DCC Configuring an ESU Lokpilot for an ABC motor

[Dog Star]

We have an ABC diesel motor bogie with a Maxon 6w motor (16:1 overall gear ratio) under control of an ESU Lokpilot V4 decoder. As supplied, the decoder appears to offer acceptable control above speed step two / three whilst running at speed step one is poor with jerky movement. We have enabled auto calibration of motor control with the result that slow speed running is good and performance degrades as the speed is increased - above speed step eight (out of 28 speed steps) the movement is jerky (worse than with the default decoder settings).

We have tried manipulating CV53, 54 and 55 as described in the user manual for Lokpilot and no obvious improvement.

Here are the default values for the CVs which impact on motor control:-

* CV51 to Cv56 are 0 / 32 / 140 / 48 / 32 / 255;

After auto calibration (by setting CV54=0 and selecting F1):-

* CV51 to CV56 are 0 / 12 / 93 / 17 / 12 / 255.

How do the CV values which we have from auto-calibration compare to the experience of other WT-ers who might be using Maxon motors in ABC products?

Suggestions on how to interpret the behaviour of our model / decoder?

thank you, Graham

 

[Simon]

Blimey, that's a right proper technical question

I just wired my Maxon/ABC combo to my Loksound 3.5 and off I went, all is good I think.

That said, it may possibly take off a bit too "nippily" but in the great scheme of things........

All of which reminds me it's about time I ran another train out in the garden

Simon

 

[Steph Dale]

Graham,
In my experience the auto-calibrate function you describe is just a good starting point for further work using the routine in the manual. Very occasionally the calibration will result in a set of completely sorted values.
Worth noting as well that the calibration can be run more than once to narrow the result.
The problem you describe is related to the relative efficiency of the coreless motor, as far as I can tell all decoders optimise on conventional motors as they're rather more common.
Steph

 

[Dog Star]

I just wired my Maxon/ABC combo to my Loksound 3.5 and off I went, all is good I think.

Simon, we had a similar scenario when fitting a LokSound V3.5 into an engine with an ABC Maxon gearbox... whilst I have no definite details, the ESU publicity and user manuals for the V4 decoders seem to suggest that the motor control which features in the V4 range does more and better than previous ESU decoders.

Very occasionally the calibration will result in a set of completely sorted values. Worth noting as well that the calibration can be run more than once to narrow the result.

Steph, I noted that the result of the auto calibration gave CV results which are very different to the values given in the "typical" motors table of the Lokpilot V4 user manual. We ran the auto calibration step several times and there was very little change in the results... the results for CV53 and 54 were +1 / -1 the values given earlier. When we did the calibration the engine wanted to run further than the length of our test plank, the test concluded with the engine spinning wheels against buffer stops - might the short length of the test run have an impact on the calibration process?

In my experience the auto-calibrate function you describe is just a good starting point for further work using the routine in the manual.

Steph, given that the slow speed control seems to be acceptable I concentrated on adjusting the values of CVs 53 /54 / 55. As our test plank is so short I have no idea as to when the motor achieves maximum speed relative to the maximum speed step - we have made no changes to CV53.

Decreasing CV54 as described in the user manual seemed to have some reduction in juddering at speed in that the problem became evident at higher speed steps (from speed step 5 with CV54 = 17 (auto-calibration) to speed step 9 with CV54 = 5 (after adjustment)). We could not get rid of the judder at speed by adjusting this CV.

Adjusting CV55 seemed to have an improvement when increasing the CV value as with CV54... we tried CV55 values across the entire CV range of 1 to 255 and could not find a value that was optimum.

To ensure that we are doing the process correctly, what is the recommended sequence for auto-calibration and subsequent CV adjustment?

thank you, Graham

 

[Steph Dale]

Graham,
Yes, the loco needs to be free running on the track. Either rolling roads or stop blocks don't work. The decoder measures the inertia of the overall loco amongst other parameters...
Steph

 

[Dog Star]

... the loco needs to be free running on the track.

Thank you Steph, we shall have to wait for warmer weather when we can erect Scrufts Junction before that we can try again.

 

[TheSnapper]

Hi Graham

You know me - I'd suggest you sell the ESU decoders & buy Zimo.....!!

Seriously though, is there nothing in the manual to suggest settings for the Maxon, such as those in the Zimo manual:


Or would Brain Clapperton possibly be able to help?

ATB

Tim

 

[Dog Star]

Thank you Brian, the ESU manual does have a table and does not mention a Maxon as such.

I shall talk to Brian in the New Year.

 

[Dog Star]

Some fine weather this morning and an early start enabled Peter and I to retrieve Scruft's Junction from long term storage and set up the layout for some play experimenting with the Lokprogrammer. Scruft's has a level track of more than ten feet so sufficient for the auto-calibration to work correctly.

Yesterday we got these CV values and a stuttering diesel:-
CV51 to 56 of 0 / 12 / 93 / 17 / 12 / 255

Today, after re-calibration on the longer track, we have:-
CV51 to 56 of 0 / 12 /104 /39 /12 / 255

which shows changes to the CVs for other than speed step one. The result is an engine which runs smoothly at the lowest speed step and is good to the maximum speed step.

Up to now we had been using a Lokpilot V4 on the basis that a mistake on our part is not as expensive as causing trouble to a LokSound V4, So having run through the procedure for auto-calibration of a Lokpilot decoder we replaced the Lokpilot with a Loksound decoder and repeated the auto-calibration process. The results of calibrating the LokSound:-
CV 51 to 56 of 0 / 12 / 105 / 45 / 12 / 255

[As Richard observes (see next post) the differences in the results are comparable with the change delta suggested by the ESU user manual - best left alone]

regards, Graham

 

[richard carr]

Graham

They are not different in the scheme of things its not that precise, as you will have read in the manual when making manual adjustments it recommends making them in steps of 5, less than that and you won't see any real difference.

The main thing is that you are happy with the results.

Richard

 

[Dog Star]

The main thing is that you are happy with the results.

Yes Richard, as things are at the moment, the performance at the lower and upper ends of the speed step range is visually good. I understand what you mean about the change delta as noted in the manual.

We have a humm at the lower end of the range and a gear whine over the middle ranges - maybe the whine shall reduce as the gear sets run in, might take a year or two given the short length of our track.

 

[Steph Dale]

Graham,
It's actually remarkably difficult to blow up a decoder, providing you remember to do a programming track test (recall CV1 value) and a reset before a running test. Under those circumstances you actually don't have enough power to blow the decoder; even if it's connected straight in to a direct short-circuit.
If the decoder fails to respond or your programmer/command station report an error on the CV1 test there's a problem with the installation which requires sorting before attempting to run.
Steph