DCC


The DCC system for the layout will be a distributed system, rather than using one central DCC station. A main JMRI station will be used to provide DCC commands for train throttles and any accessories, and will support use of wireless throttles (e.g., smartphones). The signals from this will travel over a dedicated communications system to a larger number of small boosters, which will provide the actual track power to “power districts”, a section of track, or tracks, larger than a signal/detection block, but still relatively small.

The intent here is that an individual booster will provide power to one or a few trains in mainline use, and possibly a larger number of trains on staging tracks or other less-critical areas. A double-track line will thus need one booster per track every few signal blocks.


Requirements


My requirements for DCC aren’t very complicated:

- Interface to a computer running JMRI.

- Voltage in the 12 - 14 volt range. The low end of that is probably better for the motors, but has some implications for top speed of Shinkansen. One option is to run Shinkansen at a different voltage, although that could affect motor or lighting longevity.

- Provide enough power to run all of the trains I expect to run, even if I later add sound.

- Provide the ability to run enough trains simultaneously. I’m not certain what “enough” is yet, but it’s more than 10.

- Prevent a short by one train from affecting other trains, except those nearby on the same track.

- Keep it simple: easy to assemble, easy to debug when it goes wrong. A modular approach, if possible.

- No #$@! soldering of wires except for track feeders to rails. I am never installing another board that requires soldering to 44 pins in a tight space. Life is too short.

Related to this is that I want to support both hand-held wireless throttles and smartphone-based software throttles. But that’s not strictly a “DCC” requirement, it’s a requirement on the front-end control system. Most DCC systems don’t separate the throttle bus (aka “cab bus”) from the actual command-station-to-track DCC bus though, so this potentially constricts my choices.


Power


Actual power requirements per train, for a LED-lighted N-scale passenger train without sound, are in the range of 0.25 amps. With sound, this can rise to around 0.5A from what I’ve read, although the average is probably slightly lower. A power district circuit breaker that trips at 1 Amp would thus be sufficient for one train. A train with bulb lighting (I have a couple) could be more problematic, as bulbs can draw up to 60 mA of power per car, which adds up (bulbs have a serious issue with DCC—heating—so I may not run those trains on DCC anyway). Setting the breaker somewhat higher (1.5A to 2.0 A) to allow margin for bulbs or other high-draw trains may be desirable.


Design


This is still speculative, as I need to work out some potential issues and do a bit of testing.

I’m currently planning to use the SPROG DCC Power Booster (also known as SBOOST) for my power districts. This is a 2.5A booster that retails for about US$100. This has a fixed 2.5A circuit breaker, so I can probably count on powering up to four typical trains per power district even if I add sound (you should never use 100% of the circuit breaker rating). Without additional circuit breakers to separate trains from each other I probably wouldn’t want that many, but two or three sequential trains on the same line could be powered this way, since any problem with one will require stopping the others while fixing it.

This approach is a bit pricy. Compare a 16-Amp Digitrax DB220 (US$180) and a pair of PM42 quad circuit breakers (US$65 each), which can provide 8x 2A power districts for about US$40 each. The advantage is that the SBOOST-based approach is modular. Each booster is independent of the others, so electrical problems can be more easily isolated. Plus it has no card-edge connectors (I’m aware that you can get breakout boards for the PM42 now, which address that problem). And I really like using screw-terminal connectors.

One advantage of the Digitrax approach is that you can set the breaker higher than the average level of 2A. But that’s really only an advantage in larger scales or with much older equipment that has higher per-engine power needs. Another advantage is that you could set it lower, and get more districts (by adding more circuit breakers) off the same supply. For example, a 16-district system (4x PM42) would cost about US$30 per district. So the choice is modular and simple, at a higher cost, versus flexible and more complex, at a lower cost. It’s not an easy decision, but I’m leaning to the SBOOST right now.

Another potential issue with the SPROG is what to use for throttles (other than on-screen JMRI throttles and smartphones). I have some ideas, but this needs more work.