Last night while trying to change the normal direction of travel setting on two Spectrum 2-8-0’s equipped with DH121 decoders, using Ops mode programming on the mainline, we experienced erratic behavior with both locos i.e. jerky start/stop running, with the PM42 indicating momentary short circuits in that power district.
Attempts to reprogram to factory default settings and 03 addresses, both on the mainline and on the isolated programming track, did not change the behavior of either loco and in both cases CV 29 value reverted to 07 and not 06, the Digitrax default setting.
A physical check of each loco and decoder showed nothing unusual. We have yet to see if the locos run ok on DC with the decoders removed and shorting plugs installed, also whether the decoders test ok by themselves.
Both locos had been running fine up until the attempted change in NDOT, but in reverse to the direction arrow appearing on the DT400 throttle.
The programming attempts were made while several other trains were running in the same power district or other power districts. The Command Station/Booster was the club DCS100 8 Amp unit. The person doing the programming was an experienced DCC operator, on his home layout.
Any comments would be appreciated. At the moment we have one very upset member, the owner of both loco’s) [:(!], and the perplexed and guilty feeling programmer [:(].
If they were originally running in reverse, it means the motor wires were backwards - if it was plugged in to an 8-pin plug then it was plugged in backwards.
With most recent bachmann, it is probably best to just remove the factory board and hard wire. There are some capacitors in parallel tot he motor (which you can simply cut off and remove) and some inductors in series with the motor (if you cut those, you need a jumper wire, else no power to the motor!). All that junk supposedly reduces RF emissions (since we all have TVs located right on the layotu and are trying to watch TV and run trains at the same time), but what they REALLY do is disrupt high frequency motor drives in DCC decoders, and cause problems with Back-EMF if you are using back-EMF decoders (the DH121 is not).
I do not think the DH121 implements the CV08 factory reset, it’s an older low-end decoder. This could be why you see CV29 ‘reverting’ - it’s never really getting reset anyway.
When it ran before, but in reverse of the arrows, did it operate fine in BOTH directions, or just one? If it previously ran OK forward (throttle arrow reverse) bt not reverse (throttle arrow forward) this could be a sign that the capacitors and inductors are interfering with the motor drive - switching CV29 to 7 just swapped the ‘bad’ direction.
Thanks for the comments Randy. Yours is the only response I’ve had, including a posting on the Digitrax Yahoo site.
Yes, the locos originally ran fine in both directions before the attempt at Ops mode reprogramming. We subsequently found that the decoder plugs were installed backwards and have since found that the two locos run normally, having replaced the decoders with shorting plugs.
Are the ferrite beads on the leads between the tender and the loco the inductors you refer to?
Yes, loops of wire around a ferrite bead will act as inductors. This can interfere with the supersonic motor drive. Since they’re in series with the motor leads, they have to be bypassed, not just removed. Actually,t he DH121 is not a Supersonic decoder - it might not be as affected by thepresence of inductors and capacitors. On other Bachmann locos there are capacistors and inductors on the actual circuit board that contains the DCC plug. And also possibly a capacitor directly soldered to the motor terminals. Interesting that my PCM T-1 has such a capacitor across the motor yet the Loksound decoder has both supersonic and back-emf and it works just fine - it IS possible but the capacitor value must be chosen with the drive frequency in mind - I think the likelyhood of the Bachmann capacitor being correct is minimal
You may be correct about Ops Mode - the DH121 manual says nothing about supporting Ops Mode programming, but newer decoders specifically mention this fact. When the decoder does not support Ops Mode and you attempt to program that way - who knows what could happen. It SHOULD just ignore the program packets.
Check for additional capacitors and inductors on the circuit board, and definitely try programming on the program track, not Ops Mode. Try Paged mode ont he programming track (Pg). In the Big Book, which is fromt he saem time as the DH121, it specifically mentions Paged mode being the recommend mode for Digitrax decoders. I do not believe this deocder supports the factory reset option, so you’re going to have to work through all the CV’s and fix them individually.
Using the DB150 and the isolated programming track, sanity was restored to the two loco decoders by reverting them to their factory default settings (CV08 set to CV value 08) and then reprogramming their assigned road addresses. The locos work fine now.
Another interesting incident happened while we were working on the two 2-8-0 locos. A (diesel) loco fitted with a DH121 decoder was running on the mainline when it suddenly stopped, reversed direction and ran off at a constant speed. Attempts to control it failed until track power was cut. On restoring power the loco did not respond and was found to have no address. Following the same routine reprogramming as for the two previous locos, the diesel loco operated normally.
Since there were no other locos operating on the layout and as the operator had his hands off the throttle at the time the diesel loco misbehaved, we again suspect we’ve got a system problem somewhere. The investigation continues.
Oooops! The hiccups started again with each of the two steam locos when another loco was operated in the same power district. The hiccups continued afterwards, even when the hiccuping loco was the only one in the power district. To compound the mystery we found that the locos ran ok on one module. We haven’t yet checked to see if there are any other modules that are the same.
As a correction to my earlier posting, the locos are fitted with DH123’s not DH121’s, if that makes any difference. Also in addition to reprogramming the locos, we had reversed the d
Yeah the DH123 has more features than the DH121, and I’m pretty sure supports Ops Mode programming, so that didn’t kill them. And a DH123 should have the CV08=08 reset feature. Do the trouble locos work ok so long as you don;t run an additional loco into their power district? And is it the same ‘other’ loco that always causes these to go haywire? Or will they freak out if ANY other loco runs near them? If it’s just a certain loco - check that one, it may have some wierd problem that is corrupting the DCC signal. Have you disabled analog conversion in the problem locos? Sometimes that helpd wierd issues. Otherwise, proceed witht he wiring check, you can’t have any crossed bus wires or you’d have a short and nothign would run, but if the wire length is particularly long (50 feet or more) you may need to terminate the DCC signal, reflections from the end of the wire could be corrupting the data.
Continuing the story; we confirmed that the tracks and feeders were correctly wired and cleared the DCS 's memory. We also disconnected several power bus conductor pairs that were not being used on several modules, that could have acted as antennas.
Having done that we found that the two problem 2-8-0’s ran well when together in two power districts, but continued to have erratically when together in the three other power districts. Substituting new decoders made no difference. We found that erratic behavior was reduced, but not eliminated,by using N gauge voltage setting rather than HO, also that the two locos would occassionally run well together but if stopped and restarted would resume hiccuping.
As time ran out yesterday we took a third 2-8-0 that had not had any problems and substituted a tender from one of the two problem locos, and what do you know - the third loco started hiccuping!
We did not have time to run a problem loco with the
Very strange. I haven’t see one of these locos so I don’t know exactly how they are wired but I was under the impression that all wiring was in the tender of them. Perhaps not. Bachmann locos are know for various capacitors and inductors (coils) in the motor circuit which often confuses DCC decoder drives, but shouldn’t interfere witht he track. It’s possible it might scamble the decoder. The capacitors you cna simple cut off, they are parallel to the motor. The inductors are in series and must be replaced with a jumper wire or there will be no power to the motor.
What makes me think it’s not the locos is the way they work just fine in one power district but not others. What’s different about the one where they do work? Shorter wire run? Do the bus leads to the problem districts cross over other power leads, particularly any AC power circuits?
The printed circuit board assemblies and wiring in the two problem locos appear to be the same. While perhaps not significant this loco has a DH121 whereas the two problem locos have DH123’s installed.
The two power districts where the two problem children behave have longer power bus and loconet runs than the districts where the problems occur. The track voltages as measured with a Sperry DM350A Multimeter do not vary much between modules e.g. 8.4 to 8.6 volts on AC setting with the command station output on N setting. Unfortunately we don’t have an oscilliscope handy to see what’s really there.
The long parallel runs of the track bus with 11ov AC lines may be a clue - over the longer run to the problem districts you may be gettign too much induced voltage from the 110v lines. If possible, try to seperate these by the width of the module at least - say the 110v in the back and the track bus in the front. If they need to cross - make them cross at 90 degrees. The Loconet cable probably shouldn’t be bundled with either the track bus or the 110v lines, either.
The Pricom DCC tester is a pricey unit, but probably well worth investing in for a club situation with a large and complex layout, or a modular group that sets up large displays. Easier to use than an oscilloscope (well, maybe not for some of us), and it will certainly be able to tell if there are signal problem on the track in these troublesome locations.
It appears that we’ve finally solved the problem of the hiccupping 2-8-0’s. On the theory that these two easy starting locomotives (plus a third 2-8-0 since discovered to have similar characteristics) pull a high but very short peak starting current (short as the back emf kicks in) we bumped up the short circuit setting on each of the PM42 power district breakers from 3 amps to 4.5 amps, without changing the sensitivity (reaction time) settings, also using the N voltage setting instead of the HO setting. All is well now, and hopefully will remain so.
[:)][:)]