Zimo MX644D Sound x Bachman loco(s) issues

[LittleWesternModelRailway]

Hi all,

A friend of mine is having some issues with his Bachmann LMS Jubilees, Compounds and Patriot. All are running the same MX644D decoder loaded with the same sound files, have recently been fully serviced and should be in great condition as they are well looked after.

 

The issue is between speed steps 20-45 there is noticeable 'hunting' even when no CV's have been changed we can visibly see each locomotive (some more than others) instead of being smooth seem to pickup speed 3/4 of a rotation than slow the last 1/4 and repeat. (next time I see him I will try to get some footage).

We have tried following methods for setting back EMF and such and although we can see changes to the loco(s) performance, mainly incredibly long start up times, the issue doesn't seem to go away.

This was noticed first on a Digitrax system and new and old track, stainless steel and nickel-silver a code 75 bullhead. So we tried using my DCC-Ex command station and witnessed exactly the same issues. It also doesn't matter if decoders are swapped between locos... it still occurs. so I am even less inclined to say its track or controller.. His Hornby Duchess runs and sounds much better and again is completely stock.

I'm beginning to wonder if the motors on all of them need a rigorous cleaning or replacements.... or maybe he just got bad decoders? I am truly stumped, having spend several hours on several days tweaking and checking everything I can see for changing settings for smooth running. and now I'm at a loss.

Has anyone else experience something similar? Has anyone got any more suggestions? 

Many thanks,

James - LWR

[Tim123]

Are there any suppressors / chokes / ferrite rods between the decoder output and the motor? I don't have examples of the same locos, but my other Bachmann ones have up to 3 suppressors soldered to the motor. These items could be interfering with the feedback that the decoder is sensing from the motor? Rightly or wrongly I always remove them.

[LittleWesternModelRailway]

Actually this was the first thing we thought of and removed them... Still an issues. But thanks :)

[Mike Buckner]

19 hours ago, LittleWesternModelRailway said:

It also doesn't matter if decoders are swapped between locos. 

 

instead of being smooth seem to pickup speed 3/4 of a rotation than slow the last 1/4 and repeat

 

You are willing to remove decoders - why not try putting in a blanking plate and running on straight DC.

A 3 volt or 6 volt battery would be an unimpeachable source of straight DC. See if that manifests an irregularity. If it does, then the problem might not lie with the decoders.

 

If you mean there is a regular sticking point with the same timing, once per each rotation of a driving wheel:

then this suggests there may be a mechanical issue with wheel(s),  axles,  valve gear or  quartering.

 

 

[LittleWesternModelRailway]

2 hours ago, Mike Buckner said:

 

You are willing to remove decoders - why not try putting in a blanking plate and running on straight DC.

A 3 volt or 6 volt battery would be an unimpeachable source of straight DC. See if that manifests an irregularity. If it does, then the problem might not lie with the decoders.

 

If you mean there is a regular sticking point with the same timing, once per each rotation of a driving wheel:

then this suggests there may be a mechanical issue with wheel(s),  axles,  valve gear or  quartering.

 

 

I believe he is trying with DC today to see if the problem replicates itself. So once I have more infomation I can report on it. As for the the mechanical side of things, I cannot say if this is the case but something to check in on. Still very strange it seems to be the 4 locos only though.

I will pass on the findings. Many thanks

[LittleWesternModelRailway]

So an update, and I am leaning towards either bad decoders or firmware:

Checked with a new OOTB Bachman Patriot
- DC running is perfectly fine at all speeds

- DCC shows the issue, on Digitrax and DCC-EX controllers

- We tried on Steel and Stainless Steel track and see the same issue (have no nickel-silver to test with)

- I personally disassembled each, checked B2Bs, quartering, degreased and re-oiled... everything checks out.

- Decoders are on stock settings and haven't had CV's touched - He is trying a StayAlive to see if it helps, but I'm not sure

- I see information about BackEMF and changing that etc... but I'm reluctant to try it because being stock.... It should work....

- He has a Hornby Princess with TTS decoder that does the same thing, but is a lot less noticeable

- My tank loco with different non-sound decoder ran as smooth as butter (but this hasn't been retested in maybe 2 months)
-  The capacitor has been removed as this was my first thought

- The Digitrax system is giving 19v to the track iirc and we checked and confirmed this at the time

 

I think I still need to spend some time and tweak back EMF but, it's frustrating out of the box it doesn't work.
I have suggested to contact Digitrains (source of the decoders) and see what they suggest.

[Izzy]

If it wasn’t for your remark that a Hornby Princess with a TTS decoder does basically the same thing I would have suggested that maybe the sound file being another common factor might be the issue. But if different decoders with different motor control firmware are all doing roughly the same thing to a greater or lesser degree then some other common factor is the problem it would appear. 
 

Bob

[LittleWesternModelRailway]

So, an update, I took one of my friend locomotives and decoders home with me to my layout and on 3 different types of track, same controller and power setup, the same issues occur.

After multiple suggestions I spent hours combing through BEMF CV's and settings but for the life of me couldn't see a difference, only hear the motor straining more and more, so I reset it.  I disassembled, cleaned regreased and checked the B2Bs myself, nothing seems to be snagging or otherwise. What's more is my pannier tank seems to be showing the same issue now after running again on my friends system. 

 

My system is DCC-EX and currently a 10V 2A PSU feeding the track, readings on voltmeter were consistent and no major resistance was noted.

Other than buying a New decoder and loco (I plan to buy an ESU Sound Decoder for my tank loco soon) and seeing if the issues persist with a different one, I am lost..... I will be contacting Zimo in Vienna (a hour's drive from us) to see if a firmware update might be a cause, or even if they can take a look and see if there is decoder faults. 

 

Could this simply be dirty or micro marks in the Steel DCC Concepts track? I've cleaned both layouts to the best of my ability it IPA.

Honestly, I don't want to give up but we are running out of ideas...

[Harlequin]

10V is a bit low...

Does your friend also use a (relatively) low voltage PSU?

 

[Nigelcliffe]

1 hour ago, LittleWesternModelRailway said:

.........

 

My system is DCC-EX and currently a 10V 2A PSU feeding the track, readings on voltmeter were consistent and no major resistance was noted.
....

 

As per Phil.    This would be ringing my alarm bells.   10v PSU means about 8v at the track through a DCC++EX setup, which is marginal on whether a DCC decoder will work reliably. 

 

At a minimum change to 12v.   
Ideally more than 15v,  which will require splitting the Arduino's power from the H-bridge power, as the Arduino will blow up at more than 12v.   One way is to put a regulator in after the 15v, to drop the voltage to about 8v, and use that for the Arduino,  or you need two PSUs, one for the H-bridge and one for the processor (as per the DCC++EX standard instructions). 

 

 

- Nigel

Edited April 21, 2023 by Nigelcliffe

[LittleWesternModelRailway]

2 hours ago, Harlequin said:

10V is a bit low...

Does your friend also use a (relatively) low voltage PSU?

 

So, the 10v 2A is just my test setup, my friend has (and I have too) measured his at 15v at the track. I'm going to increase mine soon as I'm still in the process of setting my system up. :) 

[LittleWesternModelRailway]

1 hour ago, Nigelcliffe said:

 

As per Phil.    This would be ringing my alarm bells.   10v PSU means about 8v at the track through a DCC++EX setup, which is marginal on whether a DCC decoder will work reliably. 

 

At a minimum change to 12v.   
Ideally more than 15v,  which will require splitting the Arduino's power from the H-bridge power, as the Arduino will blow up at more than 12v.   One way is to put a regulator in after the 15v, to drop the voltage to about 8v, and use that for the Arduino,  or you need two PSUs, one for the H-bridge and one for the processor (as per the DCC++EX standard instructions). 

 

 

- Nigel

 

So I will up the ante then in my next series of test at home, but wouldn't explain his issues while having 15v 5A on his layout.
(the arduino setup is still very much in progress and has alot of things to sort out before actually operational :D )

 

[LittleWesternModelRailway]

On 14/04/2023 at 08:36, Izzy said:

If it wasn’t for your remark that a Hornby Princess with a TTS decoder does basically the same thing I would have suggested that maybe the sound file being another common factor might be the issue. But if different decoders with different motor control firmware are all doing roughly the same thing to a greater or lesser degree then some other common factor is the problem it would appear. 
 

Bob

So, after a quick call with my friend, I was mistaken, the locomotive is Hornby but it has a MX645D again purchased some time ago. So it could still well be a sound file commonality