Stew's Garden Scale Railroad Empire

Our Model Railroad Empire

When I can have fun with this?

Most of what I've described previously took place between June of 2016 and November of 2017, since we began to remodel our kitchen around Thanksgiving time of 2017, and only recently completed it in early June of this year, 2018. I say completed, but really there are a few odds and ends that need done, like the door to the side porch. That's one of those things that's waiting on something else that needs done first. In this case, getting the garage back into some semblence of order after all of this remodelling. We're in the rains every day season now, and I can't leave the door out in the weather, so I need space it the garage for it... But that's not what I want to talk about.

(8 July 2018)
We've accomplished little else this year and it's taken nearly every weekend since Christmas just to get it all done. As with everything else, you can read about that here too. And while I still need to replaster the hallway where we closed off the doorway from the kitchen, the kitchen is done and back in operation. I'm hoping to get back to working on the railroad soon, but between having to spend every weekend on maintenance at the other house and the disaster left in the garage from all the remodelling, it seems there's always something else that needs done first. But that's my middle name I guess...

I did manage to get a few of the file cabinets as well as one of the wood cabinets I built from the other house and set up a "work cell" in my office awhile back, in preparation for some intricate remotoring and super detailing work I have planned for the the flea market finds. I previously acquired the remotor parts the last time Bachmann had them on special and I've recently acquired detailing parts during their latest sale. Again, I just need to clear out all the remaining chaos from the kitchen remodel to make room for it all. Sitting here writing this gets me no closer to that goal, but it does help me think about where we are, where we came from, and where we're going with this railroad of ours.

While the future is uncertain, we know we need to do something about the only place left with bare dirt in the backyard, inside the change of plan loop. I've been researching various structures that would be a good fit to sit alongside the long stretch of straight track between the garage and the loop. Perhaps the Village Tavern? I've long wanted to "decorate" the 4x4s that frame tha raised bed planter to make it look more like a train station, or other such massive building, that would be right at home on the railroad. When we first put it in, I left about a foot between the side of it and the patio, for two reasons. First was so the 4x4s didn't sit at the low point of the drainage along the low edge of the patio slope. Second was to leave room for gravel and a siding to that station, once the 4x4s are properly decorated.

By placing a roof or similar type of structure to overhang the top edge of the 4x4s, and a platform along the side, it would give the appearance of a protected passenger platform. Shallow relief panels with doors and windows would further enhance the illusion of a passenger station platform. In the center, where 4x4s cross between the two long sides, a turret or other structure such as a clock tower would disguise the fact that it's just 4x4s, giving the illusion of one massive structure. The fact that roses are growing out of what would be the courtyard would be the only giveaway that it wasn't really a building, but a planter. That's the plan anyway.

I'm still trying to find the plans for the proper architecture. I'll probably build a template to help assemble the roof trusses and use some of the simulated corrigated roofing panels I've been making from annealed aluminum cans and a heavy duty "crimper". I guess I need to start some sidebars of the various tools, templates, and materials I use to make these models. All of the wood I've used to model structures, like ties and trestle bents and such, are old fence panels repurposed into scale lumber. Since I'm using the Bachmann "Big Haulers", which are large and meant to represent narrow gauge equipment rather than standard gauge, you could say what I'm modelling is closer to F scale than G scale - Meaning 1:20.3 rather than 1:29 or 1:32 proportions.

I tend to use 1:24, or "half inch" or half doll house scale, since it makes the math a little easier. More like it makes the scale lumber a standard fractional size. For example, a scale 6x12 is ¼"x½", the size of a testle sill member. Cross bracing for trestles is generally 3x10 or 3x12, making it 1⁄8"x3⁄16" or 1⁄8"x¼". And 1⁄8" is about as small as I feel comfortable ripping on the table saw. I make a push shoe from ¾" plywood and keep it just for doing 1⁄8" - ¼" sized material. Another is dedicated to the ¼" - ½" range. Much bigger than that and I can just use the standard push stick that came with the table saw.

Much smaller than 1⁄8" and I'll need one of those Northwest Shortline "chopper thingies" to slice the pieces using razor blades and such. I think they may actually make a mini table saw of sorts for ripping stock in those smaller sizes, like scale 2z4s and the like. The alternative is to buy the expensive precut scale lumber. For now I'm just sticking with the sizes I can cut on the table saw. Thankfully most everything about the railroad is massive already. And since I build to a scale slightly smaller than the equipment I'm running, it will look more rickety. On the flip side, if I ever decide to run standard gauge equipment, it's going to look MASSIVE compared to the trains.

I've yet to experiment with metal framed structures and bridges and such. The closest would be my experimentation with the aluminum cans. I've seen too many commercially available structures that aren't even a poor representation of the prototype. Oversized screws or pop rivets or even tack welds holding them together. Guess I got spoiled by all of the detail in styrene and even etched metal laced girders available for HO scale. I'm sure there has to be a much better way to represent the fasteners in scale, starting with the gusset and fishplates and such, so obviosusly absent from those sad models. Bridgewerks comes the closest, if you can afford them, but they still use unrealistically large pop rivets to fasten everything together.

I guess you'd call me a rivet counter. It doesn't have to be exactly to scale, or even entirely faithful to the prototype, but a pop rivet is a pop rivet. It looks bad full size to begin with, and would be bigger than a car wheel in scale. Even the #2 brass hardware I use for timber bridge fasteners is oversized in scale, but small enough to be believable. I suppose for metal structures such as gusset plates, rivets could be simulated closer to scale using the "pounce wheel" technique, then using solder or adhesives to connect them together. I prefer something close enough to be believable, yet still uses fasteners to allow disassembly for whatever reason.

Although I have yet to add a single siding or crossing at grade, I want to add semaphore signals for visual interest. I had started to put together the tools and materials I needed to make PRR position light signals from brass in HO scale. Then I had to tear that all apart just for it to sit in pieces in storage until I can find a space to once again occupy with my HO scale empire. But we're talking about our garden scale empire and semaphore signals. Using larger brass tubing and turning parts from brass stock and shearing targets from sheet brass, I can actually make them operational. Most likely they would be soldered together, similar to the way the prototypes are welded together. I haven't quite figured out if 00-90 sized nuts and hardware would be a good scale representation of the U-bolts and such though... Yet.

The other thing that I really need to build is larger scale interlocking tower to house all the controls. Just the electronics at first, and at this point, it would just be a set of 70VA powerpacks. The entire layout is just one big block at the moment, so eventually I'll need to divide it into smaller blocks, insulating them from one another and running the power feeds to each of them. This is where converting to DCC would help reduce the wiring complexity. For some reason, DCC is still quite expensive for large scale, perhaps because the numbers aren't there for greater economies of scale. Dunno. Wire and switches are still much checper than even the least expensive DCC contoller, not to mention all the trouble you have to go to with installing decoders and all, so it's little wonder many folks still prefer low cost DC cabs to expensive DCC setups.

What I'd like to do eventually is control each block with its own motor controller electronics and an "intelligent" control system to track each train and its progress, switching control of each block to the cab controlling that train, be it DC or DCC. This would require occupancy detection as well as some sort of ID and feedback. That's where I need a new idea... DCC has expanded to transpond, and could probably identify itself when queried by the controller. The block that received the response would provide the whereabouts of at least one engine of the train. But where is the rest of the train? To the east or to the west of that engine?

Proper occupancy detection for the entire train would require "leakage" resistors to draw a small amount of current to allow the motor controller or other circuitry to detect it. I see this as a work in progress type of project, where features are added as they become available. The "smarts" for the intelligent controller could come from an embedded controller, like a PIC or similar, all the way up to a raspberry pi connected to a flat panel display. The possibilites are endless. For now I would be happy with a tower that had a flip open top and a power pack dedicated and connected to the railroad where I could run trains at the flip of a switch. What would be killer is interlocking lever frames with operational mechanical interlocking and actuation of remote devices, like turnouts and semaphores. Someday...

So now I'm thinking about how the guys in the smaller scales epoxy those "chip" resistors across the metal wheelsets to provide a current path for sensing and detection. I've also seen some modellers actually add RFID to their rolling stock to track their wherabouts. Maybe the two could be coupled together somehow, like a transponder the size of one of those chip resistors. The problem would be how to program them to be different "addresses", to have a different id in other words. Many embedded devices these days have a sort of serial computer interface, such as "I2C" or the like. It would probably require a socket of sorts to place it in just to program it, and I'm not sure how you would reprogram it once epoxied in place, but I certainly like the idea of each car have its own id and the ability to transpond its location. Again, someday...

Another thing I really wanted to setup on this page is a list of plants with links to more information on them. We need to do something to disguise the fact that the railroad is just a bunch of stringers on a bunch of plastic poles. Originally, I thought of replacing those poles with concrete blocks and building some earthworks around them, both to disguise the blocks and provide planters for trees or whatnot growing along the right of way. Part of this should explore the planter inside the change of plan loop and perhaps adding an engine house or roundhouse to it. Perhaps the water feature should be part of it. Needs more dreaming...

So two things are screaming to be done: 1) Make a list of things to do; 2) Make this more than one page, like current and future. Perhaps many pages, one per "project" I'd like to track... Puns. What a concept. Some need done sooner than later and some sooner than others. I'll give the list a shot next. This future page I like to use as a "what if"? Just to capture the ideas as they pop into my head.

(20 July 2018 - That's one small step for man...)
I sit here updating this page, pondering the significance of today's date. Nearly 50 years ago today, men first walked on the moon. I was 7 years old then, and now I'm going on 57, wondering if today's kids even understand the magnitude of that accomplishment. Hell, I wonder if folks from my generation even understand the magnitude of that accomplishment, chanting for "Space Force"... But that's a different can of worms. Today's challenge, for me at least, is how to make more progress toward additions to the backyard garden rail. To that end, I have been doing some of the research tasks listed below. Initial investigations anyway.

In researching block occupancy and transponding, I have found a number of techniques that measure the current draw, all the "same old same old" techniques - Twin-T, reversed diodes, inductive sensing, optical - all "tried and true", all with many sources of manufacture. What I did find intriguing was a new addition to DCC transponding, a means of identifying what block a train is located in when it transponds its response to a DCC command. It still needs block detection, but a new "receiver" is added to the mix that decodes the transponded response and identifies the block that response came from. Clever. But how does it know where the end of the train is? A light decoder with transponding built into the caboose? What if your pike doesn't run a caboose anymore, like the modern prototype? A block is a rather large chunk of "rail estate" too.

I always wanted to figure out how to use time domain reflectometry (TDR) to actually discern how far from the ends of the block the largest current sink resides, which should be the engine(s), then command an external surround sound system to "track" the source. Turns out there is a "SurroundTrak" system that integrates with this transponding system extension for Digitrax, but I have to wonder just how good the resolution of the sound source position is? Plus it appears that only Digitrax offers this capability. They already have the "loconet" mechanism for exchanging information between their different pieces parts, so you could likely determine the direction the train was heading, but again, where is the other end?

I also see there is a complete end-to-end control system that allows for total automation of your pike, using your existing infrastructure, even providing their own hardware and software. The starter kit is $99... Wow. Another idea of mine that's come to fruition. Well, sort of. None of them give a granularity any better than a current sink is located somewhere in this block. My idea needs much more research, with trains running and an oscilloscope connected to the rails to monitor the waveforms, and how they change based on position within the block. With DC cabs, the system would have to rely on the "noise" generated by the non-constant impedance between the wheels and the rails. With DCC, there is already a "natural" 8KHz frequency signal generated by the controller and applied to the rails either directly or via a booster.

Perhaps instead of a "chip" resistor, a sort of resonant circuit could be constructed as a chip and attached like those chip resistors? Don't think we'd want a resonance at the DCC frequency, but perhaps a higher order harmonic? I'll need to do the legwork on wavelength for those frequencies and see how they compare to block lengths in the various scales. Maybe even have the cab send out a "ping" at that frequency and see what resonates back, and where.

An alternative would be a GPS in both engine and caboose that relayed position back via WiFi, more of an IoT approach perhaps? With the constantly dropping cost of both GPS receivers and embedded microcontrollers with built in WiFi, this could be a less expensive alternative, especially if coupled together with remote control. It seems it would still be much cheaper than the radio control alternatives. Sound seems to add quite a bit more expense, but if the embedded micro also has an audio amplifier, just add a speaker. Now the motor controller needed per block becomes a motor controller per loco, and at least a lighting controller per caboose if looking for that ultimate "where's my train?" control.

(Sunday, 30 September 2018)
But that has nothing to do with what I got done this last weekend on the backyard railroad! I even ran trains this weekend! That's right, plural. Trains... I wanted to test the re-re-fit of the new base in place of my less than successful attempt at modifying the original to accept the new mechanism. It runs great, except now the front truck wants to climb the rails and derail. After trying to re-organize the garage, I had just the one flea market find engine left in the box that used to contain all three. The flea market find that's like Ann's new engine chugs along, but seems to be having some issues. Good thing I have a new mechanism for it too! I should have worked on the taxes, but it was impossible for me to resist the urge.

I guess I need to start some sidebars of the various tools, templates, and materials I use to make these models. All of the wood I've used to model structures, like ties and trestle bents and such, are old fence panels repurposed into scale lumber. Since I'm using the Bachmann "Big Haulers", which are large and meant to represent narrow gauge equipment rather than standard gauge, you could say what I'm modelling is closer to F scale than G scale - Meaning 1:20.3 rather than 1:29 or 1:32 proportions.

TDR Pulse Width Dead Zones

Pulse Width	Dead Zone
1 ns One billionth	1 foot
2 ns	3 feet
10 ns	14 feet
100 ns	55 feet
1 µs One millionth	430 feet
2 µs	810 feet
4 µs	1,600 feet