Traverser Control Dev Board

Here is my latest development(dev) board for the traverser control system for Comstock Road.

Prototype embedded systems are often embodied as a development(dev) board. This is, electronically, a complete system laid out on a printed circuit board instead of in whatever form factor the final product is expected to be in. The dev board is used by embedded software developers to create the firmware. Laying things out on a board makes it much easier to attach measuring instruments such as multimeters, oscilloscopes and logic analyzers to see what is actually happening in parts of the system.

Unlike the industry standard, my dev board is a literal pine plank with components attached to it but it serves the same purpose. All the parts are there and wired up, just not installed in the layout. Given how chilly the basement is right now and how far it is from my computer, this is a non-trivial advantage! Attaching everything to a board also allows me to move it around easily with reduced risk of damage.

Clockwise from the bottom right, the components are:

  • NEMA 23 stepper motor
  • Big Easy motor driver board with a barrel jack for power
  • 4×20 LCD display board
  • 2 limit switches
  • Arduino Uno with a screw terminal shield on top of it
  • Breadboard (so far just used as a ground bus)
  • Front panel with pushbuttons I built

Wire management could be better but it is clear enough to me in person. Now I just have to create the actual code.

Beachhead (Re-)Established

I am working on the Arduino based drive system for the traverser on Comstock Road and have arrived at the point where I need to solder leads onto buttons so I can hook them up to the mocked up system. I have also been putting that off for a while now because my workbench has been well buried in detritus. It is the place I tend to unpack things and put things that I am not putting away.

This afternoon I took a bit of time and did some tidying up. I put away things I have places for and cleaned up the general accumulation of stripwood ends, dropped straight pins, jumper wires, … Surprisingly, there was relatively little that doesn’t have a place although bottles of fluids it definitely a problem area as indicated by the upper left corner of the photo.

I got the lead soldering project started and am now set up to leap back into action with no startup costs. I have also identified a necessary improvement in the form of some sort of small bottle stabilization scheme to prevent upsetting the narrow bottle of flux I was using. Or, historically, any bottle of flux I am using. At least nothing else was in the splash area so my virtuous cleaning had immediate benefits.

Goals for a New Year

Lookit here, third post in as many days. Christmas sugar rush still in effect. I have a few things I would like to get done and putting it out there in a post is one way of adding a nudge towards action. I am blessed with the required resources of time and money but struggle with inertia, motivation, getting started, whatever you want to call it. I am also still alive so there is still hope for improvement. 🙂

At the risk of upsetting anyone who has seen too many corporate management fads go by, I am defining goals that fit my version of SMART(specific, measurable, achievable, realistic, and time bound). The specific is the important one to me, improve shop organization is vague, create a set of drawers for the lathe stand is specific. Everything can be done with infinite time so some target deadline helps me but I am not going to be too stressed about it. I am going to aim to get it all done by the end of May.

Without further ado, here are the ones I want to start with:

  • Machining – Complete Beam Engine: already ongoing
  • Shop Organization – Lathe Tool Rack: started on some widgets, need the thing to hang them on.
  • Shop Organization- Organize Measuring Tool Drawer: Kaizen foam in hand, need to work out a layout
  • Modelling – complete On30 railbus kit: this one is a bit ambitious but I think the machine shop gives me a way to fix my blocking issue
  • Comstock Road – Traverser Automation – I have already applied the paddles to this long dormant project, still working on rebuilding my development environment. With this done, I can complete the backdrop and mock all the structures.
  • Comstock Road – Rough Scenery – foam is half done. I want to finish that and get everything covered in basic ground cover.
  • Blog Regularly – I will aim for once a week at minimum. It may be me talking about why I haven’t done anything since last week…

That is all I am going to put on the list but of course there are many other projects lying about. Hopefully, this

Servos Wired

Over the last few days, I have worked my way along to completing the servo installations for Comstock Road. There was a bunch of other wiring tasks required that got done while I procrastinated on the servos themselves. When I actually got around to them, things went reasonably well although I did have to get out the multimeter at one point to debug what turned out to be an insufficiently inserted Anderson connector lug.

More or less in order, I did the following:

  • installed the board that holds the Octopus board and relays
  • connected the control panel to above. (OK, order did matter here)
  • ran wires from the relays to frog feeders
  • ran a servo cable from the controller to the runaround turnout which is the only servo not on the center baseboard section
  • applied Anderson Powerpole connectors to the runaround servo cable to span the gap between baseboards and jumper connectors to both ends
  • ran a cable and jumpered the other servo run which is not right next to the control board
  • bent up the link wires and installed them through 1/16th” brass tubes
  • actually installed the servos
  • tested everything, (multimeter comes in here)

The next step will be to make up the throw bars and link them up so I can align the servos and run some trains.

Here is the control board wired up. I used some small wire wraps to tidy things up a bit. Without the wraps, the y-cables connecting control panel, relay and controller made a bit of a rats nest.wiredcontrolboard.jpg

At the other end of the center section, I created a “cable” by wrapping all the wires going into the Anderson plugs with electrical tape. This makes it less likely that the smaller gauge wires will get damaged when the connection is made and unmade.wiringharness.jpg

Note that only the white and grey DCC bus wires look like they need these robust connectors. In fact, I had to solder the servo and frog wire to the lugs before crimping them since the crimp isn’t tight enough to grab the small gauge wires.

Octopus On A Plank

I have some task to perform this weekend that will take me out of town on Sunday so hobby time is limited but I did get something done. Permanently installing the Tam Valley Octopus III and attendant remote relays will require me to mount them in close proximity to the control panel. There isn’t any large wood surface in that location or in any other for that matter, this being the downside of non-solid wood topped baseboards. There being no wood surface available, I determined to provide my own.

I measured the boards and pondered their optimal orientation in relation to each other and to the various connections that would need to be made. I then cut a 3″x 8″ hardboard rectangle, drew some guide lines on it and attached the boards with 3/8″ #4 wood screws.

It isn’t fancy but all the boards are securely mounted so that plugging and unplugging won’t cause something to come loose. I will screw the board to the underside of two subroadbed sections so that the board is accessible between the baseboard bracing.


The blue terminal blocks all handle power of some sort: track for the relays(red boards) and servos for the Octopus board. The black connectors on the relays and the C 0-7 header on the Octopus connect to the controllers on the panel via Y-cables. The less ideal bit is that the cables to the servos connect on the header on the left side of the Ocotpus. This was the least bad choice since the connector for the Remote Alignment board (right side) needed to be easy to get at since it’s cable will get repeatedly plugged and unplugged at least until I get things all set up.

As a note, the relays are what seemed like the best solution for switching frog polarity when I bought the Ocotopus-III several years ago. If I was doing this from scratch, I would use a Frog Juicer instead especially now that I have used one and seen how simple it is to wire up.

Diamond Frog Juice

I have been test running back and forth looking for mechanical flaws in Comstock Road’s trackwork. Whilst so engaged, I discovered that the two isolated and as yet unpowered frogs on the diamond are both as long as the SW-8’s trucks and spaced almost exactly the same distance apart. This was obvious as the locomotive persistently lost power crossing the diamond at slow speeds. (Faster speeds allowed the capacitor to get it across the gap in time).

Fortunately, I recently purchased a Tam Valley Dual Frog Juicer for just this eventuality and got it installed tonight. Several post-installation test runs have failed to produce a stall so it looks like the device is earning its’ keep.diamondfrogjuice

This is probably the easiest installation I will get. Four screws, four wires and done. I think I spent more time reading the instructions than actually hooking it up.

As an encore, I broke out the servo connector wire kit I have assembled from various vendors. The combination of purpose specific three strand wire, connector kit and crimpers will allow me to connect servos to Octopus-III board and control panel. And the vast quantity of parts will allow me to mess up the occasional crimp. Score is 6-1 so far…jrkit

Electric Sunday

Today’s model railroading efforts were devoted to wiring. I connected up the feeders for the balance of the layout. I then tested the main and one path through the diamond and am happy to report that all went well except for the occasional loss of power due to dirty rails. Rail cleaning ensued. Since rail cleaning improves electrical contact I am calling that electric related.

I then drilled the holes for the three remaining turnout servo linkages. At that point, I realized that I was going to want a control panel since the minute you have two unlabeled push buttons dangling off an unattached board you have a user interface problem.

I then assembled four more Tam Valley fascia controller kits (one more than required so I have a loose one for servo installation if I need it). tvdcontrollers

I am unsure of what form the final control panel will take so I am not investing a lot in a panel right now. I therefore bodged up a quickie panel out of hardboard using basic drafting tools and the paper template supplied with the controllers. Ten minutes with a center punch and the drill press gave me all the holes required. A bit of sophistamicated graphic artistry with a straight edge and the big marker and the temporary stand-in was done.bodgepanel

This will work for me as a start and allow me to experiment with panel locations. This is my current best choice but I am not thrilled with the proximity to the creek bed to the right.

Power Strip

I made a minor investment in Comstock Road’s infrastructure by installing a dedicated power strip one the sub-baseboard frame. This will allow me to plug in the “wall wart” power supplies for the servo controller, the traverser mechanism, and whatever else I end up with (animation, structure lighting,…) and run one cord to the dedicated switch outlet. Prone as I am to getting focused on a task and obliviously kicking and tripping over things, I felt it prudent to organize things out of the way before I accidentally ripped some cord out of something.

This also represents the first time I have actually been under the layout since I set it up since I can usually just flip up a baseboard section to get at the underside. A sweep up of floor usually covered by stuff was indicated. No lost treasure was found.powerstrip

The scrounged power supply currently powering the servo controller is obstructing an outlet despite the spacing. I may have to see a squarer replacement if I need that outlet in the future.

Turnout Control Conclusion

If you have been following along, you know that I have been attempting to devise a way to throw Comstock Road’s turnout points using a servo and a rotary motion mimicking a manual switch stand. The initial attempt using the mounting scheme appropriate to the typical back and forth scheme was not a success. After much scheming, I became resigned to having to mount the servo face towards the baseboard bottom and with shaft in line with the vertical throw shaft.

Happily, I came across a similar scheme used and well documented by the Delmarva Model Railroad Club that I could adapt to meet my goal. Rather than use a couple of blocks of wood, I used a couple of pieces of aquarium bubbler hose and 1 1/2″ #6 wood screws.

Before I tipped the center section up to get at the servo location, I taped down the points, throw bar and all. This kept things centered as well as prevented the pins holding the throw bar from falling out. Family lore includes the time we tilted a sofa bed while lugging it up the stairs and it went sproing. All subsequent movements start with tying those suckers shut! Note that digital photography has not stopped me from exercising my talent for getting a finger into the shot…boughtthatfarm

Things secured and disconnected, I tipped up the section, clamped it in place and re-bent the vertical wire to the new spec. The horizontal leg has to match the distance between the servo shaft center and the last hole on the servo horn.turnoutwiremk2

I plotted out the mounting holes to put the servo horn perpendicular to axis of the servo mount at center. This turned out to work but only just. The servo shaft is not centered in the housing so the near mounting screw interferes much sooner than the far one. The interference issue is only relevant when you invert the servo like this. Future installation will offset the center point to split the difference in the available travel.servobracketmk2

I tried to capture the situation when the turnout is thrown to that side. The servo horn is right up against the tubing but the turnout is thrown so we will call that a win.servohardover

Finally, I am awaiting the arrival of appropriate bits and bobs to wire the servos, controllers and driver board permanently. I can still operate one turnout at a time at close range via temporary measures. Also note that I forgot to install the frog polarity relay while I was “under’ the layout. One more for the checklist.verytemporary

I did a run of the test train to prove things worked so I can now claim to have an operating layout. I can now perform an Inglenook scheme via this turnout, the back track and manually pushing the traverser. Or at least I could if I had enough cars converted to P:48. I will need to do an inventory and get that under way.

Getting one turnout is not a huge deal but getting a working method sorted out to my satisfaction is a mental obstacle overcome. Onward!

A Bit More Turnout Control Progress


Pictured above are in order, an Octopus III servo controller, remote relay, fascia controller and micro-servo, all from Tam Valley Depot. I bought a bunch of each for the previous layout and never did get any of it deployed so I am both pleased to be finally using it all and having to learn how to do that.

Tonight I soldered up a fascia controller kit (two LED’s, a button and a connector) and messed around with the remote alignment board to get the hang of it. I think I can do it all now including, ahem, factory reset the Octopus in case I mess it up. Hypothetically speaking. 🙂

Next concern is that the required throw for the points is about 100 degrees of rotation. To get that out of the servo will require it to be quite close to the point of rotation which creates other alignment issues. I am having a bit of a ponder about what to do about that. I will also tip the board up for the next bit of fiddling since I can’t get under the bracket location with a screwdriver due to the sub-baseboard. Hopefully the assorted point bits won’t fall off since I haven’t permanently attached any of it.