[Matthew Peddlesden]

Hi Folks,

Would someone be able to give me a plain-english explanation of how the back-emf tuning CV's work?

I've managed some success, turning an undriveable old 47 in to something that's actually half decent however i'm down to one kind of behaviour that is common to it and two other locos. The decoders are all 36-553 Bachmann 3 function with Back-EMF / Load Compensation.

The effect is that if you have the trains running at whatever speed, then drop the speed to zero, they'll gently slow down and as they are about to stop they'll lurch and sputter momentarily, about a wagon length, before stopping. The amount of sputter depends on the speed, if you go from full speed to zero then you get about a wagon length, if you gently bring the speed down yourself it runs almost perfectly.

The controller is a Lenz Compact + LH30, though I have another loco (Bachmann 56xx) with the same decoder that does not exhibit this problem so i'm pretty sure it's just a matter of balancing the EMF parameters to the motor...

Unfortunately, reading many tutorials online i'm still befuddled as to how it works.

Any recommendations greatly appreciated

Matt.

 

[SPROGman]

QUOTE (Matthew Peddlesden @ 28 Jan 2007, 16:52) <{POST_SNAPBACK}>Hi Folks,

Would someone be able to give me a plain-english explanation of how the back-emf tuning CV's work?

I've managed some success, turning an undriveable old 47 in to something that's actually half decent however i'm down to one kind of behaviour that is common to it and two other locos. The decoders are all 36-553 Bachmann 3 function with Back-EMF / Load Compensation.

The effect is that if you have the trains running at whatever speed, then drop the speed to zero, they'll gently slow down and as they are about to stop they'll lurch and sputter momentarily, about a wagon length, before stopping. The amount of sputter depends on the speed, if you go from full speed to zero then you get about a wagon length, if you gently bring the speed down yourself it runs almost perfectly.

The controller is a Lenz Compact + LH30, though I have another loco (Bachmann 56xx) with the same decoder that does not exhibit this problem so i'm pretty sure it's just a matter of balancing the EMF parameters to the motor...

Unfortunately, reading many tutorials online i'm still befuddled as to how it works.

Any recommendations greatly appreciated

Matt.
There's no standard for these CVs and the method varies between decoders and depends upon the algorithm used. Better makes such as Zimo have comprehensive instructions for the decoder. Ask the decoder manufacturer.

Andrew Crosland

 

[Doug]

Take a look at Page 18 of the ESU LokSound decoder manual. ESU LokSound is what is installed on the Bachmann DCC Sound locos.

Also, look at the DCC On Board PDF from Bachmann. On page 1 they talk of Back EMF and n page 2 they give CV values (CV 53 to CV 55) and rational behind the values.

 

[Matthew Peddlesden]

Ah yes, nice one, the ESU one is more useful than most of the documents i've found on the subject... now to get 'circle' (underground ernie) to respond smoothly...

Matt.

 

[Matthew Peddlesden]

Result!

For future reference, for the Underground Ernie "Circle" train (presumably will apply to all the ernie range as they come out), using a Bachmann 36-553 decoder, it will suffer from quite bad stuttering as you speed up from a standstill or return to a standstill.

The following CV's cure it :

CV53 = 22
CV54 = 17
CV55 = 56

A little explanation:

CV53 - "Feedback Reference" - When I tested this I found I had reached top speed on the unit by the time I got to about 60%, following the instructions this setting was reduced and it now reaches full speed at around 95% or thereabouts.

CV54 - "Feedback Parameter K" - This determines how much of an impact the load compensation has, I tweaked this down a bit just to smooth things out a little.

CV55 - "Feedback Parameter I" - This one is quite crucial on the Ernie range, it dawned on me as I tested that the motor has a lot more drag on it than a conventional one because it's also driving all the gears for the eyes to waggle back and forth. Set this as high as it will go just about and this is the one that has the most effect.

Circle is now extremely controllable and very smooth running at all speeds.

Finally, i'd just like to add, top points to Bachmann for including a NEM SOCKET even on the Underground Ernie range - it's hardly an expensive range of stock for kids.

I (nearly) fell off my chair when I took the top off (a dead easy to loco to open) and found a socket with a blanking plug in it, moments later replaced with the decoder and operational! Brilliant.

Now I just have to learn how to stop it IN the station not half a coach length beyond it... but i'll leave that to the 3yr old expert, my job here is done

Matt.

 

[Doug]

Well that's some good news. Back EMF is not a subject that comes up much and definitely not something that comes up often when talking about tuning the fine running of a loco under DCC.

I'll be looking into this in more detail too. Thanks for taking the time to post your findings.

 

[Matthew Peddlesden]

I can't remember exactly where I read about it but essentially if Back EMF is on a decoder then it is important that it is configured using those three CV's to match the characteristics of the motor. The ESU doc is one of the best i've seen at helping understand how it works.

Where Back EMF doesn't match the motor's characteristics, it doesn't know how to respond - so it does its thing with the pulse width to get the motor going, and then it checks the return in the downtime, there's nothing coming back so it widens the pulse, next time it checks there's too much coming back (the loco just shot off) so it drops it, and then the loco stops etc - when what it really needed was a little patience because the motor is a bit slower to respond (i.e. on circle there's a load of gearing for the eyes).

So if you have a chipped loco that behaves very poorly at the bottom end of the speed range, jumping or sputtering etc, it is almost guaranteed to be Back EMF - the easiest verification is to switch EMF off, it should run more smoothly at the bottom end but without the benefits of *really* smooth running that EMF gives you.

So, first switch Back EMF off, if it behaves better you know to fix these CV's. Then switch it back on and start fiddling, go with the instructions in the ESU doc as they are the least confusing (I wouldn't go so far as to call them simple though).

Now to get my old Hornby 47 to quite jumping and stuttering.

Matt.

 

[SPROGman]

QUOTE (Matthew Peddlesden @ 30 Jan 2007, 10:11) <{POST_SNAPBACK}>So, first switch Back EMF off, if it behaves better you know to fix these CV's. Then switch it back on and start fiddling, go with the instructions in the ESU doc as they are the least confusing (I wouldn't go so far as to call them simple though).
But, crucially, the ESU instructions apply only to ESU decoders. Other manufacturers will vary. Lenz give you no choice other than on/off and a few pre-defined motor types without explaining what they are. ZIMO decoders have fully tuneable BEMF parameters.

Andrew Crosland

 

[Matthew Peddlesden]

All the docs i've seen thus far talk about the same kinds of parameters, I and K, there is a third parameter that some offer as well but is less important so many don't. CV numbers aren't always the same but you can usually match the descriptions to the parameters they refer to where they exist.

Certainly it's true to say that not all decoders support the parameters, the Bachmann's are just ESU's under the hood anyway though

Wasn't aware that Lenz had such simplistic BEMF controls though, but then I'm very happy with the 36-553 decoder for my needs.

Matt.