USA Trains Speeder Tips

• The USA Trains Speeder
• DCC Installation 31 May 10

A USA Trains Speeder followed me home from the Big Train Show this afternoon. I managed to get it on the track for a little testing before actually packing to go on travel for about 8 weeks. It ran pretty well. The track hadn't been cleaned in a month and was pretty heavily oxidized but it ran. It sputtered, but it ran. I ran the pole sander over the main line and then it ran just fine. I also ran it in analog mode with DCC on the track and besides for the typical DCC buzz in an analog loco, it also ran fine. There is a light amount of gear noise but it runs quite smoothly.

The speeder is not feature heavy. All it has is a motor and a non-directional headlight. There are no conventional couplers, but there is a drawbar of sorts at the rear.

It tracks well and power pickup is reliable as long as the wheel wipers don't get bent. However, the speeder is small and light so it is intolerant of debris on the track. Hit a leaf or twig and it'll derail immediately.

The speeder picks up power on four wheels, but only the rear two are powered, similarly to the Lionel Handcar. The power pickups are metal wipers that ride on the backs of the wheels. I checked the motor current and it is very low. At 18 volts running light, it draws between 60 and 70 mA. At full stall, it draws 700 mA. An HO or even N scale DCC decoder will do fine in this one provided that I can find the room. However, this will have to wait for a couple of months as I will be out of town for most of that time.

DCC installation in the Speeder is a piece of cake. The only real issue is finding room for an HO scale decoder, I used a Digitrax DH163D (wired) decoder because I had one and I could evaluate how DCC would work in this loco without spending any money. The plan was that if it worked, I'd consider buying some HO sound decoder and return the DH163 to stock.

There are four screws in the motor block that release the block from the body. Then the bottom part of the body will come out by gently prying on the "lift handles" at either end. There is a cast resin weight in the rear of the car. This guy completely fills a cavity that is ideal for stashing the decoder. The obvious solution is to ditch the 1.4 oz weight and replace it with a similar amount of something denser, such as lead. This then leaves volume available to stash the decoder.

I used two 1 oz fishing weights, held in with a liberal amount of hot glue. You could also pour lead shot (recovered from a few 12 ga shotgun shells) and then bind it in with hot glue, silicone sealer matte medium or diluted wood glue.

The Speeder uses a black and red wire to connect the track directly to the motor. These wires are reversed in position as compared to HO standards, but large scale also reverses the track polarity. If you are interested in having the Speeder analog convert properly, then just follow the HO color guidelines and ignore the left and right stuff. Then it will work fine. I cut the headlight wires off right next to the motor and then cut the red and black wires at the highest portion of their loop when they are pulled out as far as they will go. The red and black decoder leads go to the red and black power pickup wires. The gray decoder wire goes to the black motor wire, the orange decoder wire goes to the red motor wire. I wired up the headlight to be on in the forward direction only using the blue and white decoder wires. The lamp is incandescent so it doesn't care about polarity. It also draws only 23 mA at 20 volts so it doesn't need a current limiting resistor either.

The decoder is then stuffed up into the body and the floor is reinstalled. Then the motor brick is reattached with the original screws and it's done. It worked right off, no problems.

The DH163D was working fine in the Speeder but it doesn't provide sound. The Speeder is one of the last locos that I have that ran in stealth mode. So the DH163D came out and a Soundtraxx Tsunami went it. The TSU-1000 sound decoder isn't very suitable for large scale because it is rated for 1 amp stall current, a spec that few large scale locos can meet. However, the Speeder draws only 700 mA at full stall so the Tsunami could handle it. It also just barely fit.

The wiring was very simple, all the wire colors matched those of the DH163 except for the two purple speaker wires. The Tsunami uses an 8 ohm speaker and makes pretty good volume as compared to other HO sized decoders. I originally tested it with a much better speaker, this one cannot reproduce the bass but it does well enough.

The Tsunami produces a wide range of sounds and the decoder supports BEMF, this is a good decoder for small locos that can handle it.

After filling up the back completely with weight and the decoder, the speaker had to go somewhere. This is a 1" 8Ω speaker that came with a Dallee system. I made an enclosure out of a cardboard tube and hot glue. The enclosure will have to get painted, probably silver to match the rest of the unit.

Eventually, I built a whole new speaker enclosure with a "high bass" speaker. The enclosure is still made from cardboard (model rocket body tube and some scrap cardboard) but this time I assembled it with Zap-A-Gap CA and it came out a little neater. It sounds better too as can be heard in this Quicktime movie.

Update 31 May 10

The Tsunami sound decoder never really liked the high track voltage, 22 VRMS, at the GIRR very much. It would complain by blinking it's headlight fairly often. This indicates an error condition. I finally elected to do something about it after Greg Elmassian commented about the poor reliability his Tsunami exhibited on his higher track voltage of 24 volts. The Tsunami is rated at 22 volts, but it is an HO decoder and would have not been extensively used or tested by Soundtraxx at 22 volts.

One way to drop the input voltage is with a resistor, but then the voltage drop depends on load current as well. Another way to do it is with a simple bidirectional diode stack. The four diodes in this stack produce a drop of 2.8 volts total, enough to get the voltage down near 19 volts which provides some margain against the specification. I used 1N4005 diodes for this stack, but virtually any 1 amp silicon rectifier will work since the Tusnami is rated at only 1 amp anyway.

A somewhat easier way to wire a diode stack is to use a bridge rectifier. The diagram shows an application with a high current bridge, but any cheap 1 amp bridge will work with the Tsunami. Use one bridge rectifier for each 1.4 volts of desired drop.

My parts box yielded some 1N4005 (400 volt) diodes that I got as surplus somewhere for about a penny each. The leads were cleaned and bent and the diodes soldered together in a compact stack. The soldering looks pretty bad because these diodes probably failed a solderability test, they didn't wet well even after being scraped clean.

The stack was enclosed in a piece of shrink tube and inserted in series with a track lead.

The tendency of the decoder to issue an error did not go completely away, but it happened much less often. I saw it only once during my testing, I would have expected it to be doing about half the time like it used to. It is probably telling me that it want's lower voltage yet.

I discovered by accident (a broken solder joint) that this decoder, and probably most others, will actually run from a half wave rectified DCC signal, albiet slowly. If you want to cut your voltage in half, then a single diode will do the job.