Understanding Motor Behavior Changes After Decoder Install

[tehachapifan]

Just installed a Zimo MX621 decoder in my SW1500 kitbash with a Con-Cor/Kato SW1500/1200 drive. These old drives have 5 pole motors but no flywheels. The drive I used for this used to run fairly smooth on DC but it would constantly fluctuate speed, probably having a lot to do with the lack of flywheel(s). This would occur on flat track but it would, of course, also speed up a lot on a downhill run. I installed the Zimo decoder and tried it out, leaving all the default settings as they came. Now the drive runs ultra-even speed-wise and doesn't even speed up at all on a downhill run (hey, pretty cool!) but now the loco vibrates quite a bit as it rolls along. The loco makes a different electronic-type sound too, which goes with the vibration if that makes sense. I would say it is running way better with the decoder overall but the visible vibrations are a bit distracting. The manual has pages and pages of info that I've started to go thru to better understand all the various settings and adjustments, but man is it in-depth and complicated! Figure this might be a back EMF thing but really don't know. Anyone experience something like this?

[jdcolombo]

Hi Russ.

What do you mean, precisely, by "vibration"?  Do you mean that the locomotive runs unevenly (e.g., you notice momentary surges in speed)?  If it is this momentary "surging," at what speed does it happen the most?  Low speed?  Medium? When you crack the throttle open or when it is running at 20% throttle?  1/3?  Does it eventually smooth out or do you notice it all the time no matter what the speed/throttle setting?  Or do you mean that it literally rocks sideways on the track as it is rolling along?

The Zimo has one of the most sophisticated BEMF circuits of any decoder available, and has so many adjustment options that it's hard to even know where to begin without some additional information.   Perhaps even a short video taken with your phone and posted to YouTube or some sharing service that would allow us to see what you're talking about.

Another approach would be to just start playing around with the various BEMF CV's (at least, the ones most likely to make a difference).  These are CV's 9 and 56.   The Zimo uses two digits in each of these CV's to control the BEMF circuits, and the 10's digit controls a different aspect than the 1's digit.   So, the default for CV9 is 55.  If your "vibration" is a jerky movement, you might try increasing both the 10's and 1's digits by 1 to see if that helps.  For example, try 85, instead of 55 in CV9.  In theory, the higher the 10's digit, the less "jerky" the movement will be.  If 85 improves things, try 65, 75 and 95 to see what seems to work best.  Simlarly, higher 1's digits tend to work better with older motor designs.  Once you have the 10's digit that works best, try increasing the 1's digit by 1 to see if you can get further improvement.  Zimo actually recommends a setting of 95 in CV9 for a "normal" N scale motor (not a coreless).

CV 56 sets the "I" and "P" parameters for the BEMF circuit (you don't want to know).  Once again, the 10's digit and the 1's digit affect different parameters.   The default is 55.  Try 33.  Then try 77.  Which is better/worse?  You'll do iteration here (if 77 is better, what about 87?  What about 88? 97? If 33 is better, what about 23?  22? etc.).   This is time consuming (and best done via ops mode programming - programming on the main - rather than on a programming track) but it's really the only way to know for sure that you are getting the best possible performance from this decoder.

And remember that if things go crazy, you can always just reprogram with the default 55 to get back to where you were.  Don't be afraid.  You aren't going to ruin anything by doing this.

John C.
 

[Ken Rice]

My guess would be you're seeing the EMF sampling pulses.  You can confirm that by tweaking CV9, which defaults to 55.  Try something like 22 and see if the vibration you're seeing diminishes (but don't be surprised if motor control also gets worse).  And if you set it to something like 77 or 88 it should get a lot more noticeable.  It may be that 55 is the best value, but playing with it would confirm whether that's what you're seeing.

The zimo manual (http://www.zimo.at/web2010/documents/MX-KleineDecoder_E.pdf) page 16 and 17 has some pretty good tips on how to approach tweaking the motor control CVs.  Having done that for some reluctant performers I recommend keeping notes on the various combinations you've tried and whether each combo got better or worse (and what it was better or worse than).  The defaults usually work pretty good for decent motors and drivetrains, but if you've got an older motor and, say, little or no flywheel or a lousy gear ratio or both you may get significant benefit from careful tweaking.

[reinhardtjh]

I'm not very familiar with the Zimo decoders. I've just looked over the manuals from time to time, but if it's an actual vibration you see and hear then maybe it's the number of pulses that the decoder uses.  Some decoders have an option to switch between 20khz and 40khz pulses depending on the type of motor. I'm sure Zimo has something similar.  Maybe a different setting might help.

[tehachapifan]

Thanks, guys! I'll play around with these settings and see what happens.

This isn't the greatest drive to start with, so I doubt I can achieve perfection.

John, the vibration is literal vibration (rapid yet subtle shaking) of the loco. No speed fluctuations at all.

[tehachapifan]

I did some experimenting and found that adjusting CV56 made the biggest improvement. I changed it to 22 and now the loco is running nice and smooth, at least as good as this particular chassis will probably do. Thanks!