This is Going To Take a Bit

I was inspired earlier in the week to start in on a(nother) project, converting an ancient, original center drive Weaver RS-3 to Proto:48. While Comstock Road can only plausibly host one locomotive at a time, having a backup is something I want both in case of unexpected issues and so that I can still push cars around while re-detailing one to proper CNR outline. Thus, I picked up a consignment Weaver RS-3 with drive issues and ordered the drop in conversion axles from NWSL. There it sat for a couple of years.

For a change of pace from machining type model building, I took up the project again expecting to have the swap done in an hour. Here are the two axle types for comparison.

You can see the first gearbox already converted in the background. It was non-trivial only in the sense that the little nut and bolt sets holding the gearbox together were very stubborn. I foresee a purchase of precision nut drivers in my future because getting a good grip on the nuts with the end of my needle nose pliers was challenging.

Sadly, my quick project was not to be. I got the second axle of the first truck done and reassembled the truck. I quickly discovered that the new axles which are supposed to fit into the holes in the sideframes which bear the weight of the model were too short. A bit of internet searching later turned up reports of others with similar issues. It appears that a packaging/labelling error is too blame. Being way past a reasonable return date, I elected to go with another solution.

The alternate solution is to narrow the bolsters and cut new notches in the tabs on the sideframes that clip into the bolsters. Not for the faint of heart but not undoable. The upside of this approach is that the sideframes are brought in to closer to scale width. The downside is that this is a one way trip whereas the drop-in would be reversible. (Proto:48 locos have a much smaller market than O 2-rail.)

I considered making the modifications with my mill but I have no experience in machining plastic and did not want to risk melting or shattering something that is hard to replace. I settled on a combination of jewelers saw, x-acto knife and files. The initial effort before life intervened is a qualified success. I need to tweak the notch but otherwise I can believe that I can do this. We shall see.

If it all goes pear shaped, I plan to mill a replacement bolster out of brass or aluminum. Hopefully, that won’t be necessary but knowing that I have a backup plan gave me the confidence to dive in and start hacking away.

Operating Session

This afternoon I decided to stage a simple operating session as a way of assessing the current functional state of Comstock Road. I have run the loco back and forth once in a while but haven’t really operated at all. My plan was simple, get all four of the operational (aka couplers on both ends) cars in the fiddle yard into a train, number the industrial spots on the layout, randomly generate four numbers and put the four cars wherever the numbers said. And identify any issues as I went.

Overall, it went surprisingly well considering how little maintenance I have performed in the last year. Once I was set up, it took about30 minutes to complete the setouts. The pink foam wasteland has train service!

I did encounter some issues that I will need to address. Here are the ones I remember because I am not yet systematic about it to write them down as I go…

  • Layout need to be vacuumed to remove metal swarf. Proximity to milling machine makes this a recurring problem. Not a lot of bits but how many little bits of aluminum spanning a gap or steel sticking to a motor does one want? Plan is to acquire a plastic shower curtain and hang it between mill and layout.
  • Wireless throttle battery was flat. I thought I had it set up to charge off the general shop circuit but that has not worked. Investigation required to avoid operating while plugged into charger on end of extension cord.
  • DCC base station had been unpowered long enough to lose date and time settings. Although the default 1970 date was appropriate for Comstock Road, that would not be how I would do it. Need to either power up the layout more regularly or move that unit onto another circuit that is.
  • No uncoupling tool ready to hand. Scrounged a small flat bladed screwdriver which I kept losing track of. A dedicated swizzle stick or something is in order.
  • A couple of derailments happened, all in the same place and with the same car. Subject car is very light and was coupled to the loco which is very not. Loco coupler is the stock Atlas O one with a spring designed to toss diecast cars around on three rail tinplate layouts. Car has San Juan plastic wheelsets. Is likely some combination of those.
  • Banged head on lighting valance leaning over to align the transfer table. Stepper powered operation will fix that. Work in progress.
  • Pink foam wasteland and pint paint cans to not an attractive photo backdrop make. Motivation for scenic aspects of layout construction increases.

And that is it, other than a general desire for more cars to push around. Not perfect but mostly things with well defined and reasonable solutions. Perhaps I will schedule a regularly recurring op session for myself as a way to get me moving on some of this list.

Sieve Experiments

My soil sieve arrived unexpectedly early so I took advantage of the sunny weather to experiment outside and see what I could come up with. I started with some of the contents of a leftover bag of limestone screenings with thoughts of producing my own ballast. (When it’s not cinders, ballast in southern Ontario is usually limestone.)

The sieve came with three different meshes: 1mm, 3mm, and 5 mm. 1mm is about 2 O scale inches so about right for ballast. Or that was the theory, anyway. What I failed to account for was all the smaller bits and outright dust that also passes through that mesh. What I got was good “dirt” material but not ballast.

I then hunted around the house for something with a finer mesh. I was partially successful in that I found a bit of plastic screening but it looks to be about the same as the 1mm. I tried sandwhiching it between two of the screens and did get some “ballast”. I think it looks darn good but the amount produces is such a low ratio to the total material processed that I could not reasonably produce enough to do even a small layout such as Comstock Road.

Not to be discourages, I decided to use the “dirt” as a first texture layer on the foreground test scene. It is undeniably an improvement over brown paint.

Next I need to round up a suitable brush for stippling on glue and shoot some grass on this thing.

Foreground Staging: Useful Scenery Practice

I have been sporadically putting in the foam scenery base of Comstock Road but with no sense of urgency. I think that part of that lack of drive is due to a bit of uncertainty about the next steps. I know what they are, more or less, but have not done some of them in a long time (mixed Sculptamold in various consistencies and configurations) or ever (applied static grass). What I need is a practice project with low investment, material or emotional.

Coincidentally, I have been using my scheduled reading time to catch up on my pile of partially read model magazines and came across the perfect project concept. In the November 2020 issue of Railroad Model Craftsman, George Dutka presents the idea of foreground staging. This is a small, simple, scenicked diorama used to provide foreground in layout photography and hide the front fascia. It also provides a way to temporarily deploy structures that don’t otherwise have a home. (If you are familiar with George’s work, you will know that he probably has an extra structure or two about the place. 🙂 )

Given the narrow depth and close to the edge track locations of Comstock Road, some foreground staging is something I can definitely use. The entire scene in front of the traverser will only be about 6″ deep. The diorama is inherently expendable and quick. So I am off! I have gotten as far as the “paint the Sculptamold” phase but am stalled a bit on some materials.

I am assembling appropriate ground textures from local sources, a process much slowed by the current Toronto area lockdown but not impossible. Hopefully, once I get ahold of some suitable sieves I can get this done. With curbside pickup and online shopping only, I can’t stroll the housewares aisles looking at the size of the meshes in the strainers so I have resorted to the online retailing behemoth for a set of soil sieves. Now we wait. And collect and dry used tea bags. By the time I get this done, I should have shaken out all the bugs in a basic scenery system.

As a thoroughly unimpressive illustration of the concept, here is what the aforementioned foreground track looks like with and without my work-in-progress foreground bit held in front of it.

Including of a structure to frame an edge will require a steadier setup than board held in left hand and phone in right. George recommends accumulating a suitable stack of boxes.

About That Foam Scenery

During some of my scheduled shop time, I have been working towards getting all of the spaces between the tracks filled in with foam. The first priority to is relieve my anxiety about any derailments involving rolling stock making a fast trip to the concrete. Nothing even close to that has happened and I am generally well pleased with operational performance so far but it is still something I worry about.

Getting all those non regular shapes cut and fitted is time consuming and got me thinking about what I might do differently next time. I haven’t really settled on anything but I did want to share what I have learned about this approach. Queue the list.

  • Retrofitting foam is fiddly and time consuming. No matter what methodyou chose it will take time and probably involve a mess somewhere.
  • I have tried tracing the shapes onto the underside of the foam from below. This is awkward and not as accurate as I hoped. I always have to trim things down.
  • I have tried “routing” the shapes by tracing the edges from above using hot wire tools. This produces closer shape matches but has the various drawbacks of hot wire tools. Ironically, the Hot Wire Foam Factory router tool doesn’t work as well as the “knife” tool. The router is too short and too thick.
  • Foam board aka extruded polystyrene rigid insulation is very slow to cut with hot wire tools. The Hot Wire demo videos all depict white foam “bead board” and are clearly the intended material. I will consider using that instead for future projects if I don’t need the structural feature of the rigid foam.
  • Polyurethane glue works a treat but squeezing the bottle (LePage 200ml) for long beads is hard on the hands. I have resorted to using a quick grip clamp as a squeezer.
  • T-pins are great for holding foam bits together while the glue cures. T-pin use number eleventy-one.
  • Fumes from hot wire foam cutting are unpleasant. duh.
  • Getting the pictured foam in along the back edge strongly validates the removable backdrop design. It would be a real bother to do from the front.
  • Ditto having the layout lighting working.
  • Glue instructions say to dampen surfaces prior. Choose a stable container for your water supply. Just sayin’…

I expect to have everything roughed in sometime next week and will probably roll straight into some final contouring with additional layers of foam and sundry coatings. I have a jar of Foamcoat I intend to try out in addition to the traditional Sculptamold.

Lighting Update

I previously described, I installed an LED strip in valance of Comstock Road. I applied it using the self-adhesive backing which turns out to be less than permanent. On multiple occasions, I entered the room to see one of the two strip sections dangling down, held up only by the wires at one end. This, as they say, won’t do.

After a bit of research, I decided to try double sided carpet tape. If that fails, I will go all the way to stables or some other mechanical fastener. Having got the valance down to do the taping (not trying that upside down and leaning into the layout), I decided to continue with running the wiring.

I will likely change power supplies when I add a second strip so the current power brick is temporary-ish. I have been hung up on the idea of some sort of shelf but decided to just zip-tie the brick to a cross member.

Zip ties, what can’t they do? El Kabodge strikes again.

Roller Gauges: Design Meets Reality

Two Proto:48 Roller Gauges

When I got to actually laying out the dimensions on the first gauge (ie, marks in big Sharpie with dial calipers, I belatedly realized that the .036 flange width on the ends would be very delicate and vulnerable to damage if one, er, hypothetically dropped one on a concrete floor. Which is why my previous efforts had .100 rims rather than aspire to fit right in the frog of turnouts. I opted to repeat that choice for durability’s sake.

Being an aspiring novice hobby machinist, there are things I know need improving in the execution of these parts. Measuring those small gaps with dial calipers isn’t the most precise method but all I have that works. The finish isn’t as smooth as it should be which I know how to fix but will require developing my tool bit sharpening skills. There are probably things I don’t know that should be improved, too.

Anyway, parts done and sent off into the pre-Christmas postal maelstrom.

Designing a Roller Gauge

I have a request from a fellow Proto:48 modeler for some roller gauges. These handy things are just the ticket for handlaying track and, as far as I can tell, not something you can get for Proto:48. (There are commercially available gauges for most of the regular track standards) While I am not prepared to make lots of gauges on my entirely manual Myford lathe (You would want a CNC machine to do this commercially) I am not averse to knocking out a couple. Once I finished my Milling About and rearranging the shop…

Milling trials having been concluded (more on that later) and shop having been rearranged including new task lighting and a shelf over the lathe, I am ready to actually make something which leads us to objective of this post.

The relevant dimension of the target rail is the width of the rail’s head since that is the part the gauge has to fit over. The measurement reported for ME code 125 is 0.056″. The rest of the numbers needed come from the NMRA trackwork standard for Proto and Fine Scales S3.1. From the Proto:48 line we get a gauge range of 1.177-1.203, a flangeway width of .036-.039, and a minimum flange depth of .026.

From those numbers, we work down to something I can aim for on the lathe. I say aim because I can miss a dimension by a couple of thou and still get working gauge. A couple of thou is a long way in machining(famous last words).

  • The desired target gauge is the middle of the standard so 1.190.
  • We want the gauge to fit through minimum flangeways so .036 wide and .026 deep.
  • We want the gauge to be accurate without being difficult to fit on the rail or roll along it. This suggests a loose running fit (had to look that term up) so I will add .003 to each slot to get .059. As a check, 1.196 is well within the standard’s maximum allowable gauge of 1.203.

Here is a very low tech sketch of the planned work:

Proto:48 roller gauge for ME code 125.

This Thing is Lit

While things have been off and on lately, I have finally managed to get the LED strip installed on (in?) the valance and wired up. The wired up part is only to the valance end as evidenced by the unsightly cord and power brick perched on the layout at back right. It should not take much more effort to get some concealed wires going down the back of the wing. I just need to settle on how to get from there to under the layout in a easily disconnectable manner.

I am happy with the lighting result but the intensity is on the weak side for photography. I will almost certainly add a second natural white strip in the not to distant future. I might make it switchable so I can vary intensity to suit.

I think that I will next have a go at complete the basic landforms and roll right into mocking up buildings.

Almost Done Painting

I was going to report that I was done with all the painting until I realized that I still need to do the rear of the backdrop sections. But all visible wood has been done. It is just waiting to finish drying before I reattach the front wings onto the sides. The quick install was verified as I reinstalled the sides as a convenient place to put the dry-to-the-touch pieces while they finished curing.

Things are going slower than they might have because my attempts to replace the now used up 21 year-old! can of Behr flat black paint foundered on the twin rocks of the whole paint code and formulation system having changed during that period as well as Behr seeming to have stopped shipping quart cans of flat. Gallons and sample size you can have but not a quart. No idea why but I do know that a sample size is not enough and, given how long it took me to use up the last quart, a gallon is more than a lifetime supply.

The resolution was to get a quart of Glidden flat black instead. The colour matches but the sheen is not identical. Fortunately I arranged to finish whole sections with the old can so nothing has to be half and half.

Once I get it all done and post the final result shot I will be out of excuses for hanging the LED light strip and getting on with the scenery.