Zimo MX620

[chinapig]

  Normal 0

Hi Folks,

 

I purchased a N Scale Con-Cor 2-10-2 on eBay with a Zimo decoder fitted.  After a bit of investigation I’ve deduced it’s a MX620.  I can’t find much documentation on the web but I’m lucky in that I’m using a Digitrax PR3 and JMRI to help program it.  It has got some fields for BEMF but the trouble is I just can’t understand them.  At the moment the loco runs away downhill and slows right down going uphill.  Is there anyone out who has cracked this, please?

 

Ted

[tstage]

Ted,

You might try either of these two sites:

http://www.mrsonline.net

http://translate.google.com/translate?hl=en&sl=de&u=http://www.zimo.at/&ei=gAvjSuyaBtGg8Aa5-uHyAQ&sa=X&oi=translate&ct=result&resnum=6&ved=0CBYQ7gEwBQ&prev=/search%3Fq%3Dzimo%252BDCC%26hl%3Den%26safe%3Doff

The first at least has an information contact e-mail address, as well as the link to the Zimo Forum group on Yahoo!

Hope that helps...

Tom

[chinapig]

 Thanks Tom.  Both sites look really useful, especially the first.  The second site just rubs in the fact that my four years at German night-school were pretty much a waste of time!  No, it's alright, I've just found the English translation.

Cheers, Ted

[blrrfan]

try http://www.zimo.at/web2007/pdf/MX620MX630MX640E.pdf

[chinapig]

Many thanks for that link, birrfan.  The problem is that none of the documents are a "Zimo for Dummies".  I mean they start talking about Mantissas.  My son has a Batchelors Degree in Maths and he had't come across them!  I guess it will come down to trial and error, but at least I know what CVs to put on trial.

Many thanks, Ted

[blrrfan]

That part ("mantissa") is for CV9 and follows the wording from the NMRA RP 9.2.2

see http://www.nmra.org/standards/DCC/standards_rps/rp922.html and scroll to the explanation for "Configuration Variable 9"

NMRA R.P. 9.2.2 (extracted part for CV9)

The value of CV#9 sets the nominal PWM period at the decoder output and therefore the frequency is proportional to the reciprocal of the value. The recommend formula for PWM period should be: PWM period (uS) = (131 + MANTISSA x 4)x 2 EXP ,Where MANTISSA is in bits 0-4 bits of CV#9 (low order) and EXP is bits 5-7 for CV#9. If the value programmed into CV-9 falls outside a decoder's capability, it is suggested (but not required) that the decoder "adjust" the value to the appropriate highest or lowest setting supported by the decoder.

[rrinker]

 The mantissa is the decimal part of a logarithm - surprised a math major would not know that. However, there is an alternate use in science and computer science which means the significant part of a floating point number. Some scientific calculators have a mantissa button which sets the number of significant digits in results. Yes there is a conflict here with the mathematics definition of the word.

 Based on the expression given in the NMRA RP, they are using the latter definition. SO if the number is 1.53 x 10^-3, the mantissa would be 1.53 and the exp would be -3

 BEMF on Zimo decoders is probably the best there is - however they explain the settings in very scientific terms, and don't reveal the exact equation being used (as this is proprietary information). To truly understand the ideal values for the various BEMF CVs in a Zimo decoder, you need to understand exactly how BEMF calculations are made.

 Or, you can wing it and start plugging in numbers until it runs the way you want. You're not going to burn out the motor with incorrect BEMF settings, it will just run lousy - huge jumps between speed steps, slowing on hills (if the range of settings allows it - maybe even slowing going DOWN hill!) - but the motor won't get excessive voltage or anything, so it's safe to play around. Record the default values if the decoder does not have a reset so you can get back to a known point if you go too far.

                                                   --Randy

 

[chinapig]

Randy,

Thanks for that information.  I'm learning more each day but don't think i'm going to delve any deeper!  I'm going to be winging it from here by plugging in values on a trial and error basis.

It's amazing what this hobby of ours exposes us to.

Cheers, Ted

[rrinker]

This is how I do it, and I even have a degree in electrical engineering (granted, I never took any DC motor courses). Even knowing the equation, the math is nearly beyond me, as I have no used any of that calculus I took oh so many years ago.

I don't have the link handy, but somewhere there is a comparision of the 3 values to a spring, which helps make it make some sense. It's also in the Digitrax documentation except Digitrax decoders only give you control over 2 factors in the equation - they do compare it to a spring, but I know I've seen or read somewhere else how each of the three variables that Zimo allows you to adjust to a spring.

                                   --Randy

[aluesch]

Hi Ted.

You may have the newer MX620 or an older MX62. The latter can be recoginzed on the ceramic board.

Most people here are talking about BEMF compensation settings (for smooth motor control), which are done automatically on the MX620 but can also be played around with manually using CV #9 and CV #56.

In your case however, I assume you are more concerned with turning  BEMF action ON/OFF or a setting in between. This is done in CV #58 and is quite simple. A value of 0 equals no BEMF and a value of 255 full BEMF function. If you prefer your trains to slow down a bit going uphill, set this CV to a value between 100 and 180. 
CV #10 and #113 can also be used to reduce BEMF action above a specific speed. If you don't want/need this function, leave the values at 0.

You can download a MX620 manual here: http://www.mrsonline.net/html/manuals.html
or here: http://www.zimo.at/web2007/content/handbuch.htm
The older MX62 manual can be downloaded from the ZIMO web site, just scroll down to "obsolete Decoders".

Hope that helps!

Art

ZIMO Agency of North America

www.mrsonline.net

 

[chinapig]

Thanks for all the help, folks.  I’ve solved my problem.  After tweaking of all the CVs to no affect, I gave up and set them back to the default.  I then adjusted CV5 down to around 100 as the loco had reached its top speed at about half throttle (for some strange reason this loco’s top speed is pretty slow but as it’s a 2-10-2 I’m happy with it).  Low and behold, going down the helix it didn’t speed up.

 

I guess it was trying to reach its true top speed as set by the throttle – I can’t really explain it technically.

 

Now that’s solved I can track down and nail some other problems!

 

Thanks again, Ted