Odd Speed Fluctuations: Decoder or Drive?

[tehachapifan]

I have a LokSound Select Micro istalled in an Atlas N GP9 (DCC ready drive without universals) and have been having some odd behavior with either the decoder or the drive, mostly with regards to speed fluctuations. I have been able reduce this with CV adjustments but haven't been able to completely eliminate them. The other thing that will occasionally happen is the loco will not start moving from a stop or will suddenly halt as if something was binding (during these stalls, the headlight and effects LED's will display a back EMF-type flicker). What will then happen is the loco seems like its building up pressure and will then suddenly release and lurch at a higher speed and then come back down to the correct speed. I've disassembled the loco several times looking for a source of binding in the trucks, gears, worm assemblies, etc., and can't find anything at all. I've even since "Beardenzed" the Worm assemblies with no change either and have also tried different levels and type of lubrication.

The speed fluctuations will be only slight and random in most speed steps except two, where the fluctuations will be very pronounced and rhythmic in both directions. Before I adjusted some CV's, this pronounced and rhythmic fluctuation in speed was happening at speed step 8 but was way smoother in speed steps 7 and 9. It also would happen up around speed step 20 or so (also way smoother on either side). I tried taking the trucks off to test the motor assembly itself and would notice I could still hear the rhythmic fluctuation (like a "weeeooooweeeooooweeeoooo") again a speed step 8 and around 20 but could not see any wobbling with the flywheels or anything else unusual (the motor also appeared to spin freely). I've tried turning off BEMF which didn't help and I tried adjusting CV's 52-55, which only seemed to mostly move the rhythmic fluctuation to a different speed step.

Here's something else that's odd....I found some info online that recommended changing CV52=127, CV53=160, CV54=127 and CV55=127 which then caused the loco to shake pretty violently in speed step 1 (no shaking in previous settings). I also tried doing the automatic adjustment feature for these CV's, which didn't seem to eliminate the fluctuations either. Right now I have these CV's set to CV52=30, CV53=77, CV54=60, CV55=30 and CV56=180 (fairly close to the auto settings), which has the loco running pretty smooth at the moment but not without some fluctuation (I say at the moment as it seems to be very moody and tends to behave a little differently every time I go run it).

What I don't know for sure is how the loco ran prior to the decoder install and I'm not really wanting to uninstall the decoder to test it at this point, if someone might have some experience or insight with this.

Thanks!

[Burlington Bob]

Have you tried running on plain old DC?  Tried a decoder reset?

[tehachapifan]

Have you tried running on plain old DC?  Tried a decoder reset?

Its been a while, but I believe I tried a reset. I don't want to do that now unless absolutely necessary as it took me quite a while to get the lighting effects set up. I don't think it would help, but...

No, I have not tried running it on DC. Certainly worth a try to see what happens.

Thanks for the input!

[peteski]

Its been a while, but I believe I tried a reset. I don't want to do that now unless absolutely necessary as it took me quite a while to get the lighting effects set up. I don't think it would help, but...

No, I have not tried running it on DC. Certainly worth a try to see what happens.

Thanks for the input!

Just to throw in my 2 cents, I suspect that when Bob said "on DC" he didn't mean to just run the decoder-equipped loco on DC power. It means taking the decoder totally out of the motor circuit to verify if the problem is mechanical in nature. At least that is how I would understand that.

[tehachapifan]

Right, that's how I took it too. I've been pondering this some to try to figure out how this could be accomplished without actually taking the decoder out of the equation (as in removing it). If I run it on DC, which the decoder will allow me to do with a certain CV setting, can I truly rule out the decoder? At this point, I'm not going to remove the decoder as it is a hard-wired installation.

[peteski]

If you run a decoder-equipped loco on DC, the decoder is still fullu engaged in the circuit.  The microcontrolller works form about 3V and up, and the motor still gets its voltage through the 4-transistor H-bridge which is controlled by the microcontroller. Some decoders even engage the PWM circuit to send pulses to the motor (like they do in DCC mode).

To temporarily convert the model to DC, I would leave the decoder installed but unsolder the orange and gray wires from the decoder (or motor), then run a wires from the motor terminals to the track pickups.  That way the decode and motor will be powered from DC but the decoder will not drive the motor.

[Burlington Bob]

To temporarily convert the model to DC, I would leave the decoder installed but unsolder the orange and gray wires from the decoder (or motor), then run a wires from the motor terminals to the track pickups.  That way the decode and motor will be powered from DC but the decoder will not drive the motor.

Exactly!  Only for sure way to rule out a mechanical problem or vice versa.

[tehachapifan]

While I haven't quite gone to the point of doing any de-soldering of the decoder (yet), I am thinking more and more that the motor may be bad. As mentioned earlier, every so often the loco won't start from a stop as if it may have a bad pole. Also, at speed step 1, I can see occasional bright arcs through the gray motor casing that surrounds the commutator area. If that is not all that uncommon, it's the first time I've ever noticed it occurring.

[tehachapifan]

Ok, today the unit was actually running pretty good when it suddenly would no longer start from a stop at all and the lights were all doing that back EMF-type flicker. Got fed-up enough to go ahead and pull out the decoder, reinstall the analog board run it on DC. There, it would start from a stop fine (couldn't get it to not start) and the motor actually ran quite a bit smoother, but not without a slight speed oscillation and a somewhat more than normal speed cog at extremely slow speed. Still hoping it was the motor causing the issues (they're way cheaper!), I swapped it out with another one on-hand, reinstalled the decoder and now the unit runs flawlessly. So, unless a problem develops, it would appear that the motor was indeed bad (bad pole?) and the behavior was possibly being magnified by the decoder somehow. Perhaps a bad pole might wreak havoc with the back-EMF-related features of the decoder...even though I turned back-EMF off at one point with very poor results? At any rate, glad to have this unit running well and keeping fingers crossed that nothing crops up!

[RBrodzinsky]

Russ -- my guess is that one of the poles had a broken wire. I had that, on a Kato Mike, and it drove me crazy until I figured out what was going on.  Depending on where the motor would stop, the wire was potentially contacting the motor frame and shorting

Glad you figured it out.

[peteski]

Bad Pole, eh? I resemble that remark!   

If you want to play detective you could take the motor apart and test it for open winding by measuring the resistance between commutator segments.

[tehachapifan]

Russ -- my guess is that one of the poles had a broken wire. I had that, on a Kato Mike, and it drove me crazy until I figured out what was going on.  Depending on where the motor would stop, the wire was potentially contacting the motor frame and shorting

Glad you figured it out.

I bet the broken wire is exactly what was happening! That might also explain the occasional bright arcs I could see illuminate thru the motor casing.

Bad Pole, eh? I resemble that remark!   

If you want to play detective you could take the motor apart and test it for open winding by measuring the resistance between commutator segments.

I think I may take the motor apart with one of those Feather River hammers!

Thanks for all the help on this, guys!

[central.vermont]

One of these Russ. 

[tehachapifan]

Yep, that's the therapeutic device I speak of!

[mmagliaro]

You probably don't care at this point, but if you want to try to save the motor,  the problem could also be bad carbon build up in the commutator slots.  You probably can't see the slots well without pulling the whole armature out of the motor, but if you are game, it might be worth doing that so you can see them and clean them out.