October 24, 2010

That lonesome whistle wail

Today's mail delivered a package from Tony's Train Exchange with a QSI Revolution U sound decoder and three different speakers. I selected the QSI decoder because of its higher current capacity rating compared to the Soundtraxx decoders. The three speakers are different sizes ranging from 0.75 to 1.06 inches because I didn't know which ones would fit. I ended up using the biggest speaker, the QSI High Bass 1.06 speaker with its snap-on enclosure. This speaker barely fit in the firebox.  I had to use my motor tool to trim one of the walls of the firebox so that the fit wouldn't be too snug. I was concerned that if the speaker driver cone was too tight, the bass frequencies would be degraded.

Before installing the decoder, I tested the engine. The lead driver on the engineer's side was showing a short with a resistance of 130 Kilo ohms. The back driver was not shorting. This would cause a current leak of about 1 mili-amp at 12 volts, about what you might see if I were to use resistors on the wheel sets for track occupancy detection. I tested the loco on the tracks and the short did not affect operations. Nonetheless, I tried to find the source and could not. I have no idea what could be causing the problem. I ended up disassembling the loco in the failed attempt to debug it as you can see in the first photo.  I even disconnected the drive rods to see if they were the problem.

I added pickups on the tender.  In the process I accidentally popped the drive shaft from the universal joint between the motor and one of the gear towers. That was a royal pain to get back installed. I basically had to disassemble the motor from the tender to get enough slack to pop the shaft back in place. Once I had everything back in order and the motor leads picking on both sides of the tender, I tested the tender by itself on analog (DC) power and it ran without the engine. That was good.

As I did in the other locomotive, I disconnected the wire from the pilot truck as it was causing trouble with binding. While I had the pilot truck off, I added the same number of paper washers as I did to the Haupt to raise the nose of the engine to provide clearance for the pilot truck. I also added the insulating ring on the front of the cylinder just in case.

With the decoder and speaker in the boiler and the motor in the tender, as well as all wheel pick-up on the tender, I had to make two wire harnesses. One to bring track power from the tender to the decoder. Then another to send DCC regulated power back to the motor. I used the second plug that comes with the loco lighting wires to make the return wire harness to the motor. It was an easy task to disassemble the micro-plugs and reassemble them with the wires I needed. I used heat shrink insulation to protect all electrical joints. Thus the loco now has two plugs between the tender and the engine.

One curious aspect of these locos is that the frame is hot on the engineer side on the tender, but is reversed for the engine. That is, the frame is connected to the fireman's side on the engine and to the engineer side on the tender.  Once I figured that out, installing the decoder was fairly straight forward. I tested it before shoving everything into the boiler.  It worked! I was very happy.

A photo of the installation 
Then disaster. As I shoved everything in the engine, the wires to the decoder broke at the plug-in harness at the decoder. The tight fit of the speaker but too much tension on the wires and they broke.  There was no way to re-solder the leads back to the plug as the joint is encased in the plastic plug and is a very tight package. Arrrrrgg!

I examined the wires. Only the red and black decoder leads broke. I decided to remove the plug and solder the leads direct to the board.

Bad move! The plug was not a friction fit, but actually hard soldered in place. I could not tell this from the outside because the solder joints were under the plug. It must have been assembled in a wave solder machine because you can not possibly get a soldering iron under the plug. In the process of removing the plug I damaged the printed circuit pads on the board. The pads for the head light leads were gone, but enough of the pads for the other wires were present. This allowed me to painstakingly re-solder the decoder leads directly to the board. It wasn't easy, but I got it to work. I had to use the F5 lead (number boards) to provide a signal to the headlight because the white wire pad for the headlight was totally gone. That explains why it flickers in the video. Hopefully I can adjust that with a CV value.  I tired a 220 ohm resistor to drop the light brightness, but that made it too dim. So the light is running straight off the decoder leads with no drop down resistor.

Once I got everything back together, I tested it and it worked. I was both amazed and pleased, as I thought that I had ruined this decoder.

The following video shows a quick test of the loco as it rounds the tight curve behind the engine terminal at Falmouth.  The speaker sounds pretty good.  Now I need to read the various manuals to customize the sound of the engine. The air pumps must go, but the whistle is very good.

I think you'll agree - the sound adds a whole new level of life to the layout.

With all the disassembly I have done on these locos now, I am beginning to understand how they work. This is making me more confident that I will be able to scratch build my own locos when the time comes.  I would like to build a yard engine with link and pin couplers on both ends. I'd also like to try building some of the engines that ran on the USMRR in Alexandria, such as the Devereaux and the Virginia. They would go nicely with the planned model of the Alexandria roundhouse I want to do too.


  1. I like the flickering, makes sense for an oil light.

    How do you get the exhaust sound synchronized with the drivers?

  2. So far I haven't found a way to control my ersatz headlight with a CV. The decoder assumes that number boards are always on when the engine has power. I would like to turn the light off, as locos in my period did not use headlights in the day. Since I broke the normal headlight function, I may be SOL here.

    The sound of the exhaust is only synchronized via the voltage level. I did not add a cam. It seems to work pretty well at slow speed. The decoder has several different modes that it can operate including a "Sound of Power" mode which simulates momentum in the sound effects. I haven't really tried playing with all the sound settings. Truth is, I am pretty happy with the out of the box sound and have higher priority tasks than sound programming at this point. So I'll probably leave it at that for now.

  3. Fantastic. It is extremely authentic... and all this from one visit to the Golden Spike Museum, huh? ;o)

    I agree that the sound does add a layer of realism to it and helps fuel the imagination.
    Thanks for the short video!