Since it was Friday, I decided to play with consisting my two Atlas GP40-2s with QSI sound. I have had these two locos for about 2 years and have had no trouble with either of them. Normally, when they are selected on the throttle they respond with an air let-off and then will start up with a double press of F6. Not today. After many frustrating minutes I resorted to doing factory resets on both - the magic wand worked for one and the OPs mode reset the other. After reset both responded normally and I was able to restore the road numbers and begin running again. I have seen this phenomenon mentioned in previous posts and my question is has anyone figured out why this (seemingly random) event happens? All of my locos are stored on individual tracks that are electrically isolated (DPDT switch) from the layout when not in use and the track is only powered up after the DCC system (Zephyr) has been turned on and has stabilized. Are there any theories out there?
Joe
David:
Tried the addresses given them previously. I keep them on a card just in case I have a senior moment.
They were never put into a consist. I was going to try to familiarize myself with consisting when I found the locos to be unresponsive.
I have run them as recently as a week ago with no problems and I always use the F6 double press to start them.
Sorry I wasn't more clear about "comatose". Neither loco would respond to multiple attempts to start up with F6, Turning the DCC system off and back on and making sure that they were receiving power. I didn't try reading back CV 17, 18, 19 and 29. If this happens again I'll try that. However, if these were set properly and I was able to run the locos without a problem previously what could cause an undirected change in a CV value? Perhaps that is the answer I'm looking for?
Joe, the fact that a reset solved the problem strongly points towards a scrambling/corruption of the memory portion of the locomotive. I have had decoders do this on occasion, not frequently, and frankly I can't recall if this has happened to a QSI decoder on my layout or not? I just put it down to an electrical anomaly, perhaps associated with static, electrical storm etc. Frankly I don't know. But for sure you are not the only one that has had a decoder go haywire for no apparent reason.
Simon Modelling CB&Q and Wabash See my slowly evolving layout on my picturetrail site http://www.picturetrail.com/simontrains and our videos at http://www.youtube.com/user/MrCrispybake?feature=mhum
Simon:
Thanks for confirming that this was not an isolated event. Interestingly, one loco (about a year older than the other) wouldn't reset with the magnetic magic wand but did with the OPS mode reset sequence. However, when it reset it retained its assigned number rather than resetting to address 3. The other loco reset easily with the wand and reset to address 3. Obviously there is some type of partial memory loss going on. If it happens again I will try David's suggestion and see if I can read the CVs he suggested. It will be interesting to see what is getting scrambled.
Not sure if you are a Decoder Pro user Joe, but this is to me one of the big advantages of the program, especially if you do a bit of customization. By saving all the settings for a particular loco, you can then reset all the CV's back to what you had them at, without having to remember anything.
Yes Simon I use Decoder Pro regularly and really like it. I previously had trouble reading the CVs of my Tsunami and QSI equipped locos on my program track ( I have an older Zephyr). I recently acquired a SoundTraxx PTB 100 program track booster and it works beautifully with my system. I have been able to read all of my locos CVs into Decoder Pro and they are all saved for future use. I just hope that I won't have the need to re-vitalize a dead decoder in the future.
This happened to an engine of mine, a Proto 0-6-0 with factory QSI sound. Like you, I tried all sorts of resets, both programming track and magic wand, but I had no luck. (I did not try Ops mode, and I probably should have.)
When the problem occurred, I was testing a crossing track which I later found to be defective. The piece, an Atlas 30-degree crossing, was shorting when trains went over it, and one of these shorts seems to be what put this engine into a coma.
My LHS was able to restore my engine on the QSI programmer in the shop. Even that was a slow process, but the engine has been fine ever since. (Hint: support your LHS.)
It takes an iron man to play with a toy iron horse.
Mr. B:
I can rationalize a short causing a decoder scramble but my two locos were sitting on electrically isolated tracks where they had been parked a few days previously. We had no electrical storms in the area and I had done no work of any type on the layout. I just went down and started up the layout as I had done many times before and when I went to run these two engines they were dead. Interestingly my Tsunami equipped GP15-1 was OK and ran fine as well as all of my other non-sound locos. It's a real puzzle to me and I don't recall being transported to the mother ship .
The shorting situation on my locomotive might have been coincidental. This topic shows up from time to time, and in most cases it seems to be a rare event. Mine, in fact, is the only one that I can think of where the problem occurred other than at power-up.
You didn't do anything wrong or unusual, and neither did I. This is a known but poorly understood problem with some QSI decoders. We've both managed to restore ours to normal running, so we can live happily ever after.