A few weeks ago my wife and I had started work on the Atlas O Signal Tower model, the first plastic model I’ve built in probably twenty years. I am trying to “do it right” this time and not take any shortcuts, so when I realized that the walls were going to be so thin that any interior lights would make the walls glow, construction stopped and I headed to the hobby store to buy some paint.

Now I’m finally getting around to doing that painting. I had to paint the lower level assembled, but the second story was unassembled so it was much easier to get some paint on – next time I’ll paint everything before starting.

The results are good; as you can see above Before there was quite a bit of light coming though, and now after about three uneven coats of flat black there’s very, very little. I intend to give the interior another coat of off-white to make it a bit more interesting inside, as well as possibly adding a dividing wall, furniture, etc. to provide a bit more interest.

The photo above is being lit by a single warm white LED which is pointing down. I’ll probably have a ceiling light in the lower level and try to do one or two lantern style lights in the upper floor, which will have a worker in it manning the switch levers. Not quite sure how I’ll turn a LED into a lantern yet.

There’s an exterior gooseneck lamp too, which I spent quite a bit of time researching sources for a few weeks ago. There are some cheap lamps from China that are similar on eBay, but the Cadillac seems to be the Gooseneck Light Kit from Ngineering. If they sold them assembled I’d have ordered a few immediately, but the construction process looks somewhat daunting to somebody who’s pretty new at modeling.

Progress has been slow the past couple weeks. I was away from home for work, which didn’t help, but the bigger limiting factor right now is that I’m still waiting on the larger Atlas O curves I had ordered. The remaining switches arrived, which is good, but I have decided to make alternate arrangements for the curves. It’ll probably be another week before they get here. Even with my paper track stand-ins from the prior post, I’m reluctant to actually start doing much serious work beyond setting track on foam until the final pieces arrive.

I did spend some time tinkering with grade up two inches to the inner section of track. My Lionel CSX SD80MAC locomotive had some trouble getting up the grade on my prior layout, and with this new (much gentler!) grade I’ve been seeing the same thing. Except on this layout I can actually see what’s going wrong. Essentially the two outer inner axles on each of the trucks are the motorized wheels; the outer axles spin freely. When the grade increases too sharply, the locomotive gets stuck on those two outer sets of wheels. I’m also noticing that this locomotive’s pick-up roller, which has caused problems on the Atlas O switches in the past, is not negotiating the lovely O-72/O-54 switches very well. I think the best solution may be to look at limiting its downward travel. Tyler at Legacy Station told me some people use lash-ups to solve the problem (two or more locomotives working together) – the second locomotive pushes the first one over the trouble spot. An expensive fix to the problem, to be sure.

This weekend I began doing some preliminary work on the trestle bridge in the form of building the bent jig, shown at right. I’m pretty pleased with how it is turning out. Right now I’m using 3/8” dowels for the pilings, 3/8” square for the cap, and 1/4” square pieces for the sills and diagonal supports (not shown). The posts here are 16” long, which comes out to 64’ in O gauge, if my calculations are correct (16*48/12).

The next step is to do a stain test and if all goes well begin building some bents. I’ll also need to figure out how I’m going to build out the gorge that this trestle will cross. Building up (for mountains and so forth) is pretty easy; building down will be a bit different. I imagine I will end up building up from the floor. The L-girder benchwork is so easy that when I actually have to solve a benchwork problem, I’m not sure how to go about it. Access hatches will be another such problem to solve.

If you want to nerd out on trestle construction, there’s a pretty great post by Bob Hyman, Wooden Trestle Construction on the Rio Grande Southern Railroad.

The larger Atlas O curves in my layout (O-81, O-90 and O-108) are on backorder, so in order to put down what track I do have, I printed out actual size stand-ins for the missing track. Awkward, but it works.

With my findings about operating noise from the foam, I knew I’d need a number of sheets of 1/4” plywood, as well as more foam (I want to build a mountain eventually, after all), so I got my friend Joe to help out with an acquisition trip with his trailer. We got the foam and plywood home and this weekend I’ve been working on finishing out most of the remaining joists as well as laying down the plywood and foam.

The layout really starts to feel like it’s taking shape now that I’ve got the foam cut to fit. I did a test fit with some of the track last weekend but several of the larger curves (O-81, O-90, O-108) and a couple switches are on backorder, so that’s going to make it tricky to get the track fully laid out to make sure it really fits.

One issue that I did notice was that my ideas about what I would be able to reach were quite optimistic, leading me to seriously rethink the need for access hatches. The rule of thumb seems to be that you should require no more than a 3 foot reach, but even that feels a bit long, and I’m pretty tall.

Last weekend my wife and I worked on putting together the Atlas O Signal Tower, probably the first model I have worked on in 20 years. Humbling in many ways but I’m trying to do it right this time. I’d like to paint the insides so that the walls don’t glow when it’s illuminated, as well as look into doing some very basic weathering (mortar for the bricks, mostly).

The model has a gooseneck lamp over one of the doors and that got me thinking that it would be great to have an actual light there, so I spent quite a while researching options for this. While there are quite a few cheap-looking lamps of this sort on eBay (they all ship from China), I did find a kit from Ngineering which looks very cool, although possibly at the very top of my skill level. If you know of a better source for these (preassembled, maybe?), I’d love to hear it.

Still on the thread of lighting, I thought it would be cool to have a lit lamp inside the signal tower on the operator’s desk or thereabouts, but I found it practically impossible to find any sort of kit for this. There are tons of vendors selling tiny LEDs, but lamps and lanterns are oddly hard to find. A vendor called Berkshire Valley makes little metal lanterns but I’m not certain how practical it would be to modify it. At this point my best idea is to try to create a lantern in modeling clay and then use Amazing Mold Putty to make the plastic version. I got this idea from a great little series of videos where Mark TrainMaster uses this product to duplicate dwarf signals (part 1, 2, 3). Perhaps this is an obvious task to someone with more modeling experience than I have.

I ordered a kit of hot wire foam cutting tools from Hot Wire Foam Factory: the Freehand Router and 4” Hot Knife. The plan is to use them to terraform this 2” foam. After a brief trial I think they are pretty good tools, but am concerned that my foam is quite a bit denser than the foam these tools are designed to deal with (styrofoam). In cutting a chunk out with the hot knife the wire kept bending back by 45 degrees, and the freehand router had similar issues. I did notice that the power cable leading to the tool gets pretty warm, which makes me think that I might get somewhat better (hotter) results by making my own cable out of higher gauge wire. Certainly, more practice is needed.

So with the plywood decking screwed down and the foam resting on top of it I think I’m ready to lay out what track I’ve got and start tracing it out, as well as figuring out a ditch-cutting hot wire rig.

As mentioned previously the landscape surface of my layout is going to be extruded foam, the pink stuff. In doing some tests I quickly found that there’s a significant amount of operating noise when putting the track (I’m using Atlas O) directly on the foam and rolling my heavyweight O gauge passenger cars on it. I ran a series of tests with the following materials, from top to bottom:

• K-Line Heavyweight Passenger Car
• Atlas O 40” Rigid Straight
• Woodland Scenics Track-Bed
• 2” Owens Corning Pink extruded foam
• 1/4” Lauan plywood
• L-girder benchwork topped with 1x4 joists

In short I found that by placing a sheet of 1/4” Lauan underneath the 2” extruded foam I was able to significantly reduce operating noise. Adding the Track-Bed improved upon this even further. Here’s what I found, in order of noisiest to quietest:

• Track, 2” foam – unacceptably noisy
• Track, Track-bed, 2” foam – unacceptably noisy - 1
• Track, 2” foam, 1/4” Lauan – not bad
• Track, Track-bed, 2” foam, 1/4” Lauan – very, very little added track noise

I don’t have the equipment to measure this scientifically, but my sense is that the foam amplifies whatever bass component there is to the train rolling on the track. Thankfully 1/4” of Lauan is enough to deaden that component back out.

In the interest of completeness, many people on the forums like to use Homasote boards as their sub-roadbed, but Homasote is pretty difficult to find in the southeast. If you’re thinking of doing a plain plywood sub-roadbed, Homasote seems to be worth looking into, but in my own small layout experience a plywood subroadbed was fine, even without the Track-Bed. Still though, I had cut the plywood to outline my track (like cookie-cutter style, but with the non-track areas removed, which was probably a mistake).

So I’ll be layering Lauan beneath my foam, which should give it a tiny bit of added stability too. Added expense and work, but worth it.