Upgrading an old model.

I don’t think there’s any question that I’m a better modeler today than I was a decade ago. That’s probably true for anyone trying to be their best at something. Our “best” is a constantly moving target that changes with every new skill mastered and bit of knowledge gained. So, with that in mind, I decided that it was time to take a second look at my model of Seattle & North Coast F7A #101, which I began way back in 2007 and didn’t complete until 2014. This model has always had a few annoying issues caused both by inexperience and not having the proper tools and I figured that at this point I was ready to make a second attempt.

The primary impetus for working on this model came when I decided to move away from P:87. As originally built 101 was fitted with code 64 wheels and couldn’t run on anything other than perfectly laid exact scale track. Additionally there was a considerable amount of noise from the gearbox when I put it on rollers to test the motor and I figured a good lubrication, or perhaps even a motor replacement, was in order.

I won’t bore you with photos of the wheel swap as I’ve shown similar work before. However, in the process of disassembling the model I was reminded that the coupler boxes were a shoddy hack job and the electrical system a pile of wire spaghetti. Taking 101’s shell off and putting it back on has never been a lot of fun and since I had everything apart now seemed like the right time to tackle those problems.

I elected to start with the coupler boxes. The main issue was that the Sergent couplers I use on my models have a much thicker shank than the Kadee clones that the Athearn boxes were designed for. Additionally, the Details West GN style plow on 101 sticks out farther than the standard F unit pilot necessitating a long shank coupler (the prototype plow folds to either side at the center to allow better clearance but the DW casting obviously does not). Because of this the front coupler stuck out too far to look realistic when the locomotive was leading a train and still wasn’t quite long enough to operate properly. It also had something of a, to put it delicately, flaccidity to it. The situation at the rear wasn’t any better. The DW draft gear had a narrower opening than the stock Horizon/Athearn draft gear and I ended up cutting a huge chunk off the front of the box. I did thin the shanks on the couplers down a bit to deal with the thickness issue but I’d also had to build new lids for the boxes out of styrene and really hadn’t done a good job. I believe this was partly the cause of the drooping front coupler.

I figured that for the rear coupler I just needed to make a box that was the right shape for the draft gear and tall enough to accommodate an unmolested Sergent. For the front I’d always thought that a box with an oval shaped aperture that would allow it to slide back and forth on the mounting screw might provide both the right look when the unit is leading and the necessary length when it’s coupled to something else. The tricky part was going to be fabricating them. I made a couple of mock-ups using an Accurail scale width coupler box and some styrene but they were both clumsy and difficult to assemble. Since I have two more SNCT F7’s to build this seemed like the perfect first project for the SLA 3D printer I bought last year.

To design the parts I used Autodesk’s Fusion 360 which is free for home/hobbyist use. There’s a bit of a learning curve but it’s not too bad. I watched a few tutorials and played around for a couple of hours and soon knew enough to make a stab at the coupler boxes.

I won’t go into too much detail here about designing parts for 3D printing. There are a ton of YouTube video’s and blog posts on the subject by people who actually know what they’re talking about. My process was pretty simple anyway. I took measurements of the model, drew a sketch on paper and then drew the same thing in Fusion 360. Here’s what the models looked like on the computer:

With the design complete, I exported the models to a program called a “slicer” which translates a 3D model into a series of layers that the printer can then print. An SLA printer uses an LCD screen to project an image of each layer underneath a vat of resin. A UV light then cures whatever resin is visible through the image on the LCD. The cured resin sticks to a build plate that moves up as the print advances. The slicer tells the printer how thick each layer should be and how long the resin should be cured for. I tried both the slicer that came with my printer and a free alternative called “Chitubox” most people seem to prefer Chitubox though for my purposes I didn’t see much of a difference.

With slicing complete and the resultant file ready to go I Ioaded it onto a flash drive and fired up the printer. I have an Anycubic Photon, which is an entry level machine that’s been around for awhile. I bought it on sale and while it’s worked pretty well, there are now more advanced SLA printers available with higher resolutions and/or the ability to print longer (or taller) models. If you don’t know what I mean by “SLA” check out this link for a nice overview and a comparison with the more widely known FDM printers).

Each print took a couple of hours and the resin is pretty noxious. Whenever I run the printer I set it up in my spray booth and leave the fan running for the duration. I generally also let the printer run overnight or when I’m elsewhere in the house. The smell isn’t overpowering but it’s annoying and probably isn’t particularly good for one’s health. The build area on the printer is large enough that I was able to print multiple coupler boxes at once. This was handy because it allowed me to position each one at a different angle to see which was best. I ran a number of prints, making adjustments and re-slicing the model between each one until I got a few models that I could use.

Once each print had finished I used Anycubic’s Wash and Cure station to remove any uncured resin and then give the part a final cure under UV. After that I removed the print supports from the parts, gave them a quick sand and primed them with Mr. Surfacer 5000.

With the parts primed I was ready to install them. The front coupler box is assembled outside of the model with a small screw that I salvaged from an old CD-ROM player holding the lid on over the coupler. It then slides through the opening in the plow, and a printed bushing is screwed in to the hole on the frame. The coupler box can now slide on this bushing. It’s designed to be tight enough that if the engineer is careful with his throttle (as he always should be) it will not retract when coupling or pushing a train. The only thing I may need to revisit is the screw. The bushing can turn a bit as the box is slid back and forth and it may work the screw out over time. If that happens I’ll likely secure it with a bit of Loctite thread locker.

The rear coupler box was considerably more difficult to install. Because of it’s shape the box needs to be worked in between the rear truck and the draft gear before the coupler is inserted through the hole in the pilot. Then the lid can be angled in and the entire assembly can be lined up with the hole in the frame to be screwed into place. Unfortunately I discovered that the box with the lid on it was slightly too thick to clear the truck. Since I couldn’t make the box any thinner due to the thickness of the Sergent shank I instead opted to cut the rearmost part of the gearbox off. This doesn’t affect the gear train though it may allow dirt and contaminates in. I’ll keep an eye on it but it’ll probably be fine.

With everything put back together I did a quick test. As you can see the front coupler can now be pulled out well beyond the point of the plow but can be retracted into a more visibly appealing position when not in use.

With the coupler boxes sorted it was time to move on to the electronic rats nest. When I originally finished the model I drew up a beautiful wiring diagram in Illustrator and then just soldered a bunch of wires and connectors together. It worked and it could come apart without de-soldering but one wire broke over time and others were rubbing on the flywheel and were likely to lose structural integrity at some point. Those white connectors were nearly impossible to pull apart and they often came undone only after pulling with an unreasonable amount of force, making them more dangerous to the fragile shell than de-soldering would.

With the shell removed, I ripped everything out (including the motor) and considered what I ought to keep. The speaker was ok for the early aughts but doesn’t hold up well today. So it went in favor of a new one from Scale Sound Systems. The decoder worked but I didn’t like the 9-pin connector so I swapped it with an 8 pin Loksound Select that needed a home. The wiring for the lights couldn’t easily be replaced but I did want to group it all together into a single harness that I could unplug to remove the shell. I also wanted to connect everything to some sort of motherboard, and though I have quite a few of Nix Trains Decoder Buddy’s, I didn’t want to waste a new 21 pin decoder on a locomotive that doesn’t need the additional outputs. I also would have had to request a custom board without resistors for the LED’s since those were already installed and glued to the shell.

Fortunately, I’ve always wanted to etch my own circuit board and it just so happened that I had picked up a bottle of etchant, some dry transfer trace masks and a few double sided copper clad board at a nearby Radio Shack when it was going out of business. I had my doubts about the efficacy of the masks but they weren’t expensive and I’d only be out a bit of time if they didn’t work. I had drawn up a diagram some time ago that had been intended for my scratchbuilt SD40-2’s but would work fine for an F7 if shortened a bit. Before attaching the masks I drilled through the board to accommodate the mini connectors that I would use to attach the decoder and lights to the board. I bought these from Micro-Mark but they’re pretty standard and can likely be had for much cheaper elsewhere. With the holes drilled I masked the areas of copper that I wanted to remain and found that the transfers actually went down quite easily. No different from dry transfer decals. There were a couple of ragged edge where the burnishing tool caught the material but nothing that would cause an issue. Unfortunately I forgot to take a photo of the board at this point.

With that I was ready to etch the board. The Radio Shack etchant contains Ferric Chloride, which is not a good thing to spill on oneself so I donned the appropriate PPE (long sleeves, gloves and probably safety glasses, though I don’t remember. You should certainly wear eye protection if you try something similar) before pouring it into an appropriately sized glass container (plastic would work as well but obviously metal isn’t a great idea). I then submerged the board for a few minutes, swishing it around constantly. Once the unmasked copper had disappeared I dunked the board in a container of fresh water to neutralize the Ferric Chloride. Clean up wasn’t too bad. I used a plastic funnel to pour the used etchant back in the bottle, the dissolved copper may dillute it a bit but I didn’t have another container that I felt would be adequate.

With the board clean and dry I scraped off the masking and was pleased to see how crisp and accurate all the traces were. Next I soldered on the connectors, which was fairly easy. Then I performed a quick continuity test. Somewhat surprisingly there were no issues.

With the board sorted I now needed to figure out how to mount it in the model. Because it’s double sided I needed to be careful about what the bottom of the board could touch. I ultimately decided to pad the top of the motor with some .020″ styrene to insulate it from the board and just wrapped the whole assembly with Kapton tape. The motor leads are soldered directly to the board. One is on the bottom and the other comes up through a hole and connects to a trace on the top.

Next I used some AWG36 wire from ESU in the correct colors and built my wiring harness. I bent the pins on the top of the connector 90 degrees to clear the shell and soldered the wires to those. The other end of the wires were soldered (with some difficulty) to the remaining magnet wire in the shell. Everything was insulated with heat shrink tubing from Ngineering.

Finally I turned my attention to the speaker. This is a sugar cube mounted in a 3D printed enclosure. It sounds amazing but didn’t quite fit in the frame. Fortunately the walls of the enclosure were thick enough to let me sand away enough material for it to sit properly. Once that was taken care of I hard wired it to the decoder.

I didn’t want to shorten the leads on the decoder’s harness so I coiled them up and wrapped them to the decoder with Kapton tape. I then plugged the decoder in and attached it to the board with, you guessed it, more Kapton! I adhered the motor to the frame with (nope, not Kapton, haha) clear silicone (for ease of removal) and soldered the pickup wires from both trucks to the appropriate traces on the board.

All in all I’m pleased with how everything turned out. There are no wires to rub against the flywheels and everything looks neat. It’s also very simple to remove the shell and replace the decoder or do any other work that might come up.

One other thing of note: I took the trucks apart and replaced the leads as I wanted to use thinner, appropriately colored wire. De-soldering the old wires ended up being a bit of a process due to an insane amount of corrosion on the joints. After sanding all that away I was able to get the wires off with no other issues. A bit more cleanup and a bit of tinning and the new wires were soldered in place. Not sure what caused the corrosion. It might have been from whatever flux I was using at the time or maybe humidity in one of my old apartments…

Once I had everything hooked up I did an initial programing on the decoder. I downloaded an appropriate sound file from the ESU Loksound website and set a speed curve and some other stuff. For some reason the headlight doesn’t dim but everything else works well. The Scale Sound System speaker enclosure is a huge upgrade and sounds great even with the volume turned way down, as is my preference. If you want to sort of hear what it sounds like, I posted a video to YouTube.

I’ll leave you with some proper photos of #101 as it sits now. You’ll have to excuse the bits of flerm and the shabbiness of the paint. A cleaning is in order as well as some replacement decals and a slight re-weathering. Clearly that front coupler box is far too shiny. All that will have to wait though. I have plans to start on SNCT #102 and #103 in the not-too-distant future and I’ll touch up #101 when I get to the paint and weathering stages on those units.

Thanks for reading this far. I appreciate your interest and hope some of what I’ve done here will be helpful for your own modeling. As always feel free to ask any questions either below in the comments or on Facebook.

This post is going live on January 1, 2021 (in the United States) so a happy New Year is in order. I managed 6 posts in 2020 which is not so bad for me. I’m hoping to get at least 10 out this year, hopefully on the first of each month (except July and August when I’m usually too busy with outside stuff.) I’m working on a lot of new projects and the layout will make a reappearance here sometime in March. 2020 was a surprisingly productive year for me considering everything that went down. Here’s hoping 2021 is just as productive while also being less of a complete shitshow.

Chris