degree angles while the glue dries. Add the roof rails, spray paint the roof flat black on top and leave the underside (inside) white so it will reflect the directional 5mm LED bulb that will illuminate the inside of the structure during night scenes. If the kit comes with a roof that is not white styrene, spray paint the underside of the roof white, let it dry, and then spray paint the topside flat black. After it is dry, install the roof and glue in place. 

roof If you are going to have any stacks or pipes coming up through the roof, drill holes, paint the pipes and glue them in place now. Also glue down any large roof-top structures. After everything is dry, paint the flat black roof top with a rather thick coat of  white glue, and sprinkle on some ballast. I like to use regular gray or multicolored  HO ballast. I like the look of it better than N scale ballast which is just too fine for my tastes. I will also often sprinkle on a spot or two of black or dark gray ballast or brown or tan to keep things from looking too pristine and perfect. I also sometimes leave the black roof top showing through or partly showing through in a few spots, and I sometimes add dabs of black acrylic paint to simulate tar spots or patches etc. Unless you plan some really large roof top structure, you can glue smaller details like pipes and vents, and tanks over the ballast.

Now you can add all sorts of details: the doors, if separate, smoke, colored awnings, downspouts, pipes, vents, water tanks, little shed roofs and porchs, ivy, or other climbing vines - whatever you can imagine. I keep a little box of rooftop details salvaged from other kits, pieces of sprues, and many other sources. For example, florist wire is very handy for making rooftop pipes and down-spouts etc. or a tiny round-headed screws with the slot puttied-in and painted silver or black or rust color makes a cool vent dome. You will be surprised how much you can find, if you just keep your eyes open. Remember, since you look down on the layout, you see more rooftop than anything else, so rooftop detail is very important in achieving interest, variety, and realism. 

Finally, I glue in the window glass and affix any window decals, dry transfers or signs. I like to black out a few windows using electricians tape, put a few shades down or in the half-down position using tan translucent masking tape, and occasionally I remove the mulls and board up a few windows here and there using flooring scraps from etched wooden kits. And last of all, I place an LED inside the structure, turn out the lights, and check for any cracks along the roof and wall seams where light might shine through. I patch these from the inside with putty or with black electrical tape. Et Voila! 

chev b2 b3
grange b5
b6 b7 b9
b10 b11 b12


A DPM Kit Bash
Kit bashing is an art, and it can really get elaborate. I generally don't get too carried away with kitbashing, although I do it quite often. Generally I kitbash structures to make them fit into a specific space, or conform to specific elevations dictated by the terrain. DPM kits lend themselves well to this, and the company even sells modular factory wall sections of various descriptions designed especially helping kitbashers expand, modify, and marry various kits. 

kbmw Here is a large milling factory on the A&BR1. This complex, designed to fit terracing in a steep hillside, was kitbashed from several DPM kits and a handful of extra wall sections. 

Also I often use kitbashs to create so-called "background" buildings, or "backdrop buildings,"or "half buildings" - shallow structures usually placed against or near the backdrop, that give the illusion of depth but actually are only about 1/2 inch or so deep. (See my tutorial "Marrying the Layout to the Backdrop" for informtion about this and other forced perspective techniques.)

bb A "Background" Building. Each  DPM kit will, of course, yield two of these: a front section and a rear section.

One of the shortcomings of the DPM line is that all of the shops are multi-story structures, and although most small towns contain lots of multi-story buildings, they also generally contains lots of one-story buildings.
So I often kitbash to make a few lower buildings to give variety to a town street.

Here, I am kitbashing 4 identical DPM three-story kits to make one large 4-bay building with two one-story bays in the center section and a five-story "U" shaped high-rise surrounding three sides of the low roof top.  Below is a photo of one of the four identical kits used, and a "work in progress" picture of the kitbashed structure taking shape on the bench.

kbkit kbb

kc Completed DPM kitbash. I have only to add the shop signs on the four latteral sign hanging shafts above the doors and place a small water tank to the roof, and I'll be done.

I plan to place this building on a city street, deep in the layout, in a place where the back of the building will not be visible to viewers, so I will just use plain styrene for the back and get the bonus of having the three surplus first story rear wall sections (I used one for the rooftop service structure) to use in making some additional "backdrop buildings."

Windows 10 Problem vis a vis 7.1 Surround Sound Box (ENCAB-8CM)


Several months ago, about the same time I switched to TrainController Gold on the A&BR1, I updated the trainroom computer to Windows 10. Everything worked fine, but lately a problem has popped up involving the ENCAB-8CN 7.1 USB Surround Sound Box. Somehow this unit got reset, and it now appears to be suck in the stereo mode. Under Windows 10, I cannot address it to reset it to the 7.1 Surround Sound mode because the Encore Electronic Company that manufactured the box has gone out of business, and so a Windows 10 Driver update is not available. The original driver installed a control program and it all worked fine under Windows XP. It works under Windows 10, but the control program is not present under Windows 10. So I can't change the settings. Unless I can figure a way to address this little box, I'll have to replace it. Not really a big deal: It only cost about 30 bucks.
My girlfriend has an old Windows 7 laptop that she no longer uses, so a may borrow that, load the old dirvers for the ENCAB-8CN into to it, and see if I can reset the unit to 7.1 using Windows 7. This might work, but how long this fix will hold, I cannot say.

I've got my eye on an inexpensive 7.1 Surround Sound System which includes an amplifier/receiver, and 7 speakers for under $200. I will probably buy this system for the A&BR2. The old ENCAB-8CM and three of its small speakers can then be used in stereo to power a left, center, and right setup from the networked second computer that will display the switchboard and control the long center section of the bench supplying discrete stereo sound to that section independent of the 7.1 Surround Sound for the  peripheral benchwork.

Update: 6-8-2016. I got the USB Surround box working again, by installing the driver in an old laptop running Windows 7 and reseting the software to 7.1 output. Then when I connected it to my Windows 10 laptop, I found the control program, not on the task bar as before up on the control panel. Perhaps it had been there all along. Whatever the case, it worked fine, although TrainController sound dropped out several times before it stabilized and began to function normally. The drop out appeared to be caused by a faulty USB connection. It is working for now, but I suspect there still may be a driver issue with Windows 10.

June 2016

Prototyping Layout Lighting Distribution Blocks

Note: Since I later decided to not to use the 30 agw wraping wire, and switch to a  20 agw daisy chian method, I will not use the distribution blocks detailed below. (See post of 6-24-2016.)
There is nothing on the layout more simple than the wiring for the LEDs used to light structures. I plan to use a simple 5 volt, 6 amp circuit bridged in parallel by a large number of 5mm LEDs each attached to a 200 ohm resistor, and each drawing about 10ma of current. I choose to wire in single-bulb, parallel circuits because I want to keep all the connections on the wiring mounting board and not up under the bench. This power distribution block will also send 5 volt power to the Tortoise Interfaces to power the dwarf signals, in which case, the resitors will be omiited since they are already installed in the Tortoise Interface. On the A&BR2, I will attach each LED leg or Tortoise Interface position to a blue and a green 30 agw feeder and then bring the feeders back in neat bundles to the wire mounting board and along the board to a Layout Lighting Distribution Board where I will attach labels  and resistors as needed, and connect each pair to the  + and - power bus. This puts all connections and labeling (except those at the LED itself or fro the tortoise interface to the LED) in clear view and easy access on the distribution board. These parallel runs will be on average about 2 to 3 feet long, and, given the low current requirement of a single, 5mm LED, the 30 gauge wire will be more than adequate and will create minimal resistance. Also the expense will be minimal because the 30 gauge wire is really cheap, only few dollars for 1000 feet.

lld Here is the prototype with its label and with + and - 30 agw wires and resistors attached to a 24 pin strip, enoungh  for 12 LEDS.

Note: Since I later decided to not to use the 30 agw wraping wire, and switch to a  20 agw daisy chian method, I will not use the distribution blocks detailed here (See post of 6-24-2016.)

This prototye employs a long narrow custom made circuit board with muliple posts. For the negative (green) wires, I have soldered a 30 gauge feeder to each negative post on the top side of the board, with one end of a resistor (mounted on top of the circuit board) soldered to the corresponding post on the bottom side of the board. The other end of this  resistor is soldered to the negative power bus on top of the board.
If the connection is to a Tortoise Interface to power dwarf signals, I omited the resistor and took bottom side post straight to the power bus, since the resistors are already installed in the Tortoise Interface. For the postive (blue) wires, I have attached a 30 gauge feeder to the postive posts on the top side of the board and bussed the corresponding positive pins on the bottom of the board to the positive power bus. So everything is right there in front of me, connections to LED feeders, power bus, labeling, and resistors. I did not wire one side as a single looped-through common return for it would have required larger wire, and a lot of connections under the bench. The long individual 30 gauge wires don't really cost anything to speak of, and the 30 gauge wrapping wire is very easy to install in neat bundles.

I'll now go ahead and build five or six more of these 12 LED, 24 pin boards. How many I will end up needing, I do not know.  I have  more than 150 structures on the A&BR1 with probably about 200 individual bulbs. I'll likely need about two to three times that to complete the A&BR2, so maybe 400 to 500 LEDs (that is probably 4 or 5 amps worth) and maybe 30 or 40 Distribution boards in all. That is all a  long way off  right now. Last of all, I built a little test setup on a breadboard using 4 - 5mm warm white LEDs powered from the prototype distrubution board. I measured the current draw of the circuit at about 39ma, about 10ma for each LED as calculated. 

(Note 6-20-2-16: When I began to install the dwarf signals on the first prefabricated track module for the A&BR2, I found that working with the 30 gauge wire was a real problem. It was just too fine to effectively deal with. So in the end, I did away with it and with the above described lighting distribution board in favor of a more convention scheme using a green and blue 20 agw primary wire sub-bus off the main 5 volt power bus to daisy chain to each tortoise interface block and supply all the dwarf signals on the module. I'll use a similar scheme to daisy chain to multiple LED lights for structures. Labels will be affixed to the underside of the bench at the tortoise interface or where the penetration is made for each LED. Also a label listing all the LEDs supplied will be attached on the wire mounting board where each 20 agw  sub-bus attachs to the main 16 agw 5 volt lighting supply bus.) 

Beginning to Lay Track - Perparing to Construct Prefabricated Sections
6 -7-2016

At this point I am totally comfortable with my track plan, as well as with the bench design, the new XCAD track plan renderings, the TrainController Switchboards, the wiring labels, and the Turnout and Block Lists I have created. With all of this well in hand, I can now begin to lay track even before we break ground on the new train room. Many of the most complex sections of the layout can be prefabricated on plywood/homosote boards, designed to perfectly fit the future bench work and to still allow enough play for small "fitting" adjustments when they are married to the built-in-place sections of the layout. Laying track and wiring is a lot easier on these prefabricated boards because it is easy to light and access and does not require crawling around under the bench work. I did several sections of the A&BR1 like this using a specially fabricated work bench, and it works great if you follow these basic rules. Using this approach, I a get the lions share of the track for the A&BR2 laid before building with the new train room is completed.

Here are some guidelines regarding prefabricating track sections:
1. Plan well. To prefabricate and then successfully install sections of track, all of your planning has to be prefect. Measurements  of the track plan and the bench work design must be exact
. If you prefab sections, there will be little room for modification to scheme after you begin. For this reason, I do not recommend prefabrating track sections to beginners.

XCAD is great for this kind of planning and measurement verification, and after I complete the overall track plan, I generally go back and create a very detailed, separate XCAD file for each part of the layout with special attention to the sections I wish to prefabricate. I don't generally prefabricate a section that is on a  grade, so elevation detail is not often a construction issue here, but I recommend that an elevation plan be in place and that all grades are completely worked out in advance of any prefabriation. Pinpoint turnout location accuracy is essential and special attention should be given to exact track spacing and to electrical continuity and all electrical isolation requirements.

2. Select wisely. The choice of which sections you select to prefabricate is critical. Generally, I choose flat routes between longer blocks, especially if they contain multiple turnouts. I also choose sections that connect to longer sections of in-place track at each end allowing me a just little slop to perfect the connection of the prefabricated track with the rest of the layout. It is best if all the track ends in a prefab section marry to a short section of straight track and then to an easement. This means that, when I move the prefabricated sections around to get it lined up, I can also make very fine adjustments to track sections connecting to the prefabricated section without having to modify the easements of the large radii that they lead to.

3. Take care in handling your prefab sections. I like to build a special work bench that is like a little section of the bench work - about 44 inches high, 1 x4 open framed top - just like the bench work - 2 feet wide and 8 feet long with a cross member every two feet. This bench has its own work light bar and it is outfitted with 5 volt and 12 volt power buses, as well as with track power output from my spare DCS100 and with loconet. With this set up I can fully test each completed section using a DCC throttle to throw turnouts and run locos and even short trains. I generally temporarily screw some kind of wooden cross members above the track on  the top side of each section so that the section does not rest directly on the track where I turn it over to work on the bottom side.

wb Track Layoing Work Bench

I have ordered two sheets of homosote, some 3/4"  plywood, and the 2x4 and 1x4 lumber for my little work bench, as well as some cork roadbed and an additional DCS 100. While I am waiting for delivery, I can get everything else ready.

I'll need:
3/4 " plywood
XCAD print out
turnout list
straight edge
cork road red
white glue
1/8" cork for pads
cork roadbed strips
turnout templates
3 RH #10 Atlas Code 55turnouts
3 LH #10 Atlas Code 55 turnouts
1 LH #10 Atlas Code 55 turnouts
Atlas code 55 flex track
dark brown, light brown, light gray, and white spray paint for wreathing turnouts and track
rust colored acrylic paint for painting rail heads
black marker for drawing the center oil stain.
soldering iron and solder.
22 agw uninsulated solid bus wire.
20 agw red, black, yellow, brown, and white wire
20 agw green and blue wire
7 tortoise switch machines
14 dwarf signals
Mounting screws for tortoises
wire ties
7 tortoise interfaces
printed labels
3/8 drill and bits for wires, tortoise rod and signal masks

ps Ready to lay track, wire, install tortoises on perfab section

The plan is to weather and attach droppers to the turnouts, draw the center lines, fashion and mark the cork using the turnout templates,
drill holes for wires, tortoise rods, dwarf signals and regular signal masts, drill holes for tortoise throw rods, glue down cork, lay track, install Tortoise machines, install Tortoise interfaces, wire tortoise interfaces, droppers, and feeders, label Tortoises, Tortoise interface, and feeder bundles, hook up and wiring and perform checks. I'll just tack the track in place until I get this section in place attached on bench and married to the in-place track. Then I'll tweak the alignment and glue it down using Elmers white glue just as I did the cork roadbed.  (I use white glue here because it is less permanent than other glues - should I need to pull up a section of track or roadbed for any reason, I can just wet it down a little and it will come right up nice and clean.  In this regard, when I glue down track, I generally do not put muh glue on the turnouts - just a small amount at each end- so if I need to pull a turnout up to make and adjustment, it will come up easily.)

Weathering turnouts


I have already weathered a number of sections of flex track (see details of my track weathering technique in the post of 1-19-2016), so I need only weather the 7 turnouts needed for the first Fitzhugh prefab section. Again I use canned spray paint and a clean rag to thoroughly wipe the rail-tops clean after each coat. First, several passes using a flat dark brown sprayed on at angles to insure the rail sides are coated, then a dusting of medium gray and a dusting of light tan all sprayed straight down on the turnout. Then I use a small bush to paint the nail heads rust color.

tw Here is a weathered turnout in the foreground beside an unwaethered turnout.

Wiring Turnouts and Track


As discussed in the post of 1-18-2016, I plan to use only three droppers for each Atlas Code 55 turnout: one for the frog, and one each for the left and right stock rails. I will trust the turnout's internal connections to the closure rails, the pivot rails, and the exit rails so no additional bonds will be added. I first cut away a small section of the plastic tie material  from the underside of the turnout to expose the stock rails.
Then, taking care not to touch the plastic with the hot iron and not to solder the stock rail to the guard rails, I put a very small dab of solder on each exposed stock rail and a tiny drop on the copper frog connection ring. I then solder a short piece of 22 agw pre-tinned bus wire to each of these three places. I then test the strength of each solder joint by pulling rather firmly on the bus wires. They should bond to the rails well enough to withstand a light but firm tug, and you should be able to lightly bend the bus wire back and forth lightly without weakening the solder bond. Finally, I trim the bus wire droppers to about 3/4 of an inch and solder the three colored stranded 20 agw dropper extension wires in place - yellow for the frog, and red and black for the stock rails. For the red and black wires, which side is red and which black depends on the orientation of the switch. My general rule is the the red wire is attached to the rail adjacent to he outside edges of the bench, and the black wire to the rail adjacent to the center of the bench. These colored dropper extensions should be long enough to reach from the turnout to the tortoise interface.

tow Below is a turnout with the two small sections of plastic sleeper material cut away beneath the stock rails adjacent to the frog. (you can do this anywhere you like on underside of the stock rails.

Above is the completed wired turnout. The idea here is to use the small 22 agw solid copper bus wire for a neat connection to the underside of the rail and then hide (insulate) the solder connection between that dropper and the larger insulated 10 agw dropper wire extensions inside the pass-through hole in the bench top. Some might argue that hiding this connection could create a maintenance problem in the future. However, should the connection to the rail or the joint between the wires fail, I am careful to put a short servie loop below each pass-through to facilitate repairs. Alternatively it is a simple matter to drill a new hole beside the rail, install a new, slightly longer  22 agw bypassing-dropper to the side of the rail and then attach that to the dropper extension underneath the bench without having to take up the turnout. Side-of-the-rail connections are the norm on many layouts, and hiding such connections in the ballast is very easy to do. 

Droppers on straight flex track sections can be installed in the same way by cutting away two plastic tie-ends to expost the rails. In the case of the Fitzhugh section in this example, all but one of the internal flex track sections between turnouts are so short, that I plan to simply make them a part of the frog rails of the ajoining turnout by using a metal rail joiner and soldering the rail at the joint with the frog rails. Generally, I use a dropper for each section of flex, but here where the sections are so very short, soldering to adjacent rails is acceptable as long as the polarity continuity remains intact.

Using the XCAD rendering allows me to look at the 24 inch on center cross members under the bench to determine if any will interfer with droppers or tortoise machines. Should I encounter any this type of interference, I make a note to adjust the position of the offending cross member and make the nrw postion on the prefab section.

Having to Rethink DS64 Trunout Addresses

In the Digitrax DS64 Stationary Decoder Manual it says that the address numbers used to operate turnouts, crossing gates, and other devices can be any number from 1 - 2048. What it does not say (I just found out from the SE8c Manual) is that that the numbers 1000 to 2048 can only be programmed and accessed using some kind of computer control system, like TrainCntroller. This means that the 4 digit turnout addressing scheme I have planned will not be executable using the DT 400 Digitrax Throttle. This is not a giant problem, but if I do not change these address to 3 digit numbers, it will certainly represent a flaw.  Besides, it is not going to take that much work to fix it at this early stage: simply re-number the turnout list, and the DS64 labels, and the TrainController Switchboard labels. I have yet to enter the addresses themselves into TrainController, so harm there.

I had originally planned to use 1100 thru 1112 for Fitzhugh, 1200 thru 1206 for Westridge and so. Now it will have to be 100 thru 112 for Fitzhugh, 200 thru 206 for Westridge and so on thru 1620, the last turnout at Altamont yard. I'll just subtract 1000 from each DS64 address. Also, when I get around to programming signals using the SE8c, I have to select a board address that give me the singal addresses above 630 to avoid any possible duplication on addresses.

Assigning Board Address Numbers  

I have used the same convention for assigning Board Address on the A&BR2 that I used on the A&BR1: DS64 Stationary Decoders will be in the Range of 50-89; PM42s,90-94; BDL168s, 95 -99, and SE8cs, 6-12.

Laying Track
6- 19-2016
I'll begin with a prefab section pictured below The black box defines a 8' x 48' section of track at Fithugh containing mainline two crossovers and three siding turnouts. I begin by cutting 8 x 48 piece of pluwood and a matching pice of homosote. Then using 1 1/8" sheet rock screws, I screw the homosote to the plywood -  two screws about 6 inches apart every foot - that is 10 screws for the 4 foot module.

In this case the mainline tracks spaced 2 inchs apart. My normal center to center track spacing for mianline track is 1 3/8 inches on the straigntaways and 1 1/2 inches on curves. Here I have used a lightly wider spacing because I contemplate a small yard building betweeen the tracks. I have elected to use a single piece of 1/8 inch cork instead of a system of cork switchpads connected by cork roadbed strips. This makes for a much cleaner installation and saves a lot of cork cutting etc. I begin by carefully locating the centerlines and penciling them in on the homosote using the XCAD 1 inch grid as a guide. I then pencil in the outline on the single, large cork pad (just wide enough for the double track but not so wide as to cover the location of the 1/2 inch holes I will drill for the turnout throw rods.) I then cut out the cork pad, and glue it down using white glue.  Next, I draw on the cork  - solid mainline track centerlines and doted lines
to indicate the location of any supporting cross members in the bench work. Should a cross member interfer with a switch machine or a dropper, I can mark a new location on the cork and move or notch it when I install the section. Then I again use the Xtrack CAD rendering with the 1 inch grid and my trunout temples to locate all seven turnouts, lining up the track centers on the centerline, marking all railends and all holes to be drilled. Finally, I drill the holes for the turnout wires and the large 1/2 inch holes in the bench adjacent to the edge of the cork pad for the turnout throw rods, then thread the wires  through the small holes in the pad, and position all the turnouts on the center lines using a few track nails to hold them loosly in place.

lt1 Cork glued in place, center lines and cross memeber lines drawn, holes marked and drilled, and turnouts installed and tacked down. The turnout in the right foregroud is not yet lined up with the center mark, but when it is alligned, the throw bar will be positioned directly above the large 1/2 inch hole in the bench most of which which you can see in the shadow just to the right of the cork pad in th elower left of the photo. 

Next, I install the connecting sections of straight track, taking care to follow the plan with regard to the use of insulated joiners. I like to use a Dremel Tool to cut track. I mark it just wee bit long, cut the section, and then use the cutting disk to trim the ends until the fit is perfect - snug for short sections, and a .020 inch gap at the metal jointer for longer sections. I use a piece of .020 sheet styrene to set the gap.
When track laying is complete and everything is lined up very straight and perfect, and all track and turnouts lightly tacked in place temporarily using track nails, I tack on some 2x4 wooden guards. These will rest on the bench and protect the track work when I turn this section over to in install switch machines and wire  everything up.

lt2 Completed track work. Note the 2x4 guards tacked in place to protect the track work. The entire section will rest on these guards when I turn this over to work on the back. 

Installing Tortoise Switch Machines and Tortoise Interfaces on the Prefab Section

Now that the cork is glued down and the track is all installed, connected, aligned, and tacked down, I can flip the entire section over and install switch machines and wire everything up. This is where working on a prefab section really makes things much easier. Installing Tortoise machines under the bench is awkward and difficult to say the least, whereas installing these units on a prefab section, where you can flip it over on the workbench, is much easier. I just put the Tortoise in the center of its throw, thread the throw rod into the pre-drilled 1/2 inch hole and through the tiny hole in turnout's throw bar, and, manually moving the tortoise throw from one side to the other, adjust the unit's position until you get good snug, point contact on both sides. A shaving mirror and a work light stragically positioned below the bench help me to observe the points of the turnout to ensure perfect action as I manually move the Tortoise from side to side. When the Tortoise is perfectly positioned, I  mark the four holddown screw positions, drill the four holes and mount the Tortoise. Since I am using 3/4 inch plywood and 1/2 inch homosote, I use a length of .032 inch music wire (spring steel rod) instead of the .024 wire that comes with the Tortoise, which is not long enough. This gives me the added length I need plus a little more tension on the turnout throw bar. Once the Tortoise machines are in place, I add strips of cork roadbed along the sides on the installation with narrow slots under the turnout throw bars. I then drill the holes for the dwarf signals, solder the green and blue droppers to the legs, and install the signals feeding the three wires through the holes. Next, I follow the steps outlined in my post of  3-4-2016 to wire everything up. At this point I attach labels to three sides of each Tortoise machine, and to each tortoise interface barrier block cover, and to the points where each set of track wires penetrate the bench. I then use my multimeter to check the electrical continuity and correctness for all rails and frogs and dwarf signal outputs in both the thrown and closed position for all turnouts. While I am doing this, I do a little extra labeling to help make serving each interface a little easier should the need arise. This involves marking the mainline (closed) position of each Tortoise throw bar both on the Tortoise machine and on the bench next to the machine, and marking on the bench at the penetration all of the blue (+) LED wires ti indicate which supplys the red and which supplys the green leg of the LED. I also mark on the tortoise and on the bench next to the Tortoise the side of the Tortoise edge connector whose outermost contact is the positive power to the Tortoise Machine (white wire in my color scheme).

Note: I have experienced some intermittentcies in the connection and in the solder joints to eight position  edge connectors, so I check the frog continuity and the dwarf lights while manually wiggling the connector around by the wiring harness to be sure all connections are solid. The photo below shows the underside of the prefab section with the 7 turnouts and 7 Tortoises and 14 dwarf signals wired to the interfaces and labeled.

tl4 Seven Tortoise Machines with labels and track wired up via Tortoise interface barrier blocks..If you look very closely at the thrid and forth barrier block in this picture, you can see some of the 30 agw wiriung for the layout lights, which I later ditched in favor or more robust 20 agw wiring. 

After working with the 30 agw wrapping wire, I quickly realized that this was a bad plan - the stuff is just too delicate and too tedious  to work with. So I ditched the 30 agw in favor of a more conventional 22 agw blue and green wire hook up that employed a single sub-bus feeder pair off the main 5 volt supply bus and a daisy chained blue/green parallel circuit to the lighting inputs on each of the seven Tortoise interface barrier blocks. This involved re-wiring all the dwarf signal LEDs, drilling new holes large enough to accommodate the 22 agw wire the same way I handled track droppers. This required the holes with the outer two slightly angled  and then bending the LED legs to conform to the wider three hole array.

Here is the nearly finished section with the dwarf signals illuminated. I have yet to install sleepers to replace those removed for rail joiners. Also, notice that I have added a line of cork roadbed with narrow cutouts under the turnout throw bars and dwarf signals along the sides of the intalllation. In additon I have drilled 1/2  inch holes to accomodate future signal masts. (See the my entries regarding signal placement 2-24-16 and 7-4-16.)

I am now ready to install the feeders to the input side of the Tortoise Interface Barrier Blocks, which I can do if I mark and measure the exact locations of the DS64(s). I mark the position(s) on the underside of the prefab section so I can line up the DS64  later, and then proceed to measure and cut feeders - for track power - the distance from the tortoise interface barrier block to the edge of the section + the distance from the edge of the section to the edge of the bench once the section is installed (less 4 inch for the mounting board/bench edge set back) + 8 inchs for a service loop behind the bus wire mounting board + the distance down the bus wire mounting board to the appropriate red or black track power bus + 10% for trim. For lighting sub-bus feeders, the same. For DS64 outputs the same + the distacnce along the wire mounting baord to the DS64 mouting board. 

All that is needed now is to install thin sleepers at the points where they have been removed to make room for track droppers and/or rail joiners, and to solder all the conductive metal rail joiner connections and paint those joints and the plastic joiners as well. I use a dremmel cutting disk to flatten the sleepers so they slip easily under the rail joiners without lifting the track, and a dab of white glue to hold them in place and a dab of gray paint to match the weathering if needed.  (It is best to use sleepers from a weathered piece of track.) I will not solder the metal joiner joints or paint any rail joints or joiners unitl after the track is glued down on the layout just in case I want make any adjutments or alterations. At first I thought I would solder the metal conductive rail joints on the back side of the rails - the side away from the viewer, but upon reflection, this would make it difficult to un-solder them should I need to take something up or make a change once the prefab sections is in place on the layout. So I'll make my soldering neat and keep it on the accessable side of the rail. With the little paint, it will be invisible.

Testing Installed Atlas Code 55 Turnouts and the Prefab Section Wiring.

To test this prefab section,  I first hook up DS64s and program them to the turnout addresses powering them for the 12 volt supply on on my special work bench. The two crossovers will, of course, work in tandem, so for the seven turnouts, I will only need five DS64 addresses and feeders.  I then connect all the red A track feeders and and all the black B track feeders to their respective track power bus on my track laying bench.  Since there are no detection requirements for this route between detected blocks, the A and B track feeders can be daisy-chained and connected to their respective power buses at a single point.

Now, I just wire up loconet to the DS64s and I am ready to run a few test-locos across this section using my Digitrax throttle.

test Testing prefab section track work and wiring on the the prefab bench using my spare  DCS 100, 2 DS64s and my Digitrax throttle.

The running of the locos or even short trains on prefab sections not only checks for electrical wiring continuity, but most importantly, it checks the precision of the track laying. So I like to run one loco of each type in my roster,  just to make sure they all run smoothly without derailments or hesitations through each turnout in the section in both the closed and thrown positions. Backing a short rain through each turnout is generally a good idea as well. I have found cases where all but one of my locos ran well in a certain turnout, so it is best to be thorough in your testing. 

One note about the Atlas Code 55 turnouts. Remember, I am connecting the Atlas turnouts to track power only at the stock rails and the frog, and relying on the turnout's internal connections to the pivot rails, the closure rails and the frog exit rails without the additional hard-wire bonding often recommended by so many experts. So I have to be careful to fully check the electrical connections to each rail in each turnout. I found that, after installation,
one turnout had no connection to one of the point rails, so I fashioned a bond using 30 agw wire. After a little experimenting I found the failure was intermittent, owing to stress on the turnout. Each installed turnout, when tacked down to the bench, can be placed under a small amount of twisting stress by the tension rod in the Tortoise Switch machine and sometimes by a slightly misaligned installation. This tension can vary depending of how the Tortoise is installed, so it is best to keep this in mind, and as you install your turnouts and position your Tortoises. Try to place them in a way that does not put undue twisting stress on the turnout because this twisting might cause its internal electrical connections to fail or to become intermittent.

As I noted before, it is best to test a lot of your rolling stock on a new installation. I had one case where all my Altas locos ran one turnout fine except one of my big Kato E8s that derailed at the same spot every time. A very small amount of filing on the corner of a closure rail - right at the junction with the end of the point rail - proved to be a simple fix for the offending loco. This is very delicate business in N scale. Take your time. MAKE IT ALL PERFECT, AND YOU WILL SAVE YOURSELF A LOT OF HASSLE DOWN THE ROAD. 

Rethinking the Layout Lighting Wiring Scheme

The 30 agw wrapping wire had trouble written all over it. Bad plan! Replacing it with 22 agw made for a much more robust, much more easily serviceable installation, and using the daisy chained sub-bus feeder, did away with the need for the layout lighting distribution blocks. A single feeder pair for each group of LEDs will be attached directly to the 5 volt power bus on the bus mounting board and a label at that connection will list the signals or LED lights in the group. I was trying to avoid daisy chained connections under the bench in favor of a system where all connections to the bus were on the mounting board, but the daisy chain seems makes more sense and seems unavoidable in the case of the layout LEDs.

Signal Protocol, Programming  and Placement  
Now that I have begun to lay track on my prefab modules, I am drilling holes for mainline signals as I go, so before I drill I carefully recheck the placement and propsoed function of all mainline signals. I then check to ensure that they are correctly placed on the TrainController switchboard and in all my XCAD files, especially on the detailed sectional renderings that I am using as guide to lay out track on the prefab modules.

Signal placement, protocol, and programming is tricky business, so it is a good idea to work it out, and then after some time has passed, revisit it for a final check.  Proper and consistant placement requires, an ironclad signal protocol. Which side of the track is a signal to be placed on? How many heads, aspects, etc? What do the colors actually mean? On real railroads, these protocols vary from line to line, and from era to era. It is good to do a little reading, but in the end you have to work out a protocol that works for you. There are really not set rules.

I use the Digitrax double headed masts with three aspect heads (red, yellow green) to the right of a mainline track and offset by about 1 inch. A 1/2 inch hole in the benchtop for each mast is perfect, allowing for the mast to project above and the tiny circuit board to snuggly fit in the hole below. I use modeling clay to fill the remaining gaps making any future removal or serviing of the masts easy. Since my benchtop is 1 1/4 inches thick I use a 9/16 drill to auger out about 1/4 inch deep on the bottom side to allow the protruding connector to fit in the reccess, and thus allow the full height of the mast to be exposed above the benchtop. 

Electrically these masts can be used in a four aspect mode with a blinking yellow as the fourth aspect. I prefer to use them in the three aspect (red, yellow, green) mode, with the top head indicating the status of the mainline (closed) route through an upcoming turnout and the bottom head indicating the status of the diverging (thrown) route. Green indicates a clear path, unobstructed by any upcoming turnout positions or by loco occupancy or reservation through the next two blocks, or to the next signal. Yellow indicates that the next block is clear of obstruction, but that the second block ahead is either occupied or reserved or that a turnout after the first block but before or within the second is in a blocking position. Red indicates that the next block is occupied or reserved or that a turnout before the next block is in a blocking position.

Programming signals in TrainController is as straightforward and at the same time as flexible as this kind of thing can be, but it is often, by its nature, pretty confounding, so I take my time and try to conider all the possibilities. I am currently visiting family and away from my shop, so with a lot of downtime over the next few days, I am programming signal triggers for theA&BR2 in TrainController off line on my laptop. 

Breaking Ground

At last! We broke gound last Thursday. Here is the first photo of the new train room.

First photo of the new train room. 7-5-2016

Tortoise Interfaces on a Prefab Module

I am continuing to construct prefabricated modules, laying track and wiring turnouts. Here is a photo of the underside of a simple module (a mainline crossover and one siding switch - three turnouts with three Tortoise machines, and five dwarf signals.) I have yet to install feeders on this module. Notice the service loops on all droppers. I construct droppers by soldering 20 agw insulated wire to 3/4 inch lengths of uninsulated 22 agw bus wire that is soldered to the bottom of the rails of a section of track or the rails of a turnout, or to the turnout frog ring connection on the Atlas Code 55 turnouts; I solder my 20 agw droppers directy to the legs of LEDs in the case of layout lights or dwarf signals. Since my bench top and roadbed is 1 3/8 inches thick (3/4" plywood, 1/2" homosote, 1/8" cork)  these solder joints are hidden and insulated in the bench-top pass-through-holes,) I always create a small service loop to facilitate future accessing these solder connections and to make it easy to free up a section of track or a turnout should the need arise.

pmti Prefab Module on the work bench with Tortoise Machines and Tortoise interface and  track (red, black), frog (yellow) and LED dwarf signal light (blue+, green-) droppers attached. Note the small service loops. 

Footings Poured Today
July 13, 2016


Block Walls - Train Room Takes Shape


Foundation Walls Complete

The foundation walls are complete, reinforced with rebar and filled with concrete. Waterproofing is under way, and they are preparing to pour the slab. Notice the 3" conduit under the floor. This will connect all the "along-the-wall" bench work to the center section bench work and thus allow me to locate the booster in the center of the room and minimize many power bus wire runs.

Track Laying Using Xtrack CAD

Xtrack CAD is a pretty remarkable program, and its free. It takes a little time the get fully facile with it, but once you get the hang of it, it can be great help in creating perfect track work when using flex track. Among other things, it will monitor all radii and grades to insure you have not gone above or below your set maxs and minimums, create perfect track alignment and spacing, and automatically calculate and insert easements and curves. If you use the automatic track joining feature, the program will not let you  join two sections until the positioning of the sections to be joined is correct, and then it will insert all the segments needed to join your sections, including any easements, while allowing you to control the size and sweep of any radius to be inserted. 

As I mentioned in an earlier post, I use detailed Xtrack CAD renderings to layout the prefab sections I am now building. Some modelers printout 1 to 1 full scale pages and paste them to the homasote and lay track right over the top them. For me, a large scale print out (say 6 to 1 or so)  with the grid set to 1 inch allows me to locate track on the homasote surface with very precise accuracy. The print layout routine in Xtrack CAD is a bit difficult to master, but again once you get the hang of it, it is great.  

xc Here is an Xtrack CAD printout of a 48x 8 inch prefab section with my notes for locating track, turnouts, and cross members of the bench. This gives me accuarcy down to about 1/8 of an inch, and sould any finer adjust be needed, I can take care of it  when I position the prefab section on the bench work.  

September 2016

Slab Poured 

sp Concrete finishers at work in the new train room. We should begin rough carpentry soon.

Painting Figures

I have completed all of the prefabricated track sections except one, the long section that will contain the two mainline tracks and a parallel section of the yard lead and ladder at Altamont Yard. I am still giving thought as to exactly how I want to built this last section.

In the the meantime, since I like to place all of my bridges, bridge abutments, and station platforms as I lay track, I have bridge kits and station platform kits to build. I think I will start with the platforms, and I'll start that process by painting some figures to go on the platforms. I generally buy unpainted figures because I enjoy painting them and because the painted figures are horribly expensive. I like the Preiser unpainted sets best (~$23 for a set of 120), but they have been hard to find lately so I am today using Model Power sets of 72 unpainted N scale figures (about $8.00 per set), which are fine. Both the Preiser and the Model Power sets are 160:1. Some other brands are 150:1. You can use both if you like. The difference is not enough to notice if you don't place the 150:1 figure right next to the 160:1 on the layout. On the A&BR1, I have one town that is 160 and one town that is 150. I do not think anyone has ever noticed the difference.

pf Here are some Model Power figures being painted. I hold them by hand and use a piece of  duct tape, sticky side out aroud a small piece of .040 styrene to hold them upright while drying. Not pictured is my large fluorescent magnifying lamp.

Since Testors discontinued the manufacture of their Railroad Colors line. I have been in the process of changing out all my paints to the Vallejo brand of water soluble flat acrylics. I particularly like the plastic squeeze bottles, which make it easy. Simply shake it up, and dispense just a tiny drop of well-mixed paint, without having the clean the lid and worry too much about the proper resealing of the bottle. I think that these paints are going to last much longer than those that come in little glass bottles, like Model Master. In these plastic bottles, they seal more fully and easily and appear to be much less likely to dry up.

Elevation Calculator

One on the nifty things about Xtrack CAD is that, if you enter the elevations at the beginning and at the end of any run of track, it will allow you to read the grade % (out to 3 decimal places) as part of the properties of each of the sections of track in the run. This greatly facilitates the accurate building of riser supports for roadbed along any graded run. You just have to do the math. For example, if I have a 12 foot run of track on and even grade with a beginning elevation 2 inches above the homasote bench top and an ending elevation 4.5 inches above the homasote bench top, Xtrack CAD will tell me the grade is 1.736%. So, if the beginning  elevation is 2" above the homasote bench top, the support below the beginning of the run is 3/4 inches above the bench cross-member (2 inches minus 1.25 inches [3/4" plywood + 1/2" homasote]), and the next support is to be at the next cross-member 24 inches away, how high above that cross-member should the next support be to to give me a 1.736% grade? Well, its is 24 inches long times the 1.736 grade% plus the beginning elevation of 3/4" or 1.16664 inches, roughly 1 5/32  inches above the cross-member and 2 13/32 above the homoasote bech top.

I've written a little program in C# using a form I created in Microsoft Visual Studio. This allows me to enter the beginning elevation (
referencing either the homasote surface or the top of the cross-member), the grade % , and the length (distance to the next support). I then press the "calculate button" and I have the height of the next support, to the nearest 32nd of an inch, measured either from the homasote top or from the top of the cross-member.

Cool, huh?

Lighting a Covered Station Platform

The photo below, taken on my A&BR1, illustrates how dramatic the addition of directional LED lighting can be along an open covered station platform.
pl A night view of the platform at Westridge on the A&BR1. Notice how the directionality of the warm white 3mm LEDs I have selected, creates distinct shadows and realistic, soft pools of light on the platform.
I have finished painting 144 N Scale figures, and other platform details like freight boxes, misc. baggage, schedule signboards, and clocks, and I have painted the parts for eight 6 inch sections of butterfly style platform (Walters #3258). Each section will get four 3 mm LED bulbs wired in series attached to the underside of the roof so that, along the entire 24-inch-long platform, there will be soft pools of light halfway between all the support posts except outside of the two end posts.This is accomplished by measuring and marking the location of the 4 LEDs on the underside of the rooftop, and then gluing the LEDS to underside of the roof  before you glue it on. Use a good sized drop of CA cement to glue the 3mm bulbs in place (If you are using a covered platform that does not have such a space, you can notch the supports  (just large enough for LED leg wires) in the center of the roof-support trusses just below the junction with the roof. Be sure that you orient all bulbs so that they are daisy chained + end to - end. When the LEDs are securely in place, drill two more small holes in the platform floor just to the outside of the two end posts for the vertical power feeds. Then carefully solder the LED legs together where they overlap taking care not melt the adjacent plastic, glue on the roof, and attach two vertical 22 agw bare bus wire vertical power feeds up through the holes in the platform. The two vertical power feeds are left visible next to the support, like the kind straight electrical conduits one would often see on old-style real platforms.  You can paint them if you like. The kind of soldering needed to connect all of this must be very quick to avoid melting any plastic: quickly put a little solder on each LED leg to add a little extra tinning at the joint, and then get a little solder on the tip of the iron and quickly touch it to the LED legs where you have tinned them. With a little practice you can get a very quick bond without ever heating the wires so much that they melt the plastic. 

Here is the completed underside  light assembly. Not pretty, but it doesn't need to be, becaue except for the two vertical "conduits." the wirung and the top part of the bulbs are hidden from view.

Note: The choice of LEDs and the single accompanying resistor is critical. F
or this project, you will need tiny 3mm LEDs That emit a soft yellow/white light. You can calculate the resistance, if you like, but it is best to build a little 4 bulb-one resistor circuit on a bread board and experiment with differnet bulbs and resistor values unitl you get the exact color, directionarity, and brightness you like.  The LEDs I have selected are from Newark (MCL 034SWC: LED, T-1 (3mm), Warm White, 3.7 cd, 36 degree, 20ma, 3.4v - $.39 each). They are a nice soft warm color, and very directional so I get my nice pools of light. Each bulb is rated at 3.4 volts forward, but they seem to work fine in parallel groups of four with a single 1k ohm to 10k reisitor (depending on how bright you want them) on the auxillary 12 volt circuit I use to power Tortoise machines. Unlike the 5 volt layout lights, these bulbs will not be switchable and they will remain on all the time. I do not see this as a problem because they are dimmed down so far to get the correct effect, that in full daylight conditions, noone will notice that they are on.

Testing four 3mm LEDs on a breadboard using a 10K resistor hooked to 12 volt supply. 

After it is all put together, you may need to adjust each LED so it's pool of light falls in the exact center of the platform. Since you used a lot of CA glue, you can gently bend each bulb in place to get the exact right angle. 

fp A finished 4 bulb platform section. The concrete floor and the black metal roof both need a little weatherng ( a light black wash for the floor will bring out the pattern on the slab sections, and just a littel RustAll on the roof will give some highights and articulate the crimping joints. I think the LEDs are still too bright here. A larger resistor will dim them a bit and give me softer, more diffuse pools of light. I am using a 1K ohm resistor here,. After some experimenting, I settled on a 8.6k resistor. You have alot of latitude here.

Floor Joists Installed


fji New train room for the A&BR2 - 38' (widest) x 22' with a 9' ceiling. For N Scale, this is enormous. I calculate that, if a train were to run a complete loop starting by the entrance and continuing all the way around the room, back to its point of beginning, including a loop along the long center section of bench work, it would travel almost 9 scale miles. 

Contemplating the Completion of the New Building.

di Roofers and Plumbers next week. Finish date Jan 1 ?????

The new building that will house the new A&BR2 train room in the basement is progressing nicely. We are fully "dried in," the framing and sheathing complete, windows are in, the shingle siding is installed, and all of the exterior trim is compete. Next week we get the slate roof and some exterior painting, followed by the hanging of the exterior doors, plumbers, electricians. septic tank installation, etc. - then sheet rock and interior finishes. All the while, the masons will be working on the stone exterior, which will probably take another 6 weeks or so. I suspect that the building will be largely finished by the first of the year.

The train room has the 2x4 firing installed and they are ready for sheetrock on the outer walls.  I can now begin to construct the train room backdrop walls and the bench work and install train room AC wiring and lights. I had always planned to hire professional carpenters to frame the backdrop walls and the lighting sofit, but having watched the pros work for the past months or so, I now plan to get professionals to build the bench work framework and my control panel as well. They can make it much better, and can finish in a day what it would take me weeks to build. So, I'll have them frame the sofits, the backdrop walls including the tricky radial corners, and then get the sheet rockers in.
After that will come the electricians to install the lighting and all of the electrical outlets around the walls. Finally, I'll paint the backdrop wall sky-and-cloud murals and the sofits and ceiling, and then get the carpenters back in to build the basic bench-top framework and legs. After that will come the electricians to install the lighting and all of the electrical outlets  in the bench work, as wells as the main electrical and lighting switch panel.

Train Room Lighting and Electrical Control Panel

All the electrical in the new train room will be switched at a control panel by the computer control panel excedpt of the power to the HUE lights. There will be one light switch by the door to turn on the HUE lights, which will always come up full daylight white after being turned off at the main switch. The main switch panel will be mounted vertically on the bench work next to the computer control desk. From this panel I can  power up AC to everything, computers, DDC boosters and peripherals, including room lights, sound system, servos and servo controllers, and all of my power supplies for layout lights, dwarf signal lights, and stationary decoders. 

Train Room Lighting and AC Electrical Revisited

It won't be long before the electrician begins work, so I spent the morning giving a final once-over to the lighting and AC wiring plan. The original plan was for 18 T8 dimmable single bulb florescent fixtures switched on and off at the control panel. The dimmers will be controlled mechanacally by a servo motor which will in turn be controled by custom software that I will write to communicate with a servo motor control card via a USB connection. The 18 sockets for Phillips HUE A19 colored wifi-controlled LEDs will be switched on and off at the entrance door and the HUe light's color and intenisty will be controlled by custom software that I will write. In addition, a circuit of 110 volt AC outlets will be installed around the exterior walls for work lights, tools etc, as well as a switchable circuit of 110 volt outlets switched on and off at the control panel for all DCC components and all layout power buses.  This is the kind of set up I have on the A&BR1; but at the time I built that layout LED lighting technology was in its infancy, so I spent some time this morning thinking about replacing the florescents with the new daylight LEDs.

Using 10 watt, 800 lumin LED floods, would be perhaps just a little cheaper and much simpler to install than the florescents, and they would run on a smaller wattage dimmer. Still, I would need  60 of them to get the same light output I will get with 18 fluorescents T8s (18 T8 at 2750 lumins each = 49000 lumins ~=  60 10 watt LED floods at 800 lumins each  = 48,000 lumins.) LEDS work better with a dimmer than fluorescents do. Fluorescents only dim down to 1% and require occasional adjustments, while LEDs dim virtually fully and are quite stable. The 6500 Kelvin fluorescents I use give a lovely daylight, but I can get LEDs that put out 6500 Kelvin as well. Anyway, after a little experimentation, I found that the only real problem with the LEDs is shadows. Even if I have an LED light every foot and a half or so, they still cast multiple light shadows on the bench top, whereas the fluorescents give a nice, realistic, diffuse, noontime light with virtually no shadows at all.

I plan to discuss all of this with a lighting consultant at the lighting supply company I use, but all things being equal, I think I will stay with my original plan and use dimmable fluorescents with HUE light accents and effects. Given my experience with this setup on the A&BR1, I am of the mind that this is the optimal model railroad lighting method: the most light with the most realism, the most complete effects capibilities, and the best and most complete automated control possibilities (that is, if you can write your own programs).

Laying Out the Backdrop Walls and the Overhead Lighting Soffit

Now that the train room is ready with all the 2 x 4 firing attached to the block walls all the way around the room, we are ready to frame the backdrop walls that will support one side of the bench and to frame the overhead fir-down soffit partitions that will hide the overhead lighting. All of this framing will be 2 x 4 stud walls. The backdrop wall studs will be 32 inches on center below the bench level to allow easy under-the-bench access and will conform to one of several different plans above bench height as detailed in the drawings below. These walls will include the 4 foot radius sheet rock corners. The soffit fir-down partition wall studs will be 16 inches on center, attach to the ceiling and will dropdown to 6 feet 6 inches above the slab floor. These dropdown partitions will effectively hide all lights mouted on the ceiling from viewers as long as they are very very  short and/or very very close to the bench. 

ws Backdrop Wall and Lighting Fir-Down  Wall Sections

XCAD rendering with new dimensions and details of wall section placement.

All of my planning drawings so far have assumed the new train room to be 38 feet by 22 feet, which is the designed distance from block wall to block wall. With the fired in perimeter walls with their 1/2 inch sheet rock to come, the as-built inside dimensions sheet rock to sheet rock is  37' 8 3/8" x 21' 8 1/4". As you can see in the drawing above, I have adjusted the master XCAD rendering to reflect these tweaks by stealing a few inches from the access aisles behind the back drop walls. I then created the simplified drawing below for the carpenters, detailing the exact location to the framed backdrop walls and the overhead fir-down partitions. Using these drawings, they can snap lines on the floor and cieling for the base plates of all of this framing. We will then double check everything and then proceed to build the walls.  (I used PowerPoint to create this and some other drawings in this blog. For creating and labeling simple shapes PowerPoint is very easy to use and it will export almost any kind of file,  jpeg, pdf etc. whatever.)

Backdrop Wall Initial Framing Completed

The initial framing of the backdrop wall (shown in brown in the drawing above and pictured below) is complete. This initial stage of framing is a simple 2x4 stud wall with studs 32 inches on center including four 4 foot radius rounded corners above the bench. The wide spacing of the studs in the initial framing will allow easy access to the area underneath the bench. The horizontal 2 x 4 support rail 39 3/4" above the floor will support the 1x4  bench framework grid. The bench grid will have a finished height of 44 1/2" (39 3/4" + 3 1/2" for the 1x4 grid members + 3/4' for plywood top or roadbed + 1/2" homosote top or roadbed.)  The upper section of the wall above the bench will have studs 16" on center and will be sheetrocked over to form the backdrop sky-cloud-mountain mural. The above-the-bench sections of the wall will each conform to one of several different wall-type designs to allow access to tunnels from behind the backdrop wall (see the wall sections drawing and the wall type floor plan above.)

bf A portion of the framing of the initial stage of the backdrop wall showing a 4 foot radius corner, the basic 32"-on-center framing.Notice the access aisle behind the backdrop wall. The drop-down-from-the-ceiling lighting soffit walls are not yet constructed.  

Final Train Room Lighting and Electrical Plan

Now that the backdrop walls are framed, the electrician can begin work in the train room to install lights and outlets. In anticipation of this, I made a final review of the train room lighting and electrical plan,which I have beenworking on,
on and off, for the past year or so and keeping up to date in XTrackCAD. The drawing capabilities of XtrackCAD are limited, and after review, I decided to use a different application to create the final drawing for the electricians, so I exported a to-scale JPG rendering of the train room and the bench work from XtrackCAD, and I then inserted this floor plan into Power Point, which I often use for simple drawing and labeling tasks. Over the top of this floor plan, I created a new scale drawing of the lighting and electrical in Power Point, showing the exact location of all fluorescent and HUE lighting fixtures and all electrical outlets as well as the location of the switchs for each. I further delineated Phase1 (wiring in the walls and ceiling to be completed now) and Phase2 (wiring in the bench work to be completed after the bench work is constructed.)

lep Final Lighting and Electrical. The long blue rectangles are the 4 foot long single bulb fluorescent fixtures and the light blue circles are conventional screw-in sockets for HUE lights.The squares are double duplex outlets. Switches are marked with an "S". Each square in the grid = 1 foot.

December 2016

Lighting Soffit Framing Complete
All of the framing in the train room is now complete, except for the lateral bench support rail, the HVAC crew is here today, and we are ready for the electricians and then the sheetrockers.
trf1 Standing in a viewing aisle and looking toward a corner of the train room. The drop-down framing attached to the ceiling forms deep soffits to hide the lighting fixtures and duct work. The bottom of the radial corner is about 14 inches above the future bench top to allow a generous reach-through portal behind the curved wall to access the tunnel that will go in this corner. I plan to sheet rock all the way down to the bench top, and then cut away the reach-through to match the contour of the mountain.
trf2 More detail of the drop-down soffit walls.  
trf3 A view down the bench, which will go just below the lowest cross member. Again we are looking toward a corner radius, and in this photo we can see the access aisle behind the backdrop walls. 

Final Tweaks to Block and Turnout Names and Addresses and to the Power District Layout

This morning I am going over the TrainController Switchboard to re-check all my block and turnout names and addresses against my block and turnout wiring labels and my Master Block and Turnout Lists to make sure they all agree. I have made a number of changes over the last few months, and I want to be sure all my documentation got updated. While doing this, I am also checking to ensure that the naming and the numbering of these blocks makes sense in the context of their physical locations on the layout, and that the blocks in each power district form a contiguous logical cluster that corresponds to a specific area on the layout. For example, I want to make sure that all 16 of the blocks in the Altamont Terminal Power District #90-2 [Brown] [BDL168 #95 with addresses 95-1 thru 95-16] are all in the Altamont Terminal area and that they are named and address-numbered in a logical sequence vis a vis their location on the layout. Most importantly, I want to ensure that I don't have any 90-2 blocks in far flung areas away from Altamont Terminal.
With a few tweaks, this has all worked out well, with only one block (a remote spur) that is not contiguous to the other blocks in its power district and corresponding BDL feeds.

Similarly, I am checking the turnout names and addresses to ensure they they represent logical spacial groupings and that the sequence of these numbers makes sense in relation to the turnout's location on the layout. I am also checking the DS64 stationary decoder labels to make sure they correctly correspond to my list and to the turnouts on the Switchboard and on the layout itself. Finally I checked the proposed locations of the DS64s that control the turnouts looking for both logical placement and the most efficient wiring runs. In short, I want the four turnouts controlled by each DS64 to be near to one another on the layout and I want the DS64 on the wire mounting board as close as possible to this group of turnouts.

Manipulating Color in HUE Light Programs

It has been almost a year since I wrote the first little room lighting program that emulates a sunrise on the A&BR1 using my Phillips HUE lights. (See my post of 4-14-20216.) I have since modified this original application and written applications for sunset and a thunder storm.  In these programs and throughout all of my initial work with the Q42.HueApi C# reference library, which is designed especially for HUE light communications, I have changed color by individually entering, line by line, the hexadecimal strings that represent the color coordinates of each color I want the lights to output. So a fade from blue to red/orange in 10 two-second steps looks like this:

            //morph color from blue-violet to red-orange




            etc. to



I noted at the time, that I knew there was a way to employ a loop to do this, but the hex numbers kind of put me off, and I just never got around to fooling this or with any of the other ways one can create colors in the Q42 library. Over the past few days I have been working on HUE color coding methods, and I have devised a way to use a loop to accomplish color morphing tasks using hex strings like the one above. The method is pretty obvious. Increment normal integers in the loop, and after each step convert each integer (r, g, and b) to a two digit hex string, and then paste the three strings together to get the 6 digit string that HUE uses to specify a specific color. The code below is an example, although the numbers here are arbitrary and I have no idea what colors this creates. Working with RGB in this way requires that you consult a RGB color chart the get the numbers of the red, green, and blue components for your beginning and ending colors and then work out the size of the steps needed, the number of steps needed, and the ratios of i for each step. Thus, it is still really clunky, and since it moves in a linear fashion for each component, the colors you get along the may not be exactly what you want or they may not even fall within the range of RGB combinations supported by HUE lights, which cannot reproduce all RGB colors. To be sure, the ratios of i you use could be arranged to change as the loop executes and thus scribe a non-linear path, but this starts to get kind of hairy; or you could use several loops to make things non-linear.

            //initialize new hue client and new lightcommand 
            ILocalHueClient client = new LocalHueClient(ip);

            var command = new LightCommand();
            //declare and define integer and string avrriables
            int r = 27;
            int g = 0;
            int b = 120;
            string rString;
            string gString;
            string bString;
            string color;
            // loop - incremental color calculation, tranformations to hex and commands with 2 sec delay
            for (int i = 0; i < 25; i = i + 5)

                r = r + i ;
                g = 0;
                b = b + (i * .5);
                rString = r.String("X2");
                gString = g.String("X2");
                bString = b.String("X2");
                color = rString + gString+ bString;

Fortunately there is a better way. There is an "xy" system based on the CIE color chart. On this chart all of the colors are expressed in terms of the x and y axis (numbers between 0 and 1). The Q42 web page also indicates area on this chart that are supported by HUE lights. So a nice deep blue might be (.2, .2) and a red/orange might be (.5, 3.5). So using this system, the loop to get from blue to red/orange in 30 two-second steps would  look like this:

            x = .2;
            y = .2;

            for (i  = 0; i  <  31; i++)

              x = x + (i * .01);
              y =  y + (i  * .005);

This is a much better approach in every regard. I created a little application with two trackBar sliders each with 100 increments (0 to .99) in order to preview all the colors on the chart using the lights themselves. This is a big help because the color the light produces is not alway exactly what one would expect when looking at the same color on a color chart. 
The use of the timing "sleep" pauses in these loops ( and indeed after all commands to HUE lights) is critical. The HUE lights are slow to react, and without adequate pauses. the system may miss steps and then catch up resulting in visually jerky color transitions. I recommend at least a 1000ms delay after any command.
Construction Blues

I always find the construction process maddening. It seems as though it always moves from one disaster to another in a series of unpredictable jumps and starts that scribe a two-step-forward-one-step-back arc. It is always over budget and always behind schedule. That is just the way it is. Progress has been slow, but the quality of the work is excellent, and yesterday we passed the rough-in inspection, which means the electrical, the plumping, and the HVAC are all roughed in, and we are ready to begin finishing the interior - insulation first and then sheet rock and finally the finished trim, floors and tile. In the meantime, we have had some problems with the subterranean waterproofing and all the backfill had to be removed and the block walls had to be re-treated. Also the exterior stone work has progressed excruciating slowly and is nearly 100% over budget. Still, while I am generally very happy with the result, I am beginning to get impatient to be done with the dust and the mud, and I want to return to my normal life. Most of all, I am anxious to begin work on the A&BR2.

I think the interior finishes will go quickly, but finishing the stone work will probably take another 6 weeks, and then another week or two to finish grading and lay the cobbles in the court yard. So perhaps early March?  Just in time to plant a little grass and some shrubs. Or at least one can hope.


January 2017

Wait Time Preperations


It is frustrating. As the new train room nears completion, I am anxious to get started, but unable to begin. Still there is plenty to do, if I can force myself to address some rather long, arduous, and tedious tasks. For one thing, I have 300 30 inch sections of Atlas Code 55 Flex Track to weather; and for another, I have 50 more dwarf signals to construct. I began working on weathering the flex track yesterday. My goal is to weather five or ten sections a day for the next month or so. Working in batches of five or ten, it takes a lttle over ten minutes to weather one section. Along the way I will construct the 50 dwarf signals in two batchs of twenty-five. Divide and conquer.

wft Using dark brown and a light tan "camouflage color" and a medium gray primer, all in spray cans, I put on a heavy coat of the dark brown making sure to coat the side of the rails, followed by a misting of gray and then of tan., wiping the rails clean immdeiately after each coat. Check out the 6-8-2016 entry in this blog for details regarding my technique. I then paint the spike-heads rust color and use a slightly dried out felt-tipped marker and a straight edge to dark the oil stain line in the center of the sleepers.

February 2017

Train Room Takes Shape

The sheetrock in the train room is hung and taped. At last, things are beginning to take shape. The next moves will be to skim, sand, and prime, and then, before the bench work goes in, to install the lights and other electrical in the walls and ceiling, and to paint the aisle ceilings and lighting soffit aisle-sides black and the backdrop and the inside of the lighting soffit sky-blue. Next, I'll construct the basic open frame bench work, bench legs, and install the wire mounting board. Finally I'll  complete the AC wiring in the bench and construct the control panel cabinetry. Then I'm ready to begin layout wiring, construct roadbed, and lay track.


Creating Signs

The Southern Railway used "head-of-train" stop markers to position a train at a platform. I think the protocol was to use a "S" sign for all trains, or to use a number (e.g. "5") marker for trains with a specific number of cars. On the A&BR2, I'll use "S" signs with a red cap to indicate the locomotive stop position beyond a platform. Like the prototype, I'll also use "W" markers with a yellow cap to locate the exact point before a crossing where a warning whistle is to begin. Finally, I will use numbered signs with a white cap to identify all turnouts by their name, which corresponds to the turnout's loconet address.

Straight pins painted white with signs made in MicroSoft WORD using the table function printed on light card stock. Turnout signs are two sided, "S" and "W:" sign are one-sided. 

March 2017

Train Room Initial Wall Painting Complete

Progress on the new building is excruciatingly slow, but the end is in sight. The painters were at work in the train room last week; and with a little more electrical work, I will be able to begin constructing the bench work. In the meantime, I'll begin painting clouds on the backdrop walls.

trp Here is a photo of one corner of the train room. The backdrop walls and the lighting soffit above the bench are painted sky blue and inside of the soffits are painted black above the viewing aisles. The bench will be 44.5 inches above the floor. That is even with the top of the open stud wall on the left. It separates the behind-the-backdrop service aisle from the train room and allows for under-bench access from both sides of the bench. 

Painting Clouds on the Backdrop

At last I am at work in the train room, and it really feels good after nearly two years of planning.

My first step will be to paint clouds on the backdrop all the way around. Since there are almost 120 running feet of backdrop, this will take a while. On the A&BR1, I created very simple, very stylized tall, light cumulus clouds. I painted them first, and then added my layers of mountains that form the horizon on top of them, so the illusion is that the clouds' rise from behind the mountains.

Keeping the backdrop simple. In this scene on the Altamont and Blue Ridge 1, there is a tree line at the top of the rise of the 3 dimensional hill in the distance, then a cut-out mountain made by gluing foliage 1/2"  homasote, and finally two layers of hazy, distant mountains painted over the clouds on the back drop. This gives the illusion that the clouds are rising from behind the mountains.

On the A&BR2, I again plan to paint cumulus clouds but with a little more detail and depth.

fc The very first project on the A&BR2: painting clouds. John Constable, move over!

I'll paint the sky below the cloud banks a lighter shade of blue, and I'll feather this lighter color into the darker sky above in the clear sky patches between the cloud banks. This overall approach should enhance the illusion of depth.

clouds I am no great mural painter, but I was able to get pretty good results after a bit of practice. I got a panel of sheet rock, cut it into 24 x 36 sections, and painted them the sky color. Then on the Internet I read  articles and watched videos regarding cloud painting techniques. Most of these tutorials deal with painting with oils or acrylics, so I had to alter  the technique slightly to work with plain old flat latex wall paint and larger brushes This is not rocket science, and after a few trys, I got the hang of it, and developed some techniques that work pretty well.

Notice here my utility work cart (a restaurant busing cart) with two lighting posts clamped to the side. This supplies plenty of space to keep brushes, paint, water-wash, rags, sponges etc. close at hand. Normally I use this cart for tools and stuff I use regularly in the train room. Without the lighting posts, it rolls right up under the bench and is very handy especially when working in places where the layout is nearly finished and there is no place to stage your tools, paints, glues, landscaping  materials  etc. Highly recommended.

I'll not add the distant painted mountains now, as I like to wait until the landscape is contoured and then, working front to back, I can fashion a believable transition across the point where the backdrop attachers to the bench.  

Rethinking Painted Clouds

After painting several large clouds like the one above on about 20 feet of the backdrop mural painting, I began to get a bad feeling about the arrow-straight cloud bottoms at a uniform height. They look like soldiers in a row - unnaturally predictable. The technique is good , but the composition is flawed.

cir Clouds in a row.  This just isn't quite right. 

My idea was to have the clouds float above the mountains, but this "soldiers in a row" impression worries me. After observing clouds in nature, searching the Internet again to look at backdrop painting work by others, and studying a few famous landscape canvases, I have decided that it is best to do away with the perfectly flat bottoms and to paint the cloud banks all the way to the bottom of the back drop, thus again reverting to the illusion that the clouds rise from behind the mountains. When the layout is complete, a good portion of the bottom of these rather large new cloulds will be covered by mountains.

cr Here's the revised effort. This should work much better. I know this looks like a lot of coulds, but over half of the bottoms of these will be hidden behind the mountans when the scenery is competed.

This work is very forgiving. If you want to change something, just lightly sand the surface of the old pianted cloud, paint over it with the sky color, let it dry, and start over.

The mountains will cover anywhere from 8 to 15 inches of the bottom of these large cloud banks, so they will not be so overwhelming and the end effect should be good. Just ot be sure, I will build a short, little, mock-up section of the bench with a few sturctures, and some mountains made with homasote and styrene cut-outs. Then I'll tack it in place. This will allow me to fully access my new cloud painting approach before painting the rest of the room.

Working Drawings for the Hinged Bench Tops at Altamont

It has been almost two years since I did the first drawings for the hinged bench tops that will allow access to the hidden yard  below the City of Altamont (see post of 9-7-2015). Now that the backdrop walls are complete, I have made detailed drawings for the construction of these hinged bench top sections. On the Type B wall, where there is access to the track underneath the city from behind the backdrop wall, I will hinge the bench top section from the stationary raised double track runners on the aisle side of the bench. On the Type BB wall with no access from behind, I will hinge the bench top section from the backdrop wall so it will open the other way and allow access to track underneath from the viewing aisle.

hsb This section will hinge from the stationary riser supported by the 2 x 4 braces to the right. The hinged section will allow the City of Altamont to be laid out on a terraced parallel grid, three stepped terraces, each 1.5 inches above the next. The lower, stationary terrace will accommodate the mainline and the siding, a row of warehouses and stores, and finally a street next to the bench edge. The middle level will accommodate two rows of buildings facing the main street (a divided two lane affair with a streetcar track in the center). This layer will have to be notched  at the cross-street intersections to allow for rather steep streets going down to the level below. The upper level will accommodate two rows of buildings facing a 1.5 inch, two-way street (the back row of buildings will be narrow "background" buildings), and there will be homasote cut-outs to form two tree lines set out from the backdrop wall. 
hbb The hinged bench top to access the rest of the track below the city of Altamont will hinge from a 2x4 runner nailed to the backdrop wall. This will allow access to the hidden yard from the viewing aisle side of the bench. Note that there is no upper level here, only open bench frame for low hills The city will narrow here for a residential district on the north side and a warehouse district above Altamont Yard on the south side.

Elavuating the Painted Clouds on the Backdrop

Just to make sure my towering cloud banks on the backdrop work with the scenery, I have constructed a little mock-up using a piece of homasote at bench top height as a temporary bench top, a couple of rough mountain cut-outs, and a few half buildings designed to fit right up against the backdrop. Rough and crude as this scene is, the clouds against the new clouds against the mountains work pretty well in this context, so I will continue painting.

ct This this a very crude  rough-in, but my new cloulds seem to work pretty well against the mountains, especially since this is a a mock-up at bench top height. Many of the strucutures on the layout will be at higher elevations than this, up to 8 inches above bench top level, and some of the the 3 dimensional mountains will rise above that. So in spots, only the top of the cloud bank will be visible. 

April 2017

Working with Mountain Silhouettes on the Backdrop

I am well over half way done painting clouds, and I should finish next week. Before I move on, I think it is now good time to work out the way I want to handle the mountain forms on and against the backdrop, even though I'll not be putting any mountains on the backdrop any time soon. Their exact positioning and coloring will depend on the 3 dimensional scenery on the layout in front of them. Still, it is a good idea to work out exactly how I will proceed when the time comes, and once I work this out, I can create a more finished looking, to-scale mock up to re-check the look of my clouds against mountains in a more fiinished form.

On the A&BR1, I used a 6 tier system. 
er Layer 6 was a low tree line made with homasote painted dark green and covered with clump foliage. Where I could, I placed this cutout tree line about 1/2 inch in front of layer 5 which is screwed to the backdrop.

Layer 5 was a 1/2 inch homasote cutout painted dark green with clump foliage glued on. I sprayed this with a duller, less vibrant green and little gray for distancing.

Layer 4  was a .040 styrene cutout painted dark green with ground foam glued on, and then spayed painted to look kind of misty.

Layers 2 and 3 were misty painted .040 styrene cutouts.

Layer 1 consisted of one or two ranges of very distant mountain silhouettes painted right on the back drop.

On the A&BR1, I struggled to find a satisfactory method for attaching my cutouts to the backdrop wall. I wanted to to move them around to get the positioning just right and then attach them in a more permanent way. I tried double-sided tape, various glues, sheet rock nails. In the end I got things in place, but the methodology is not consistent.

Upon reflection, I now think the best appraoch on the A&BR2 will be not to paint mountains on the wall at all, but rather to use painted, thin, .020 styrene cutouts for my distant mountains, and then to attach  all of my cutouts to the thicker homasote or 1/2 inch plywood layer in front, first using double-side tape for positioning, and then using very spot glue welds, short screws, or staples for the permanent installation. This way I will end up with a nice sturdy, removable, homasote and styrene sandwich which can be attached to the wall with two long sheet rock screws - neat, clean, modular, easy to remove and to change around and adjust. 

I also want to experiment with the painting technique for the middle layers of styrene cutouts - perhaps a little more detail, perhaps I can develop a technique for painting mid-level silhouette cutouts that looks better than the foam covered cutouts. Perhaps fewer layers will look better. (Contined: See entry of 4-4-17) 

Testing the HUE Lighting System in the New Train Room

One of the things that has been on my mind regarding the lighting scheme in the new train room involves the distance of the new building from my house and the range of both my WIFI home network and the HUE control bridge. This morning I took my little WIFI range extender out the new building and plugged it into an electrical outlet in the bedroom closest to the house, which is just above the train room. It acquired the WIFI signal nicely. I then plugged the HUE bridge directly in to the WIFI extender, and went down stairs and plugged in a HUE bulb in the train room. With all of this in place, I used my laptop to wirelessly connect to my WIFI Ethernet home network, and once connected, I ran one of the programs I have written to acquire and address the HUE bridge. Viola, it all worked fine. I am not worried about HUE bulbs elsewhere in the train room, because if any bulb gets a signal, it will relay it to adjacent bulbs.

Working Drawings for the Bench Work

Since the froward edge of the bench work lies directly below the aisle-side edge of the lighting soffit, it will be any easy matter to lay out the bench on the floor using just a plumb bob and a chalk line, a bench work plan is not really necessary. All that is needed is an elevation drawing of a cross section of the bench detailing the various sized lumber and plywood components. I have also included a detail of the concave and convex bench-corner construction.

bwp I have not drawn all the lateral supports, which will be initially tacked in using sleet rock screws. These will be 2 feet on center and will later be adjusted where needed to accommodate switch machines etc.
bwc Bench work cross section detail
Bench work corners

Further Thoughts Regarding Mountain Silhouettes

As I paint clouds on the backdrop, I have been giving some thought to the mountain silhouettes that will go in front of them, and I have boiled this down to five elements or five layers if you like. Closest to the viewer, there will be either a 3 dimensional mountain that abuts the backdrop wall or a low tree line (about 3 iches high), which will be made by gluing coarse clump foliage to painted homasote or 1/2 inch plywood. Ideally, this first homasote tree line silhouette will be attached to the bench about 1/2 inch out from the backdrop wall. This 1/2 inch space adds depth and allows a space to place small silhouettes of distant buildings, steeples, water towers, or smokestacks etc. but to be effective, it must be viewed from the front. On the A&BR2 there are many places where it is possible to look down the line of the layout, and in these cases the space is visible. This is not ideal, so in many places I will use only 1/4 of space, or even attach the tree line cutout to the first mountain cutout. Next, there will be a thick mountain silhouette against the backdrop wall with coarse foliage glued to a painted (black) 1/2 inch board.  Behind that, a painted .040 styrene cutout with fine foliage stapled onto it, and behind that, maybe a .040 styrene sheet with green turf glued to it. (I want to experiment with painted highlights and shadows on the glued-down clump foliage layers and on the turf textured layer as well. On the A&BR1, I used the truf covered cutouts, and I am not totally happy with the look of these. Still, there may be a way to highlight these with paint and make this technique more effective)  Finally,
a painted mountain cutout depicting several layers of distant mounts on 020. styrene cutout  (This  layer will require some painting technique to create distant misty mountains with solid light blue very distant mountains behind.) I also want to experiment using light gray and light blue spray paint to achieve increasingly misty ranges front to back.

For more information about this technique, see my tutorial "Marrying the Layout to the Backdrop" elsewhere on the web site. This tutorial is all about developing methods to effectively hide the point of transition from the 3 dimensional layout space to the flat painted backdrop. The layered mountain cutout technique makes this transition very gradually in the last inch or so of the 3 dimensional space. When you think about it, the actual transition from layout to backdrop does not take place here until the fianl painted distant mountains.  

This is an ambitious scheme, and it will take some doing to get it fully worked out. I think there is a danger here of getting things too complicated, so I'll have to be careful as I progress. Certainly, I will not use all of these layers in any one spot. I suspect 3 or 4 will work fine in any one spot of the backdrop. Whatever the case, now seems a good time to construct a test module, evaluate it against the cloud backdrop, establish scale and fix the horizon level, and to work out details and techniques for future construction.

Getting the scale right is critical. I have spent a good deal of time over the last few days trying to determine how high above the bench the horizon should be. This is tricky because in places the 3 dimensional scenery will rise as much as 5 inch above the bench top, and it might be unrealistic to have the horizon rise up in all of these spots. In order to help visualize this scaling issue, I created a scaled sketch depicting in overview 14 feet of the backdrop wall behind the town of Fitzhugh. Then I drew to scale what I think is a realistic overview of the scene. I next selected a 4 foot section of this overview, and created a more detailed scale drawing of this showing all the layers. I will build my 4 foot long test mock-up from this.

mp Preliminary sketches. I think a horizin 12" above bench top will work, but I will know better when I get this mock-up built. Notice that in this section the roadbed grade rises from 3 to 4 inchs above the bench, so with a 2 or 3 inch tree line, this leaves only from 5 to 7 inches or so for the layered mountains. 

Working with these sketches and experimenting with rough cutouts, I have concluded that 12 inches above the bench will probably work best, except on the east wall where the highest streets of City of Altamont meet the backdrop 8 inches above the bench-level hidden yard below. Here, I will cheat the horizon up to 16.5 inches above the bench. With a roughly 3 inch tree line, this will give me 5  inches for layered mountains. Since there are high, 3 dimensional mountains in the corners of the room that flank both sides of the City of Altamont and separate this east wall from the adjoining north and south walls, this transition to a slightly higher mountain horizon along the entire east wall should not leap out at the viewer as an inconsistnacy.  

Once I have the first three layers on my mock-up completed and assembled, I can work on the painted layers until I develop a technique that works and is repeatable. It is clear from this that the painted layers will still have to glue either to the wall or to the back of the .040 style layer, so I will continue to try to find a glue that will not warp the styrene, set up firmly enough for the long haul, and still allow me to disassemble these pieces if I decide to make a change. 

In preparation for painting the critical fourth .020 styrene layer, I have been looking at the mountains around me and watching mountain painting tutorials on line. Although the below photo is a winter scene taken from my front porch, it illustrates the terrain I what to reproduce. I can clearly now see that the mountains on the A&BR1 a far too steep. The Blue Ridge Mountains are rugged in places, but from a distance they present a very soft panorama.

ms The view from my front porch. A soft panorama. A summer version of this view is the terrain I want to recreate. It is interesting to note the light source is from the right here. This is a detail I will observe in order to achieve depth and and interenst and to  make images on all four walls appear to have a single light source. I have done this already with my clouds. On the west wall, the clouds are front-lit; on the east wall they are back-lit; on the south wall the light comes from the left , and one the  north wall it come from the right. This is all consistent with a summer afternoon, with the sun in the Western sky. I can do the same thing with the mountains. It is a subtle point, but every little detail adds a little something to the overall effect. 

Constructing a 4 Foot-long Mountain Silhouette and Checking the Scale

Below is a photo of the framework of the 4 foot-long, mountain layer backdrop module, which I constructed as an in-place test against my clouds to determine scale, horizon height, and to develop texturing and painting techniques.

Here is the painted, roughed-out, 4 foot-long, mountain-layer mock-up positioned next to the future 1.9% grade at Fitzhugh, which at this point is between 3 and 4 inches above the bench top. At its highest points the silhouettes are about 8 to 9 inchs high forming a horizon that is about 12 inches above the bench top. This horizon line and the scale look good to me. I tried this with slightly higher mountains (15 inches above the bench) and that did not work as well.  There is a 2.5 to 3 inch 1/2 inch ply wood tree line layer in front, with a plywood layer for clump foliage behind that; next a .040 styrene layer that will be covered with fine foliage; and finally the .020 distant blue mountain layer, upon which several nearer ranges may be painted. At this point all the silhouettes are painted black except for the distant light blue mountain silhouette in the back. The application of various shades of foliage, misting with spray paint, creating shadows on the side away from the sun (in this case the left side) and painting highlights will follow. 

Continuing with the construction of the 4 foot test mockup, I next added some Woodland Scenics clump foliage to the first two layers. In the photo below we see the tree line and the first layer of moutains covered. The next range of mountians (which will get finer foliage) and the light blue distant mountain cutouts are uncovered and unpainted respectively at this stage.

mm2 Here the tree line, which is tight-up against the first layer of mountains with no space in between, is covered with medium green clump foliage with a good bit of light green foliage mixed in, especially along the top. The light green top helps to highlight the silhouette and to articulate it from the darker, more uniformly colored layer of mountains behind, which are covered with a combination of medium green and dark green clump foliage. I think when I mask and apply a little spray paint I can make the tree line stand out a bit more by highlighting the top of the tree line with a little very light green or yellow and by spraying a subtle mist of gray on the first layer of mountains behind. The second range of mountains behind that, which will be covered with finer foliage and will be sprayed with even more gray and some light blue, and the next range, which will either be painted a misty blue-green-gray color (or perhaps just a flat color or perhaps with more painted detail) in front of the light blue horizon ridg . Since this mountan layer "sandwich" is not attached to the wall at this point and since all of these layers easily come apart, all of this spary painting can be done on my paint-bench and the masking or shielding from layer to layer will be easy. Once I establish the spraying shades for the different layers, future layers can be sprayed before assembly and there wil be not need for tedious shielding or masking. 

Rethinking the Tree Line and Completing the Mountain Mockup

The more or less flat, uniform tree line layer on my Mountain "sandwich" is not working fully for me. I have used this technique on the A&BR1, but I always had a row of low buildings in front to break up the regular shape. Here the track will be pretty much right up against this tree line, so no nearer modeling can be inserted in between. To solve this, I have decided to add a flat photograph on the forest floor in places along the length of the tree line. Below are photos detailing this simple process.

muff1 Here is the "before" - a straight uninterrupted tree line right down to the  future roadbed.
Here is the cut out to be pasted in  -  a photo I found on the Internet and sized, cropped, and retouched in photshop to make it the right scale and soft, dark, and misty. This is printed on 110 point card stock.
Here is the "after" photo with the photo cutout  in place. I  glued on a few 3 dimensional tree trunks made from painted wire, toothpicks, and kitchen skewers to establish a foreground and to push the misty photo to the back. Finally, a few clump foliage bushes at ground level break up the straight line on the bottom of the form.

Below is the completed 4 foot section in place against the backdrop, with the light green highlight sprayed on the top of the tree line, the next layer of mountains sprayed with a little medium gray to give it a little distance, and the next layer of mountains covered with course turf and sprayed, first gray, and then liberally with light blue to enhance the contrast and add more distance. Also I altered the clouds a bit to better frame the scene. Not Bad. I am still a little unhappy with the look of the next to the last layer, that is the .040 styrene layer covered with corse truf. It is a little ragged looking and the coloring is not yet just right. It work OK where we just see the the top edge of it, althoug it could be lighter still in color. However, where there is a large portion of this mountain exposed, it looks uneven in texture. I will continue to experiment to refine the technique for this layer. The rest looks good to me. 

mc After a bit of experimentation, I found that the mountain module looked best without the extra layer of far mountains that I had planned. The light blue cutout  looks great against the clouds. Adding another  layer in front of it, no matter how well painted, proved distracting, So I'll keep it clean and simple. I have not added the shadows to any of these mountains to establish the light source as planned. I may experiment with that later, but something tells me this will work better without that complication. 

Around the layout, the foreground will vary, so all of these modules will vary to accommodate what is in front of them, but this is the general model I will follow: at low tree line, a near mountain range of clump foliage, a misty medium-distance mountain range of corse turf, and a distant mountain range painted solid light blue.

Cloud Painting on the Backdrop Complete


The completed clouds may look a bit over-the-top, but remember only the top half or so will show once the mountains and the foreground scenery is complete. Also I will re-touch each section of  clouds to fit the scenery in front just as I did with the 4-foot section above. (The cutout section of wall on the far side is for behind-the-backdrop hidden yard access under the City of Altamont. This may be higher than needed, and so I may sheet rock over some or all of this depending on the clearances afforded by the hinged top that will go in front of it.) 

Test Section of the Bench Completed

The first section of the bench work for the A&BR2 was completed today. The 8 foot section allows me to evaluate the new bench height (44 1/2 inches from floor to the top of bench level, plywood and homosote roadbed.) From this prototype, I can also get a feel for accessiblity of the 12 inch wire mounting board afixed to the recessed legs behind the 8 inch fascia board.

bt Bench height seems good. The 36 inch front-to-back dimension is a reach but I think workable. Access to the wire mounting board is pretty good if you are seated on a low bench or stool, but the top two or three inches are perhaps a bit hard to get to. I don't want to lower it, because the duck under is already pretty low, so I plan to reviese the wiring layout for the board so that wires in the top few positions are wires with very few intermendiate connections like loconet, signal cables and system grounds, etc.

The only other change involves the height of the fascia. I have raised it up 1 1/4 inches so that it will cover the edges of the plywood/homasote layers of decking or roadbed. This will necessitate making it 9 1/4 inches wide instead of the 8 inch width pictured here. 

Basic Bench Work Completed


Watching professional carpenters work is a wonder. Two men did this in a two days, and they did it better than I could have done working alone for many weeks. Below are a few photos of the finished basic bench work.

b2 Looking east down the aisle between the center and  the south sections. The center setion is huge - 5 feet wide and 22 feet long. I'll not lay track on this section until I have pretty well finished the rest so I can use it as a work bench and staging area for materials. Here we see some plywood ready for future use.
b3 Looking east down the aisle between the center and the north sections. In the left foreground we can see on of the two 5 foot wide turn around sections by the door. Track access in tunnels on these large sections is going to be tricky. Also construction of the mountain in the difficult-to-reach center of the turn around section will be a challenge,

Now that all of the bench work is in place, I can fully evaluate its design in detail. In general, it looks very good indeed. However, I can see a few potential problem areas.

First, coming out a full, 36 inch from the backdrop wall was ambitious. I knew this. With the bench in place, I now feel perhaps 32 to 34 inches would have been better, but there are not many places where the track is right up against the backdrop wall and I am tall, so I do not think this represents a real problem, just a little stretching when working close to the backdrop wall. I feel that wider is better as long as I can reach everything without too much strain.

Second, and certainly a little more problematic, are the two 5 foot wide turn-around sections of the bench by the door. This comes as no surprise. I wanted large radii for the turnarounds, and I knew that the center of this area would be difficult to reach, so I designed this section to accommodate a large mountain with the track turn-around radii in a tunnel hugging the outside edges next to the bench edge and next to the backdrop wall. With the sheet rock cut-away in the backdrop wall
beneath-the-mountain and with a cut-away in the 5 foot side panel I should have adequate access to all the track in the tunnel from behind the backdrop walls and from the side walls by the entrance. That said, to insure really complete track access, I will have to place a few turnouts inside the tunnel and make the mountain footprint a little larger than I had originally envisioned and then cut away a slightly larger section of beneath-the-mountain sheet rock from the backdrop wall. I am a not delighted with the prospect of turnouts in the tunnel, but with this done, I don't think I'll have an access problem.

Another, and perhaps better, way to solve this might be to leave the mountain footprint smaller and the turnouts outside the tunnel as planned and then to devise a way to access these difficult-to-reach turnouts from behind the backdrop via a removable section of the against-the-backdrop mountain silhouette framework to reveal a sheet rock cutaway just behind these turnouts. This plan would involve only two removable sections of silhouette, so this might be the best and the easiest solution.

The construction of the central part of the mountain itself will be a challenge. It is definitely in a bench-top area I cannot reach. The solution here, I think is to construct the large central section of the mountain on the work bench and then poison the finished section on the layout and tie it into the adjoining on-layout, stationary parts of the mountain. I can probably manage it so that this central mountain section can be removable to provide really wide open access to the entire section if need be.

How all of this will work out remains to be seen. I wanted good track access everywhere, but I also wanted large, more realistic radii. There is a trade off here, and it becomes a balancing act. My best guess that the 36 inch bench will make me stretch a little but that's OK, and although building the two central mountain sections as removable modules will be tricky, it can be done if required. If the under-mountain access thorough the cut-aways feels adequate to me, I can forget making the central mountain section removable and just tie it in permanently. This will of course be much easier, and I can probably build it in place by leaving a portion of the adjacent bench-top open for me to stand in during its construction. Certainly making it permanent is better, but I'll do so only if I am able to easily access all covered or out-of-reach track and turnouts on this entire section through the wall cut-aways.

Last of all, seeing the massive bench in place reinforces my concerns regarding access to the hidden yard beneath Altamont City. I have been sensitive to this from the beginning and throughout the design stage. Accordingly, I have paid close attention to the design of hinged tops that will afford hidden yard access: one from behind the backdrop wall and one from the end of the southern viewing aisle. I will build and evaluate these hinged lids as soon as the roadbed for the hidden yard is in place, and should there be any problems, I'll take whatever steps necessary to assure good open access to all nine track in the hidden yard from one end to the other, even if I have the redesign the terracing for city of Altamont.

Taking Stock and Planning the Next Move

Looking at this enormous bench is daunting indeed. I am beginning to obsess about future problems and hard to reach areas, and this is not good. I need to slow down, and take things one at time. All of my potential problems, whether they be real or just perceived, will be addressed in due time. What is needed now is an initial plan of attack. I'll begin with a task list in logical order with some distinct priorities.

First on the list will be a through cleanup. Dust down all the walls, vacuum and sweep thoroughly, and mop the floor. Then I'll build some small work benches in the corners of the behind-the-backdrop area and some shelving and see to work lighting in this area. Once this is done. I can move my tools and all my supplies and materials from my current shop into the new space. Then I be ready to begin work of the layout.

Layout construction will begin with the installation of a temporary control panel. I'll then have the electrician install all of the outlets in the bench work and the switches for layout power and room lights in the temporary panel. When this is done, I will install the Command Station/Booster, mount both of the PM42 Power Managers and all five of the BDL168 boards, run AC power and track power and grounds from the DCS100 to these components. Next I will roughly position and wire up track power to all of the prefabricated turnout modules I built last fall. After some testing to insure I am getting good, clean digital signal to all turnout modules and that they all successfully detect shorts both when wired directly to the command station and when wired through a power manager, I will be ready to begin building roadbed and laying and wiring track to connect these modules. 

Alphabetical Index

What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Giving Some Thought to the Backdrop Wall Design 7/25/2015
More About Marrying the Layout to the Backdrop 12/31/2015
Laying Out the Backdrop Walls and the Overhead Lighting Soffit
Backdrop Wall Initial Framing Completed 11/11/2016
Painting Clouds on the Backdrop 3-20-2017
Rethinking Painted Clouds 3-26-2017
Working with Mountain Silhouettes on the Backdrop 4-1-2017
Further Thoughts Regarding Mountain Silhouettes 4-4-2017
Constructing a 4 Foot-long Mountain Silhouette and Checking the Scale 4-6-2017
Rethinking the Tree Line and Completing the Mountain Mockup 4-11-2017
Cloud Painting on the Backdrop Complete 4-13-12017

What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Still MoreTweaking of the Benchwork and Track Plan 9/6/2105
Accessing the Hidden Yard 9/7/2015
Thinking About the Bench Work 2-19-2016
Elevation Calculator 9-19-2016
Laying Out the Backdrop Walls and the Overhead Lighting Soffit
Backdrop Wall Initial Framing Completed 11/11/2016
Working Drawings for the Hinged Bench Tops at Altamont 3-25-2017
Working Drawings for the Benchwork 4-3-2017
Test Section of the Bench Completed 4/18/2017
Basic Bench Work Completed 4/18/2017

Design and Planning
What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
New Layout Design 8/5/2015
Planning the Flow of the Landscape: Tweaking the Track Plan 8-7-2015
Road Grading, Structure Pads, and Town Terracing: Preliminary Considerations 8-8-2015
The Wiring Scheme: Thinking About Detection Blocks 8/10/2015
The Naming of Power Districts, Detection Blocks and Turnouts 8/11/2015
What Do I Need 8-12-2015 (updated 9-20-2015)
Selecting Vendors 8/13/2015
Further Tweaking of the Track Plan 8/14/2015
Tweaking the Wiring Schematic 8/15/2015
More Work with Block and Turnout Names 8/17/2015
Block Lists 8/19/2015
Turnout Lists 8/20/2015
Still More Track Plan Tweaking 8/22/12015
Thinking About the Yard 8/25/2015
Signal Placement and Terminal Track Plan 8/26/2015
Further Considerations Arising from a Smaller Trainroom 9/1/2015
Still MoreTweaking of the Benchwork and Track Plan 9/6/2105
Accessing the Hidden Yard 9/7/2015
Checking Elevations and Grades using XTrackCAD 9/9/2015
More Track Tweaks 9/12/2015
Rethinking the Removable Corner Mountains  and Resizing the Corner Radii 9/19/2015
Bridge, Viaduct, Over/Underpass, and Culvert Planning 9/21/2015
Road, Terrain, and Topography on the Revised Track Plan 9/22/2015
Another Design 8/30/2015
Labels 11/26/2015
Designing and Locating the Main Lighting and Electrical Panel 1-15-2016
Revisiting the Track Plan 1-24-2016
Do-ahead Projects 2/10/2016
Revisiting the PM42 and BDL168 Barrier Block Labeing Scheme 2/11/2016
Relocating the PM42 and the BDL168 on the Layout 2-17-16
Mounting PM42s and BDL168s 2-18-16

Update to the Train Room Floor Plan Including the Under-the-Slab Conduit Locations 2-25-16
Labeling the Wire Mounting Board 3-1-2016
        Having to Rethink DS64 Turnout Addresses 6-11-2016
        Assigning Board Address Numbers 6-13-2016
Train Room Lighting and AC Electrical Revisited 11-8-2016
Laying Out the Backdrop Walls and the Overhead Lighting Soffit 11/9/2016
       Final Train Room Lighitng and Electrical Plan 11-28-2016
        Final Tweaks to Block and Turnout Names and Addresses and to the Power District Layout 12/2/2016
        Working with Mountain Silhouettes on the Backdrop 4-1-2017
Further Thought Regarding Mountain Silhouettes 4-4-2017
Constructing a 4 Foot-long Mountain Silhouette and Checking the Scale 4-6-2017
Rethinking the Tree Line and Completing the Mountain Mockup 4-11-2017
Taking Stock and Planning the Next Move 4-22-2017
Elevations and Grades
Checking Elevations and Grades using XTrackCAD 9/9/2015
Elevation Calculator 9-19-2016

Lighting, Layout
Layout Lighting Power Bus and Lighting Distribution Blocks 2-29-2016
Aux Power Supply 3-22-2016
Prototyping Layout Lighting Distribution Blocks 6-1-2016

Lighting, Train Room
What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Further Thoughts Regarding Sound and Lighting 7-23-2015
Rethingking the Train Room Lighting Scheme 12/5/2015
More about the Phillips HUE lighting system 12/13/2015
Preliminary Testing of Phillips HUE System 12/18/1015
Phillips Hue Program Up and Running 12/19/2015
Installing and Programming a Phillips Hue Lighting System on the AB&R1 1/9/2016
Tweaks to HUE Lighting Sunrise and Sunsets Completed 1/30/2016
Servo Controlled Dimmer for Trainroom Fluorescent Lights 4-11-2016
Programming the Lights on the A&BR 4/14/2016
          Train Room Lighting and Electrical Control Panel 10-31-2016
        Train Room Lighting and AC Electrical Revisited 11-8-2016
        Final Train Room Lighitng and Electrical Plan 11-28-2016
        Manipulating Color in HUE Light Programs 12-7-2016
Testing the HUE Lighting System in the New train Room 4-2-2017

What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
HO vs N Scale 7/17/2015


What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Planning the Flow of the Landscape: Tweaking the Track Plan 8-7-2015
Road Grading, Structure Pads, and Town Terracing: Preliminary Considerations 8-8-2015
Rethinking the Removable Corner Mountains  and Resizing the Corner Radii 9/19/2015
Bridge, Viaduct, Over/Underpass, and Culvert Planning 9/21/2015
Road, Terrain, and Topography on the Revised Track Plan 9/22/2015
Weathering Track 1/19/2016
          Painting Figures 9-10-2016
        Painting Clouds on the Backdrop 3-20-2017
        Rethinking Painted Clouds 3-26-2017
        Working with Mountain Silhouettes on the Backdrop 4-1-2017
Further Thought Regarding Mountain Silhouettes 4-4-2017
Constructing a 4 Foot-long Mountain Silhouette and Checking the Scale 4-6-2017
Rethinking the Tree Line and Completing the Mountain Mockup 4-11-2017
Cloud Painting on the Backdrop Complete 4-13-12017

Signs And Signals
Signal Placement and Terminal Track Plan 8/26/2015
Dwarf Signals 11/21/2015
Signals and Signal Placement 2-24-2016
        Signal Protocol, Programming  and Placement  7-4-2016
        Creating Signs

What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Further Thoughts Regarding Sound and Lighting 7-23-2015
        Windows 10 Problem vis a vis 7.1 Surround Sound Box (ENCAB-8CM)

Bridge, Viaduct, Over/Underpass, and Culvert Planning 9/21/2015
DPM Kits 5-10-2016
A DPM Kit Bash 5-24-1016
Lighting a Covered Station Platform  9/23/2016

Track Work

Atlas or Peco Track? 1/13/2106
Switching to Atlas Code 55 1/18/2016
Weathering Track 1/19/2016
Testing the Atlas Code 55 Turnouts 1/22/2016
Turnout Templates 1/29/2016
Some Thoughts Regarding Cork Roadbed 1/31/2016
An Easement Template 2/4/2016
        Beginning to Lay Track - Peparing to Construct Prefabricated Sections 6 -7- 2016  
Weathering Turnouts 6-8-2016
        Wiring Turnouts 6-9-2016
          Laying Track 6-19-1016
          Installing Tortoise Switch Machines and Tortoise Interfaces on the Prefab Section 6-20-2016
        Testing Installed Atlas Code 55 Turnouts and the Prefab Section Wiring 6-22-2016
Track Laying Using Xtrack CAD 8-17-2016

Train Room Construction
        Architectural Drawings for the New A&BR2 Train Room 1/12/2016
        Update tothe Train Room Floor Plan Including the Under-the-Slab Conduit Locations 2-25-16
        Breaking Ground 7-5-2016
          Footings Poured Today 7 -13-2016
          Block Walls - Train Room Takes Shape 7-28-2015
          Foundation Walls Complete  8-12-2016
        Slab Poured  9-7-2016
          Floor Joists Installed  9/24/2016

        Contemplating the Completion of the New Building

        Laying Out the Backdrop Walls and the Overhead Lighting Soffit

Backdrop Wall Initial Framing Completed 11/11/2016
          Lighting Soffit Framing Complete 12/1/2016
        Construction Blues 12-22-2016
        Train Room Takes Shape
Train Room Initial Wall Painting Complete 3-19-2017

TrainController (Railroad & Co)
What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
Creating a Switchboard in TrainController 8/21/2015
TrainController Gold 4-10-2016
Some File Management Housekeeping for TrainController 4/23/2016
Signal Protocol, Programming  and Placement  7-4-2016


What Would I Do Differently? 7/15/2015
Things I Would Not Change  7/16/2015
The Wiring Scheme: Thinking About Detection Blocks 8/10/2015
The Naming of Power Districts, Detection Blocks and Turnouts 8/11/2015
Tweaking the Wiring Schematic 8/15/2015
Tortoise Interface 11/4/2015
PM42 and BDL68 Interfaces 11/10/2015
Labels 11/26/2015
Reconsidering Twisting and Bundling of Power Buses - The Bus Wire Mounting Board 11/30/2015
Wiring Color Scheme 12/1/2015
Some Thoughts Regarding Droppers 12/1/2015
Tortoise Interface Prototype 12/3/2015
Designing and Locating the Main Lighting and Electrical Panel 1-15-2016
Switching to Atlas Code 55
Revisiting the PM42 and BDL168 Barrier Block Labeing Scheme 2/11/2016
Relocating the PM42 and the BDL168 on the Layout 2-17-16
Mounting PM42s and BDL168s 2-18-16

A Power Bus and a Prototype Mounting Board for DS64s 2-26-16
Layout Lighting Power Bus and Lighting Distribution Blocks 2-29-2016
Modifications to the Tortoise Interface 2-29-2016
Revisiting the PM42 and BDL168 Barrier Block Labeing Scheme 2/11/2016
Labeling the Wire Mounting Board 3-1-2016
Installing Turnout Interfaces 3-4-2016
Aux Power Supply 3-22-2016
T Splicing Wire 2-23-2016
TroubleShooting Loconet After a Lightning Hit 5-7-2016
Prototyping Layout Lighting Distribution Blocks 6-1-2016
          Wiring Turnouts 6-9-2016
Installing Tortoise Switch Machines and Tortoise Interfaces on the Prefab Section 6-20-2016
        Testing Installed Atlas Code 55 Turnouts and the Prfab Section Wiring 6-22-2016
Rethinking the Layout Lighting Wiring Scheme 6-24-2016

        Train Room Lighting and Electrical Control Panel 10-31-2016

           Thinking About the Yard 8/25/2015