Sumida Crossing, August 29, 2009

And not only another month, but another year has passed. Not much happened in August; as I mentioned last time I’ve mostly been working on the Arduino project. So this month’s post will focus on the past, but will also look forward to the future.

This month marks the fourth anniversary of Sumida Crossing, dating things from the start of construction. Planning actually started earlier, around June of 2009 in earnest although there had been a lot of thought prior to that. And the first real train didn’t run until early 2010 (unless you count a test on a loop of temporary track). And actually, although the first post in this blog dates from September 16, it wasn’t actually online until the end of November. Prior to that I’d been working on the initial version of the website offline, and hadn’t bought the domain name or space on a server until I judged it ready. I don’t think it even had a name before November; I’m pretty sure I made that up when I bought the domain name.

But I bought lumber and took my first “layout photo” on August 29, 2009. That’s it up above: my first photo of the layout, or rather the pile of new lumber and recycled bits from the previous layout that would eventually become Sumida Crossing. It’s all stacked up against my workbench, which as usual has too much junk on it to be used for work. Overhead you can see the original halogen spot lights I spent so much time trying to get to work (a relic of the old layout I should have thrown out, instead I invested two years and hundreds of dollars in special bulbs, UV filters, and extra lights, none of which got me what I wanted). Eventually I came to my senses and started using fluorescent tubes.

I don’t, unfortunately, have a date for the first train. I know which one it was: my first Kato Yamanote Line E231, bought for entirely too much money, before I discovered how easy (and cheap) it was to buy direct from Japan. And I think it was sometime in 2007, but for the life of me I’ve no clue when. My first photos weren’t taken until December of 2008, but that was well after I’d set up the Kitchen Table Layout, and except for a couple of times I never photographed that layout, so I have no history of it except what I remember, which is spotty. It might have started at a train show in January of 2007, but I may be remembering buying a later train in January of 2008.

At first I had just the E231 and a 253 Series Narita Express (the old one, with bulb interior lights). For at least a year, and possibly closer to two, I ran those trains around in circles on my dining room table, and gradually became addicted to Japanese passenger trains. And I picked up several others in that period: the original table layout had six tracks of passenger-train storage yard, which I outgrew before starting the new layout.

When I was originally keeping this website on iWeb I kept a construction log, but when I converted to the new software, I rewrote all those pages to be more readable. The result is that my first construction page no longer has anything on it to show that it was the first, but it’s this page describing the construction and painting of my tables.

So, looking back, how has it been? Well, quite a bit more fun than I expected, to be honest, and more work as well. The early years went quickly, although I started to bog down when I was converting from DC to DCC (and I’m still not done). This was due mostly to my intent to wire up a whole mess of block detectors (which I still really need) based on a component that required a whole mess of fiddly soldering of wires, and in the end didn’t do half of what it was supposed to, which didn’t help motivate me to spend days soldering, something I really dislike at the best of times.

That product alone, the BDL168, pretty much killed my faith in Digitrax. The parts that work, do work well. That’s probably the worst part. I wanted the thing to work, and it almost did, so I couldn’t walk away. But transponding never worked reliably for me. And trying to install a half-dozen of the things with sixteen blocks each was just too much soldering. I’m at the point now where I’m thinking of scrapping all the BDL168s and switching to something else that will be much less work. I’m not sure what yet, but I don’t see myself ever finishing them.

The Past

But lets look further back in time: How did I get to Sumida Crossing in the first place, and where will I go from here?

My previous layout, the first since I’d had a loop of sectional track with some sidings as a kid, was an HO freight shortline built in 1990, and actively developed for about five years, although only in fits and starts. And despite having relatively little to that one (it was a point-to-point layout, less than 24’ long in a U-shape) I had more done on Sumida Crossing in its first year than the HO layout ever managed. It stagnated fairly early on, actually, and eventually work stopped altogether for more than ten years when I did little but dust it off once a year or so and run a train back and forth on the small length of functional track. Half the time it derailed, or uncoupled its cars halfway up a grade. Or got stuck on a dead frog on one of the switches.

I learned a few lessons from that one:

- Wiring is a pain, particularly if you don’t leave yourself enough room.
- Track plans need to be defined before you cut lumber, not after.
- Track joints across tables are problematic.
- Sectional layouts need something to keep them precisely in alignment.
- Trains don’t like sharp turns or steep grades; be conservative.
- Get the track right; fixing it later is MUCH more work.
- I don’t like building detailed structure kits. Or boxcar kits.
- DCC isn’t really very hard, as long as you plan for it.
- Ballasting is too much work, but the results can be incredible.
- Point-to-point layouts may be realistic, but I don’t find them very interesting.

So when I set out to plan Sumida Crossing, I set myself some goals. It got edited a bit as I went, but eventually stabilized to the version I put online. Those focused on the things I wanted to do, but there was a list in the back of my head based on those lessons noted above of the things I wanted to be sure not to do. Those drove decisions like the oval structure and the ability to run at least six trains at once, so there’d always be some interesting visual activity.

And after I stagnated on the old layout, I’d periodically revisit model railroading as a hobby. For a time I collected 19th Century style On30 trains, thinking about building a small rural shortline. But that never quite clicked for me. I actually started to build the layout structure, before realizing I didn’t really like the track plan, and stalled out trying to pick one I liked, ultimately deciding it was the very point-to-point concept I had a problem with. And then I decided to play with passenger trains, which led to Japanese passenger trains, which led to that first table-top layout and then on to Sumida Crossing itself.

The Future

I’m nowhere near abandoning Sumida Crossing, but I do find work slowing on it, and I’ve begun thinking more about a successor. I’m probably some time away from making a start, and once started I expect to be several years before it’s ready to run, so I want to get Sumida Crossing to the point where I can play with it in that interval. I’ve learned a few lessons this time out too:

- No soldering. Repeat No bloody soldering!
- Pay more attention to planning scenery along with the track.
- Give yourself room to work, particularly on wiring.
- Structures need room: roads, fields, etc. Plan for it.
- Do the backdrops before the layout.
- Do the lighting before the layout.
- Really, get the track plan right before you touch lumber. Really.
- I want broader curves than I can get with Unitrack.
- I really like the four-track curve above the village, have something similar.
- I like the subway, but cut-and-cover isn’t accessible once you have scenery; do something else.
- Catenary looks good; maybe do something better than Kato though.
- Plan for shelving for tools, parts, beverages, etc, otherwise it ends up on the track.
- Do something about dust control (dropcloths, drop ceiling, air-cleaners, etc).

So, given that and that I’m thinking of the next layout, what’s the next one going to look like?

Well, it will still be a Tokyo-centric N-scale contemporary layout, with lots of passenger train action. I’m very heavily invested in those models, both financially and emotionally (I like them!) at this point. But here’s what I’ll probably do differently:

- A semi-permanent around-the-wall layout (still internally sectional in case I move).
- A lighting valence built, painted and wired before the first table-top (or L-girder) is built.
- Ballasted flex track instead of Unitrack.
- DCC from day one, maybe one line rigged for permanent DC-only use for a test track.
- Wiring done “against the wall” under the layout, where it can easily be accessed. No more working overhead.
- Planned shelving distributed around the layout for power supplies, outlet strips, wall-warts for lighting, DCC gear.
- Distributed 3A power boosters rather than a big central one. Maybe even set to 1A output rather than adding circuit breakers.
- Lineside signals from day one; no trying to add them later (they may be dark at first, but they’ll be wired).
- No electronics that require soldering wires; screw terminals on everything. No more Digitrax BDL168s.
- Maybe replace Loconet with NMRANet or a simple serial bus (you can do interesting things with C/MRI and Arduinos).
- More scenery, less track (maybe; I still like having four above-ground tracks plus a subway).
- More care paid to track-too-close-to-backdrop design problems.
- Freight as well as passenger (a container terminal, perhaps).
- A visible storage yard for the most-used passenger trains; I miss that from the old table layout.

There’s more thinking yet. I haven’t even begun to sketch track plans. And there’s the small matter of buying a nice basement with a house over it (planned, but timing still uncertain). I doubt I’ll be buying lumber for another couple of years. But that’s okay. I enjoy thinking about design issues almost as much as sawdust (and I enjoy sawdust a LOT more than solder). Maybe I’ll have more to say next August, but probably not sooner.

So, that taken care of, for now, it’s time for the actual “annual musing” itself, where I recap the past year of work. Which is going to be brief, as very little “work” got done on the layout itself.

The Last Year

So what all happened on the layout in 2012-2013? Well, not much really. The first half of the year was mostly spent thinking about DCC decoders for converting my fleet. That stalled both from my own distraction, and because Don never got around to finishing his decoders, which were likely to be my preferred model out of the ones I evaluated (although I never got a final version, so I never finished my testing).

I got back into buying trains, after a bit of a dry spell, adding a commuter E233, a model of the current Imperial Train, and my first Tomix train (other than freight locos), the E1 MAX Shinkansen. I also picked up a couple of others that I now realize I’ve never posted about other than noting them on my Recent Arrivals and Collection pages: the “green wrapped” Yamanote E231 celebrating the 50th anniversary of the original green Yamanote trains, my MicroAce E926 Shinkansen inspection train, and a pair of Marunouchi Line Series 02 Subway trains. I’d set out to get one new train per quarter, and ended up with nearly twice that. Well, I’m clearly still addicted to the things.

And I did some structure work, disassembling, painting, and reassembling Kato pre-built models. Well, mostly the first two, as I’ve got a half-dozen in various stages of taken-apart, that I need to get cleaned up and repainted while it’s still warm enough out for spray painting. Then I need to actually work on them over the winter. Hopefully the “one point five meter line” will help motivate me to do the latter.

And I’ve spent a lot of time this past six months on turning an Arduino into a DC train controller. As with most of my programming projects, it’s taking a LOT longer than anticipated. You’d think I’d have learned by now that I’m not a fast software developer.

But that’s it. I need to make more happen in the next year. More new trains. More finished structures. Some actual working DCC track so I can run things. Some trains converted to DCC so I can run them. I’ve no shortage of to-do list items. It’s time to do some of them.

Tags: Administrivia, BDL168, Arduino, Design, Digitrax