Zimo Decoders and Train Controller profiling

[melmerby]

Hi all

Although Zimo decoders give smoth motor cointrol and repeatability, I find it difficult to get a straight line speed/throttle setting

I have tried MX618N18, MX600R & MX617, all are the same "S" shape.

 

e.g., Hornby W1 with Zimo MX600R with Vmid set to Half of Vmax, which according to the instructions should give a straight line:

 

 

Any idea how to get it flat without using the 28step speed adj?

As a comparison

Hornby W1 with Lenz Standard +:

 

 

 

Edited June 17, 2022 by melmerby

[RFS]

Been there, done that ..🙂

 

Two identical Hornby rebuilt West Country locos as an example.

 

34008 with Zimo MX600R

 

 

And 34013 with Lenz Standard+

 

These are both profiles with 128 speed steps, but TC in offline mode always displays them as 28.  I've tried all sorts to get a smooth profile with the Zimos, but have given up. However, the locos do run smoothly though. 

[jpendle]

On 17/06/2022 at 14:01, melmerby said:

which according to the instructions should give a straight line:

 

Hi,

 

I'm using Zimo decoders in N Gauge locos with iTrain and the speed curves which I see are very similar. There's always a curve at the low end of the speed table and then a straight (ish) line until the loco speed maxes out, then its all over the place! THis is without any adjustments to CV's.

 

 

I usually set Vmax to around 120 to eliminate the top end. I then find that there's no issue running under automatic control. iTrain compensates for the fact that the speed table isn't a straight line,. and I would have thought that Traincontroller would as well.

 

Regards,

 

John P

Edited June 18, 2022 by jpendle
Typo

[melmerby]

1 hour ago, jpendle said:

 

 iTrain compensates for the fact that the speed table isn't a straight line,. and I would have thought that Traincontroller would as well.

 

Regards,

 

John P

Hi John

It does.

 

I bought some Zimo decoders because everyone was raving about them and more fixings were available at "value price"

Admittedly they give good smooth motor operation but unlike Lenz which give an almost dead stright profile, they are all S shaped, which really IMHO they shouldn't be.

That one I posted above has the acceleration and deceleration reduced to zero (TC takes care of that), Vhigh set to 190 and Vmid set to 95. I haven't fiddled with any other settings apart from DCC only operation and 4 digit addressing.

 

Having tried a Train-o-Matic Lokommander 2 mini at a similar price, I would be inclined to use them as they, like Lenz and unlike any others I have tried, give a straight line response.

[RFS]

I've had a look at my others with Zimo decoders. Here's a Bachmann 2-HAP EMU with an MX638 decoder at firmware level 40.5 with CV5=1 and CV6=1.  Looks much better to me.

 

To control max speed I used CV57=70 (max voltage). 

 

Edited June 19, 2022 by RFS
Added CV57

[jpendle]

17 hours ago, melmerby said:

Having tried a Train-o-Matic Lokommander 2 mini at a similar price, I would be inclined to use them as they, like Lenz and unlike any others I have tried, give a straight line response.

Hi Keith,

 

Why is it important for you that you get a straight line when speed profiling your loco's? Has it led to any issues?

 

Thanks,

 

John P

[melmerby]

3 hours ago, jpendle said:

Hi Keith,

 

Why is it important for you that you get a straight line when speed profiling your loco's? Has it led to any issues?

 

Thanks,

 

John P

i'll turn it on it's head.

Do you think a decoder with a response like this is worth using:

 

 

[Crosland]

Are you certain the bit in CV29 to use the speed table, rather than the Vstart, Vhigh, Vmid CVs, ..., is not set?

[jpendle]

1 hour ago, melmerby said:

i'll turn it on it's head.

Do you think a decoder with a response like this is worth using:

 

 

I honestly don’t know!

I’ve been working on the assumption that the measured response is a combination of decoder and loco characteristics. If I had a decoder tester I’d use that to try to work out if it’s  the decoder or the loco.

 

Regards,

 

John P

[Mike Buckner]

5 hours ago, jpendle said:

Why is it important for you that you get a straight line when speed profiling your loco's? Has it led to any issues?

 

Speaking for myself:  Because that's what I want.

 

[WIMorrison]

Never actually bothered about the shape of a speed 'curve' as I have always been more interested in the actual performance and controllability of the loco. That said, a slight S curve should be preferred to a straight line anyway because you want more speed steps with less variation in speed at low speed because this is where you want fine control for shunting, stopping, starting, etc. Similarly, at top speeds you need less control and can have larger speed changes for fewer speeds steps.

 

 

[WIMorrison]

23 hours ago, melmerby said:

i'll turn it on it's head.

Do you think a decoder with a response like this is worth using:

 

 

 

If this is a Zimo decoder then use of CV #57 will enable you to shift the flat bit at the top to be 'shunted' off to the right and disappear thus providing a less steep progression in the middle speeds.

[Nigelcliffe]

Have those seeking a straight line graph in a Zimo decoder looked at CV's 135 and 136 ?    

 

I've never used them, but the manual implies a straight line speed relationship is established with, eg. speed step 1 = 1kph, step 10 = 10kph, step100 = 100kph, step126=126kph,  etc..   Other ranges to higher or lower top speeds are possible.  

 

 

- Nigel

[WIMorrison]

@Nigelcliffe Aren't these CVs something to do with constant braking distance (or something like that?). I have also never used them therefore also don't know what direct effect that they may have on speed curves.

[melmerby]

3 hours ago, WIMorrison said:

 

If this is a Zimo decoder then use of CV #57 will enable you to shift the flat bit at the top to be 'shunted' off to the right and disappear thus providing a less steep progression in the middle speeds.

Hi Iain

That's a Digitrax decoder, the worst I've had.

The Zimo one is the first post, with the smooth "S" shape.

 

The thing that is bugging me is that Zimo, in the instructions p.15, first speed example, show a straight line using Vstart = 1, Vmid = 127 & Vtop = 252. I don't get that with any of their decoders.

Also It states in example three that an "S" shape curve can be attained using the 28 step mode (CVs 67-94). However, I get that with example one's setting.

Edited June 20, 2022 by melmerby

[RFS]

Keith - as I posted earlier I get a straight line with CV5&6 set to 1, and with CV57 reduced till max speed is what I want. I have not as yet reprofiled earlier locos as they run OK even with the S speed curve.

 

Try it with one of yours.

Edited June 20, 2022 by RFS

[Nigelcliffe]

12 hours ago, WIMorrison said:

@Nigelcliffe Aren't these CVs something to do with constant braking distance (or something like that?). I have also never used them therefore also don't know what direct effect that they may have on speed curves.

 

They affect speed restriction devices, such as HLU or some elements of ABC.   But that's a consequence of using them, rather than their primary aim. 

 

According to the manual, they are to calibrate the decoder and model to achieve a direct throttle-to-speed relationship.   One speed step = 1kph  (or 2kph or 0.5kph if different options taken).   So, ten speed steps = 10kph, 100 speed steps = 100kph.      If the manual is correct, then its a linear speed behaviour.   

 

(I've low interest in trying this, and setting up devices to measure the results is a lot of time I don't want to spend on it ).  

 

- Nigel