The internet is such a wealth of information that one can develop the misleading impression that all information is available online. Not but sometimes you turn up something useful. CN Engineering Specifications for Industrial Tracks is one thing I found recently while attempting to answer questions about various track details for Comstock Road. While this is the current CN document intended for modern customers, it does give me help in making more plausible guesses where required.
Some of the information is not of direct interest to modelers since model railway engineering and permitting processes are somewhat less formalized but some of it is bang on. For instance, I gleaned the following useful numbers from this document:
- Minimum ballast shoulder for jointed rail is 6″ beyond the tie ends. 12″ for welded rail. Hmmm.
- Additional sub-ballast width is required on the diverging side of turnouts to provide a working area for the rail crew to stand.
- A spur of over half a mile requires a runaround. Long backing move without a caboose are neither fun nor safe for the guy riding the last car.
- 25′ between main and industrial track unless space is unavailable, 14′ minimum between adjacent industrial tracks.
- 20″ spacing on leads, 22″ on body tracks.
And some we are just going to pretend either didn’t apply in the 1970’s or just isn’t doable in the space I have:
- Minimum rail weight of 115lbs. Depending on sources and rail profile that is either code 125 or code 138. I have Right O’ Way code 100 steel. That is plausible for a circa 1900 era leftover but not especially likely for 1940/50’s new construction. I am claiming reuse as a wartime expedient.
- A minimum radius of 9° of curvature. Which is something like 160″ in 1:48 O scale. The entirety of Comstock Road is 144″ long… We are just going to prohibit autoracks and proceed.
There is also a bunch of details regarding clearances and grades that I will have to look at more closely although model standards will likely suffice. The real GECO spur crosses the actual Comstock Road and ascends to the last customer on what looks like a fierce grade of 5% or more so I doubt I will come up with anything implausible. (My intent is for the back industry siding to be up a grade to give a bit of topological interest.)
Inspired by the improved plan for Comstock Road, I have been diligently relearning the ins and outs of Templot in order to validate the new plan. I have reached the point where I prefer to redo something that isn’t quite right rather than rationalize leaving it be. If only I did this often enough to retain my current skill level. Oh well.
It doesn’t look as grand as the AnyRail version but this is what I intend to build. Of particular note is the fact that one leg of the diamond has a transition curve going through it. It was not too hard to do given the excellent instructional video made by Martin, Templot’s creator but being able to follow along tells me I have figured out where most of the knobs are.
In the past, I have printed out my templates on a stack of 8.5×11 sheets with the household printer. This time I am going to see if a local print shop can do it as one big sheet.
Whilst finishing up the final tie location adjustments in the Templot version of Comstock Road 12.1, a nagging little voice kept telling me that my two facing turnouts with crossing was very close to a single outside slip switch. Which it is. This is a plausibility problem since there is essentially zero chance such expensive specialty trackwork would get used on an industrial spur in Scarborough. The area I am claiming inspiration from went from fields to factories during World War II. There was no shortage of space and thus no need for slips and the like.
So, problems with plan 12.1:
- Almost a slip switch crammed trackwork.
- One of the turnouts is on an obvious curve.
- Runaround turnout spanning baseboard sections.
After a bit of pondering with the urge to start cutting Homasote vying with the doubts, I started doodling some more. My conclusion is to make the following changes:
- Flip the runaround so the main is at the front. This gives me room to move the two turnouts feeding the crossing farther apart.
- Reduce the headroom on the runaround. This gets the turnout off the joint and makes the runaround capacity equal to the traverser.
- Reduce the traverser length to 3 cars from 4. This matches the runaround and more importantly eases the curve into the siding in front of the traverser. That siding needed to move towards the front to allow the second track on the traverser to reach the main.
- Reduce the right front to a single siding. This loses a couple of spots but opens things up a bit.
- Add an “abandoned” turnout running off the front edge for scenic interest. To assuage my feelings of loss at those extra spots, I am allowing for a future extension around the corner just in case. I will either lay the turnout but no rails beyond or just the switch ties and straight rails into the industry.
Anyway, here is the AnyRail version of Comstock 12.2 which I definitely prefer now that I have it. A bit less cluttered and more plausible track geometry.
I took advantage of a quiet holiday Saturday to move Comstock Road along. I had delayed installing levelers on the naive hope that my basement floor was level enough as-is. With the middle baseboard section spanning the two support sections, nothing short of dead level was going to avoid rocking so levelers it is.
Following my local sourcing policy, I bought cabinet levelers such as are used on the bottoms of kitchen cabinets. They are top adjustable and can certainly hold the weight. I just hope that there isn’t a reason the standard item in the UK is different. Especially since I feel that I have invested a fair amount of patience in getting these installed.
When working out my fancy nesting folding leg system, I utterly failed to consider clearances for the levelers. It quickly became apparent that alterations needed to be made if things were going to fit at all. Uncharacteristically, I forged ahead without the usual thinking things to death. By the time I was done, I had: swapped the detachable and fixed ends of the leg braces, swapped them back, swapped them again, moved the cross brace attached to the brace supports to clear one set of levelers, set the works up, realized I forgot to put all the screws in, take it down, put it back up, realized it was wrong side to the wall, take it all down and set it back up the right way round. I may have said a few uncomplimentary things about myself in the process.
I pressed on and the baseboard frames are set up in place ready to receive interior bracing and trackbed. Here is what the levelers look like (That spruce 2×2 leg goes at the back but I knew I should have just bough one more finished pine stick):
And here is the current state of Comstock Road just before I roll the two storage carts back under it:
As I previously mentioned, a major cleanup and reorganization are required to site the new Comstock Road layout against the wall. A primary task was removing and disposing of the last section of the old layout benchwork. Here is a shot of the bottom which is the interesting side.
This section never had anything on it that was supposed to be there and has been serving as a big, very solid shelf since it was built. (It was going to be harbor water and piers.) I cleared it off this morning, carted it out to the driveway and cut it up so that it could join other detritus in a trip to the waste transfer station.
Always one to try a new benchwork construction method, I built the previous layout using the “waffle” approach. The results are rigid to an extreme. Tapdancing elephants might crush it but it probably still wouldn’t flex. Construction time is probably on the same order as the plywood sandwich I am using on Comstock Road but with more sawing with a jig saw required.
But, you knew there must be a but, the rigidity was at the price of weight. 1/2″ plywood + Homasote over all made the sections firmly in the movable not portable category. There are clubs who base show layouts on waffles so portability standards obviously differ. I suspect such clubs are well provided with extra hands and hauling capacity.
I made a push this morning to get the planned traverser motor control system assembled. This involved soldering screw terminal blocks onto the Big Easy stepper driver board, tinning the motor leads, soldering leads to the barrel jack for the power supply, affixing the components to piece of wood and wiring it all up.
Wiring connections were made as per the SparkFun hookup guide for the Big Easy. Clear instructions for the layperson are not to be sniffed at. This all went appeared to go well but one is never sure before you turn on the power.
A plank of wood as a way to hold things is a quick way to keep things from getting tangled or worse. The motor is quite heavy and could break toe if it landed on it. It would also rip out the carefully traced wiring as it went.
The hookup guide includes matching source code and tonight I used that as the basis for my test firmware. The pin reset function in the example was MIA so I had to write my own with reference to the default settings for the six control pins.
Finally ready for the big test, I plugged things in. LED’s on the driver board lit up and no magic smoke release was observed so on we went and things actually worked on the first try. As veteran developers will tell you, this is unexpected and somewhat unnerving. A tribute to the quality of the documentation, I think.
I took a video but have yet to devise a delivery mechanism for it. Meanwhile, here is a promotional still.
I have acquired the necessary electronic components to implement my plan for automating the traverser. Working with stepper motors is a new thing for me so I wanted to start in on that sooner to qualify the technical risk aka figure out if it can be done. In the interests of speed, I decided to go with fast and effective for the prototype.
Initial major components are:
- A NEMA23 stepper motor which the size typically used by small CNC machines. The traverser table will be rolling on drawer slides and hopefully this motor will have enough torque to move the table when loaded with cars and loco.
- A Big Easy Driver board to control the motor. This motor can draw up to 2 amps so you can’t drive it directly from microcontroller PWM outputs like a small servo. This driver is one that has good Arduino support.
- An Arduino Uno R3 microcontroller board. This is major overkill for this application in terms of processor power and functionality but the quickest to get something running. Final version will probably be an Arduino mini but that board doesn’t have on-board debugging support so I wanted to defer that until later.
I am going to use the Arduino development environment with its associated cornucopia of software and community support. I would not base a commercial product on Arduino but nobody is paying me for this so I am trying to focus on the model railroad goal and avoiding the temptation to start from scratch firmware-wise.
In preparation for cutting out roadbed and sub-roadbed, I have been creating the detailed track templates. General track planning software such as AnyRail which I use is good for general planning but does not produce templates for hand laying track. For that, the go-to tool is Templot by Martin Wynne.
Templot is free to use although I bought it when it was a licensed product. It is a powerful specialized CAD program specifically for producing custom trackwork templates according to prototype practice. If all of your turnouts are a standard, regular size then something like a FastTracks jig may be all you need. If, however, you need to produce more challenging track formations such as a yard ladder or a crossover on a curve, Templot can’t be beat.
That’s the good part. Not so good is that Templot, like other serious CAD programs, has a steep learning curve. It takes time to get up to speed which is an additional hurdle if you, like me, only need to produce templates every once in a great while. All of the terminology is based on UK prototype practice which also takes a bit of learning.
That being said, I have blown most of the dust off of my modest Templotting skills and am mostly done the tough bit of the Comstock Road plan. Here is a screenshot of some of the work in progress.
The lower right turnout is not a model railroad-esque right hand turnout with what should be the normal route going through the curved leg, it is a #6 left hand turnout diverging from a right hand transition curve. And incidentally going straight into an asymmetric crossing. Once I had got back up to speed with the Templot interface and watch a couple of subject specific videos, the actual template creating didn’t take that long.
Still to do are:
- spurs out from the crossing
- assorted tweaking of ties such as those overlapping and/or too widely spaced around the crossing
- check turnout motor locations against module boundaries.
Once everything is as good as I can make it, I will print out the whole thing full size on my home laser printer and there will then ensue a lot of trimming margins and taping sheets together. The whole thing then gets used as a template to cut out the Homasote roadbed and off we go.
Tonight I finished gluing up the baseboard frames. I decided to do a test setup just to make sure everything fit before the glue dried. There was a bad moment where I was looking at two section ends both labelled 1 and top but had pins on the same side. After a little bit of panicking and preparing to rip and end out I flipped the end section over so the pins and sockets aligned and…. it all fit. Which leaves me pondering where I mislabelled an end or accidentally made things symmetrical enough to work upside down. Almost certainly the former. Still, I don’t like it when software bugs magically go away without explanation either. Oh well.
Here, crammed into the aisle in the shop, is the current state of the layout. Internal cross bracing will wait on getting the trackplan laid out full size so that turnout linkages can be avoided. Next up is a big tidy/declutter/rearrangement so that the layout can go up against the wall to the left and a start on the backdrop ends and valance.