Haunted Locomotive?

[jacon12]

I had something happen tonight that I've not seen yet. 'Course, I'm so new at all this everything is a first. While waiting for the glue to dry I decided I'd set up an oval of track on a 4x8 plywood laid on sawhorses and run a train around. (this is another reason I don't get things done!)
Anyway, when the train got to one of the turnouts, the dcc locomotive with sound, got about halfway across the turnout it just went dead. No sound, no movement... nothing. I figured a short of some type so turned off the power to the track via the power off button on the digitrax zephyr. I also turned the power dial all the way back, got the loco off the turnout and set it back down on the track. It started to move forward fairly fast and I looked down at the dial on the control box and it was set to zero. I unplugged the power cord from the zephyr and of course the loco then stopped.
How the heck did that happen? How did the locmotive still go forward when the dial was turned all the way down?
Jarrell

[grandeman]

I'm no DCC expert but I have seen a few runaway trains. My guess is that you have some dirty track and when the decoder powered back up in "remembered" it's last instructions. Then again, that could be way off base... Anyone else?

[selector]

I suspect that your turnout frog is dead (isolated) and the loco lost power over it and went dead before the pickups could get power once again, on the far side. The decoder might have surmised that it was a power bump and recalled the last instruction when it was repowered.

Locos with a few, or widely separated pickups, or those with that deficiency AND no flywheel will stall dead immediately when they encounter dead frogs. (I just step over 'em, personally, but the dog likes to sniff, you know...[:D]).

[jacon12]

I should have mentioned the locomotive was a P2K SD-40, not that it may matter at all.
Jarrell

[rrinker]

What selector and grande man said is EXACTLY what happened.

The loco lost power while still being commanded to move at a certain speed.

You pulled the power from the Zephyr - meaning it no longer had that loco selected for direct control.

You put the loco back on the track and turned on track power - the loco immediately took off at the last commanded speed.

You don't have enough track in place to have done this, but if you had left the speed dial at 0 and called up that loco address, it would have stopped as soon as commands went out to that address with speed 0. This is why it's always important to make sure the speed control is at stop before selecting another loco or shutting things down. It can get really tricky if you have the control ALMOST to stop so at first glance it looks like the loco IS stopped. But in realtity it's getting speed step 1 or 2 commands, and it's once of those better locos that really can run at a fraction of a volt - one tie ever 10 minutes or so. Half an hour later, the loco you THOUGHT you stopped is now on the other end of the siding you left it on.

A little more explanstion: The Zephyr has 10 'slots' which means it can remember the setting and send commands for up to 10 locos at a time. But obnly one of those can be controlled by the knob at any given time. The other 9 keep getting sent the speed, direction, and function settings from their slots. Unplugging the Zephyr does not necessarily clear all the slots - but it does clear the selected loco. Selecting a different address to be the one controlled by the knob does not automatically stop the previous one - it continues at the speed and direction it was last set to, because that it what is in its slot.

--Randy

[ereimer]

as a reminder ('cause i'm sure you already know this) with DC power there is only voltage on the rails when control knob is moved above stop . with DCC there is always voltage on the rails and when you move the control knob the decoder in the loco sends some voltage to the motor . it's probably good to remember this when working on track etc.

[jhuntjr]

I have a BL SD40-2 that did this at the club the other night. One of the switches was dead so I pulled the loco back to live track. Of course it took off right away. When it hit the switch, it died and I pulled it off the track. I put in on a section of track beyond the bad switch and it responded to the throttle but once it started moving I could not gain control of it or stop it except with the old 5-0 manual brake. If rrinker is right (and from the comments I have read he is probably the reigning king of info [:)] ) how do I reset the loco? I was just about to put it on my programming track at home (using DecoderPro w/Digitrax super chief) and try to figure something out but this does not seem to be the solution. Any ideas out there.

[jacon12]

Hmmm... jhuntjr, if I'm reading all this correctly reseting to factory defaults and reprogramming won't do any good, it's just a fact of dcc. Most of the time it happened to me exactly the way it did with you, on a dead switch.
Jarrell

[Anonymous]

I've had one interesting one with a Bachmann Class 37 (British diesel loco) - fitted a Lenz decoder (my very first installation), programmed it and all was well until I did the following things on a section of the layout:

turned lights off
turned lights back on (they didn't come on)

discovered that loco was now happily trundling along with no control whatsoever - was moving (so obviously there was track power) but I couldn't stop it without hitting the emergency button. It's only happened a few times and not recently, I have no idea what caused it. I removed all the capacitors around the motor when I fitted the decoder as I'd read that they could cause this sort of problem, and none of my other locos (all using Lenz decoders) have this problem. The loco runs superbly everywhere else on the layout - just seems to have a mind of it's own here!

[jacon12]

It did it again with me tonight. Loco 1 hits a bad switch, causes short.. turn everything OFF, pick up loco 1 and loco 2 takes off like a bat outta Hades. Hmmm... maybe if I completely unplug the Zephyr from the electrical outlet THEN pick up loco one, then plug it back in.. hmmmmm...
Maybe Digitrax knows the answer..[banghead]
I dunno..
Jarrell

[ericboone]

There was a person in the club I was in that had some Lenz decoders that acted funny. When another locomotive went by the locomotive with the weird decoder, that locomotive would just take off. It was kind of weird. I think Lenz knew about the problem and was woring on it.

[rrinker]

Definitely make sure you truly turn the speed control all the way down, and remember to dispatch the loco before selecting a new one. And definitely change CV29 to disable the analog conversion. THis will help greatly.
I think I had the symptons you described just once, I have since started disabling the analog conversion and had no repeats. Kind of disturbing because now I have a Locobuffer and my laptop set up that I could monitor the commands being sent out at times like that in an attempt to see just what is happening - but I suspect it's a lack of DCC signal at the restoration of power from the short, making the decoders think they are on regular DC track, and operating accordingly.

--Randy

[jhuntjr]

Well, I finally determined that the loco is responding but it is as if momentum got turned on. It will slow down gradually and will not respond to anyother command until it has come to a complete stop. I cannot find any info in the BLI docs that came with the loco about reprogramming this. I have tried resetting everything back to factory defaults to no avail. Funny, because it starts moving normally but takes forever to stop. Anybody have any ideas?

[selector]

If it is a QSI loco, you have two CVs for momentum, one for acceleration, and the other for deceleration. I think they are 3 and 4 respectively. I have my Hudson programmed with a value of about 70 in each, and it works marvelously...except that I have to dial up to 7 to get it to move on my DT400, and even then, it takes its sweet time turning a wheel...often ten or more seconds. I probably need to set the V-start down a notch or two. Randy, does that sound right?

[rrinker]

If it truly doesn't start moving until speed step 7, then yes, bump up V_Start a little bit. Or is it just the momentum delaying the response? That's pretty high, but it should give a nice effect. I set an Atlas Trainmaster up with a huge momentum value and it works great, crank the throttle with a big train behind it and the motor revs up and it chugs like crazy getting the tran started, and then as the speed levels out it quiets down, sounds GREAT doing that.
With these gradual slowdowns to stop on a particular section of track - is the track wiring OK? This sounds a lot like "DC stopping section" behavior, where a stretch of track gets DC power not the DCC signal, and the loco will stop when it encounters this section. Without a way to revert that section to normal DCC power, there would be no way to get the loco going again once it stopped in such a situation.

--Randy

[jhuntjr]

Well, I tried to reset the decoder to factory defaults with no effect. Finally had to pull the shell and reset the thing using the procedure for removing the jumper. Now the loco operates just fine.

[selector]

Well, at least THAT worked! Last time I had the need to do a reset, even my hard reset wouldn't take. I was dumbfounded. I turned off the lights and went to bed, fuming...amd worried that I would be sending it back to BLI. On a hunch, the next morning, I put it back on the track, powered it up, and it was all set up like it was before the problems, multiple 'resets' not withstanding. Has run like a charm ever since.

Were you programming on the main, in Ops mode, or in page with a programming track? There IS a difference. Some decoders won't accept programming certain CVs on the mains unless you have tons of amps.

Go figure.

[jacon12]

For some strange reason the Kato, a late model, that I converted to dcc..hasn't done this in the past couple of days. I have no earthly idea why. Maybe it just hasn't come across the right 'conditions'. We'll see.
Jarrell