Calibrating locomotives in a consist

Post all your DCC only problems, solutions and discoverys here.
User avatar
raistlin295
Posts: 686
Joined: Thu Dec 14, 2006 1:04 pm
Location: Wolverhampton

Calibrating locomotives in a consist

Postby raistlin295 » Thu Feb 10, 2011 8:38 am

Hi all,

I'm hoping for some advice please.

I run a pair of Bachmann class 20 locomotives nose to nose in 'consist'. My problem is that although I set such things as start voltage etc to the same value, they do not respond in synch. Sometimes one will pull the other and sometimes one will push the other.

I have a Loksound PC programmer which makes viewing and altering CVs simple but my problem is that I don't know which CVs to adjust or, indeed, in which sequence to go about matching the performance of the two locomotives.

I'd be grateful for any guidance.

Thank you.
Paul

Finsbury Road My Layout

Bigmet
Posts: 7647
Joined: Tue Aug 14, 2007 2:19 pm

Re: Calibrating locomotives in a consist

Postby Bigmet » Thu Feb 10, 2011 9:27 am

Mechanical performance first. The job is much easier with both locos having had some running, say an hour in each direction at moderate speed, so that they have settled down to reliable operation. This is best done on a basic DC controller, with no decoder in the loco, so that you see any running defects that may be present, and can attempt to correct them if the running in doesn't eliminate them. You want the locos to be performing stably from a mechanical perspective, otherwise careful adjustment of CV values can be negated by a change in mechanical performance; they don't need to perform identically, as the decoders can compensate for that.

Now install the decoders. Here's a thing to be aware of, decoders can vary in performance too from example to example. Sensitivity to capacitors being left on the chassis is an example. If the performance when first tested seems further apart on the pair of locos than they were on DC, you can try exchanging the decoders to see if that brings performance closer. Allocate the decoder numbers as usual on the programme track. Set CV's 3 and 4 to zero. My preference would also to be to set for 128 speed steps.

First thing to do is match speed ranges. Take the couplers off the locos so they don't couple up when in contact. Set the maximum speed first individually, altering CV5 until you get the maximum speed you want from each at top speed step, then run them together at maximum speed to see that they are reasonably close, in both directions, in the nose to nose configuration you plan to use. Now set CV2, so that each loco makes whatever dead slow speed you like on speed step 1, check for match in both directions. then check match for top speed again, adjust CV5 if necessary. If there is a difference between forward and reverse speed matching, check the decoder manual for a 'reverse trim' CV that will allow adjustment and apply that to obtain the best match possible. At this stage you may have anything from very similar CV values on both decoders, to quite significantly different. (My experience is of fairly closely matched values most of the time, but I do have a couple of pairings of nominally identical chassis, with nominally identical decoders, where CV2 settings are roughly 10 apart, and CV5 setting is in the more exteme case 110 on one, 180 on the other, 255 steps in CV range.)

Now assess the mid range speed match in both directions at speed step 64, and adjust the CV6 value to get the best match there. Then assess at the quarter points, speed steps 32 and 96. If these match then you are about done on the speed curve. You can test further at any speed step to confirm, an obvious to point to check is the normal maximum running speed that the pair will operate at. (I have yet to encounter a centre motor diesel chassis in a UK OO model (Bach, Hornby or Heljan) which failed to make a decent match on the basic three step speed curve, but doubtless it can happen. If so it will be time to write a custom curve for whichever you consider the less ideal performer to drag it into line with the other loco.)

That done you can apply the CV3 and CV4 values you like, and check for match. Decoders can show quite significant variation here if large CV3 and 4 values are used, presumably down to the precise component values in the IC that determines the time constant. When the locos are under load on the layout, if any slight surging or bumping is noticed, it can be worth trying out switching off BEMF on one of the locos.

That's a fairly simple and straightforward approach. There is a lot more capability available on many decoders, for example to set reference voltage for BEMF, and to adjust sampling rate and sensitivity to optimise matching of the decoder to the motor characteristic, and this could well be worth using in stubborn cases where a stable match proves elusive. So far I have not had to resort to this in setting up past fity OO locos (RTR and kit build) into one of three groups for stable pairing 'any with any' within group, for double heading, banking, or moving locos on and off shed in a group.

User avatar
Flashbang
Posts: 3569
Joined: Thu Oct 18, 2007 1:07 pm
Location: SE United Kingdom
Contact:

Re: Calibrating locomotives in a consist

Postby Flashbang » Thu Feb 10, 2011 10:07 am

Hi
Excellent advice from Bigmet, which if followed closely should give good consist running.

I would as a simple 'just to see' test, initially try turning Off the Bemf on one of the locos.
Both locos having their Bemf On can result in one of the two becoming over Bemf'ed (if there is such a word?) and results in one locos motor pushing harder, then because of this extra loading it receives more volts from the Bemf setting and tries even harder to push and to keep a more constant speed. Turning one (or even both Off) may make them more compatible? Certainly an easy trial anyway. :D
[Image << Click the Icon to go to my website
Broken? It was working correctly when I left it.

User avatar
raistlin295
Posts: 686
Joined: Thu Dec 14, 2006 1:04 pm
Location: Wolverhampton

Re: Calibrating locomotives in a consist

Postby raistlin295 » Thu Feb 10, 2011 11:04 am

Wow,

A masterclass :) Thanks for taking the trouble to put together such a comprehensive guide :)

Thanks to you both for your help.
Paul

Finsbury Road My Layout

noel
Posts: 435
Joined: Sun Dec 23, 2007 1:27 pm

Re: Calibrating locomotives in a consist

Postby noel » Wed Mar 16, 2011 6:21 pm

+

Has anyone come up with a "hard systems" monitor to set all
locos to the same performance curves without the customary
suck-it-and-see approach each time the requirement occurs?

+

peroni
Posts: 24
Joined: Wed Aug 22, 2007 12:03 am
Location: Australia

Re: Calibrating locomotives in a consist

Postby peroni » Wed Mar 16, 2011 10:14 pm

Here is a good sight where you can download a PDF file explaining the how to etc:

http://www.tonystrains.com/technews/con ... -guide.htm

twilightflyer
Posts: 11
Joined: Fri Mar 06, 2015 7:37 pm

Re: Calibrating locomotives in a consist

Postby twilightflyer » Tue Jan 19, 2016 9:37 am

Bigmet wrote:Mechanical performance first. The job is much easier with both locos having had some running, say an hour in each direction at moderate speed, so that they have settled down to reliable operation. This is best done on a basic DC controller, with no decoder in the loco, so that you see any running defects that may be present, and can attempt to correct them if the running in doesn't eliminate them. You want the locos to be performing stably from a mechanical perspective, otherwise careful adjustment of CV values can be negated by a change in mechanical performance; they don't need to perform identically, as the decoders can compensate for that.

Now install the decoders. Here's a thing to be aware of, decoders can vary in performance too from example to example. Sensitivity to capacitors being left on the chassis is an example. If the performance when first tested seems further apart on the pair of locos than they were on DC, you can try exchanging the decoders to see if that brings performance closer. Allocate the decoder numbers as usual on the programme track. Set CV's 3 and 4 to zero. My preference would also to be to set for 128 speed steps.

First thing to do is match speed ranges. Take the couplers off the locos so they don't couple up when in contact. Set the maximum speed first individually, altering CV5 until you get the maximum speed you want from each at top speed step, then run them together at maximum speed to see that they are reasonably close, in both directions, in the nose to nose configuration you plan to use. Now set CV2, so that each loco makes whatever dead slow speed you like on speed step 1, check for match in both directions. then check match for top speed again, adjust CV5 if necessary. If there is a difference between forward and reverse speed matching, check the decoder manual for a 'reverse trim' CV that will allow adjustment and apply that to obtain the best match possible. At this stage you may have anything from very similar CV values on both decoders, to quite significantly different. (My experience is of fairly closely matched values most of the time, but I do have a couple of pairings of nominally identical chassis, with nominally identical decoders, where CV2 settings are roughly 10 apart, and CV5 setting is in the more exteme case 110 on one, 180 on the other, 255 steps in CV range.)

Now assess the mid range speed match in both directions at speed step 64, and adjust the CV6 value to get the best match there. Then assess at the quarter points, speed steps 32 and 96. If these match then you are about done on the speed curve. You can test further at any speed step to confirm, an obvious to point to check is the normal maximum running speed that the pair will operate at. (I have yet to encounter a centre motor diesel chassis in a UK OO model (Bach, Hornby or Heljan) which failed to make a decent match on the basic three step speed curve, but doubtless it can happen. If so it will be time to write a custom curve for whichever you consider the less ideal performer to drag it into line with the other loco.)

That done you can apply the CV3 and CV4 values you like, and check for match. Decoders can show quite significant variation here if large CV3 and 4 values are used, presumably down to the precise component values in the IC that determines the time constant. When the locos are under load on the layout, if any slight surging or bumping is noticed, it can be worth trying out switching off BEMF on one of the locos.

That's a fairly simple and straightforward approach. There is a lot more capability available on many decoders, for example to set reference voltage for BEMF, and to adjust sampling rate and sensitivity to optimise matching of the decoder to the motor characteristic, and this could well be worth using in stubborn cases where a stable match proves elusive. So far I have not had to resort to this in setting up past fity OO locos (RTR and kit build) into one of three groups for stable pairing 'any with any' within group, for double heading, banking, or moving locos on and off shed in a group.



I have just read this post after joining this forum site and you have just answered my very problem that I have been trying to solve on my two class 20's, brilliant information and easy to understand.


Return to “DCC Forum”

Who is online

Users browsing this forum: No registered users and 10 guests