Speed matching issue on different layouts

[jagged ben]

I have an issue I'm a little baffled by...

I've built a small test oval at my house and I have a Digitrax Zephyr (newer DCS52 with built in USB) that I've been using to speed match locomotives before bringing them to the club where I do my real running.  The club system is also Digitrax, not sure the command station model, probably an older DCS100 or 200.

I'm using basic speed control (start, mid, max voltage) not speed tables except in a few cases.

The issue I'm having is that I can get locos nicely matched at home, but when I take them to the club maybe 25% of the locos are much slower than the others.  I've checked the Digitrax Status Edit feature and those settings are correct. So I dont think it's a speed step issue.  It doesn't seem to be related to the brand of decoder.  I'm using universal consisting, but it doesn't seem to matter if locos are consisted.

My leading theory is that the locos are getting a lower track voltage at the club, and that the locos with more friction in the drives are therefore getting lower voltage at the motor terminals, and that the effects on speed are not proportionate particularly at lower speed.   I was surprised tonight to measure the home test track voltage at over 18 VAC.  Club voltage measures 14-15 vac (we have it on the HO setting), with the important caveat that I haven't been using the same meter and with DCC waveforms that could be a big factor.  So taking my home multi meter to the club to get apples to apples voltage readings  is one task here.  Anyway, I thought I'd ask these questions...

Can you think of anything else I should be checking for that could cause this?
Have you had this problem and have you come up with any way to deal with it?
Does anyone know if it's possible to adjust the track voltage on the DCS52?
Supposing I have to deal with different track voltages, can anyone suggest a mathematical approach to making adjustments?

Thanks in advance.

[tehachapifan]

I'm not sure you said what brand decoders these are but, in ESU decoders, there's a "switching mode" function that can cut the speed way down. If your decoders are ESU or have a similar function, is it possible that the function key for this mode is somehow active on the club's controllers?

Also, I'm assuming this is occurring when running locos in a consist? Are you running them in a consist at home too? I know of at least one decoder brand that behaves entirely different speed-wise when running solo or when in a consist.

[jagged ben]

The decoders are mostly Digitrax.  There are some NCE and some ESU sound decoders in the mix as well, but that doesn't seem to correlate to the locomotives with issues.  Pretty sure that switching mode is not what's going on, I could see that being an issue with a random locomotive but not a pattern.  How woudld I double check that?  My unscientific impression is that the locos that are slower at the club were slower to begin with before I did the speed matching.

It really doesn't matter whether they are consisted or not.  The difference in speed between home and club appears to be the same in or out of consist.

[John]

I think you are dealing with a voltage difference problem ..  just remove the friction variable from your hypothesis -- there is little you can do about that ..

On my layout, the track voltage is usually around 11.5 volts measured across the rails with a digital VM in AC mode.  It's not completely accurate, but close enough .. that is with the booster on the Nscale setting ..

My locomotives are mostly speed matched with a top scale speed of 80 smph.    It stands to reason, that when I calibrated the speed table the low mid max CV settings were based on the value of the input at the time that they were programmed.

If you now were to switch the layout to the HO setting, the voltage would be higher -- so the CV settings you put in are probably going to be lower for the same scale speed.

Try this ..

Take a VOM and measure the track voltage at your layout .. then measure the time for the loco to travel across a measured distance .. 4 feet is a good value.   Use this app to calculate spped   https://www.modelbuildings.org/scale-speed-calculator/  then repeat the process with a higher and lower track voltage ..Since the DCS52 is set for 13 Volts at the track, with no option to change it you would have to drop the voltage with a couple of diodes .. 3amp diodes -- put them in series -- each should lower track volts by about 1

[jdcolombo]

Agree with John that the issue is likely to be caused by different voltage at the rails.  Most decoders have a CV that sets the throttle response given the input voltage.  On ESU decoders, this is CV53, and it should be set at roughly 10 times the track voltage to get appropriate throttle response across the entire range of the throttle.  Move engines set correctly for 13v to a layout that is putting out 11 or 15v, and the throttle response of the decoders will be different.  Add in the fact that the motor will respond differently over the speed step range at a higher or lower track voltage (e.g., each speed step becomes a different voltage); no motor is perfectly linear in response, so the motor will respond slightly differently at different voltages, causing speed mismatch between engines.

As for measuring track voltage, remember that DCC is an AC signal, not DC.  You can only accurately measure the voltage with a multi-meter if it will measure low AC volts, although I think there is a way you can connect a DC meter to one rail and ground on the command station, take a reading, and then double it for track voltage.  Most cheap multimeters won't measure low AC volts (even if they have an AC v setting, it often will only measure 100v and above - house line level).

John C.

[John]

As for measuring track voltage, remember that DCC is an AC signal, not DC.  You can only accurately measure the voltage with a multi-meter if it will measure low AC volts, although I think there is a way you can connect a DC meter to one rail and ground on the command station, take a reading, and then double it for track voltage.  Most cheap multimeters won't measure low AC volts (even if they have an AC v setting, it often will only measure 100v and above - house line level).

John C.

WARNING - THREAD DRIFT BELOW     

John .. you are absolutely correct .. Digitrax has this method on their website for measuring output voltage .. https://www.digitrax.com/tsd/KB909/track-voltage-measurement-on-dcc-layouts-with-dire/

I have a cheap digital volt meter .. when I want to measure my track voltage -- I set it to AC volts, set it on a high range setting -- and it will measure usually 11-12 volts on the track .. while it's not absolutely accurate -it's "good enough" ..

Track voltage on DC (analog) layouts can be measured from rail to rail.  However, with DCC layouts that use direct home wiring as recommended by Digitrax, this method of measurement does not work with most voltmeters.

Track voltage (output from the booster to the rail) on DCC layouts with direct home wiring can be measured as follows:

1.     With the system powered up and with the TRACK STATUS LED on, use your throttle to select the analog address 00 and set the speed to 00.

2.     With a multimeter set to the 20 volt DC scale, measure the voltage from RAIL A to ground (you can use the SCALE or MODE toggle switch body, the case screw on the back, or the GROUND terminal on the front of the booster as ground for this measurement).  Repeat this measurement from RAIL B to ground.

3.     Total track voltage is the sum of the voltage measurements from RAIL A to ground & RAIL B to ground.

4.     The difference between the RAIL A & RAIL B voltages should not exceed 0.2 volts.  Satisfactory operation will still occur if the difference is as much as 0.5 volts.

5.     Track voltage can be adjusted by using the trim pot inside the booster located between the LOCONET B port and the SCALE switch.  This is useful for balancing the track voltage between power districts.

6.     If you find that the track voltage on one of your boosters falls outside these ranges, be sure that there are no analog addresses running in the system by using your throttle to select the analog address 00 and setting the speed to 00.  Once you have done this, re-measure the voltages. 

[jagged ben]

Thanks for the responses guys, and the sanity check that it points to voltage.  Yeah, seems like the best long term solution is to get the test track voltage to match the club track voltage.

I can't believe that it's not possible to adjust the voltage on the DCS52.  Well, I guess I can believe that, I just don't want to yet.    

Take a VOM and measure the track voltage at your layout .. then measure the time for the loco to travel across a measured distance .. 4 feet is a good value.   Use this app to calculate spped   https://www.modelbuildings.org/scale-speed-calculator/  then repeat the process with a higher and lower track voltage ..Since the DCS52 is set for 13 Volts at the track, with no option to change it you would have to drop the voltage with a couple of diodes .. 3amp diodes -- put them in series -- each should lower track volts by about 1

I'm measuring over 18 volts from DCS52, for what that's worth.  I think that is the better multi-meter, not that I completely mistrust the one at the club (but it will be interesting to compare the results apples-to-apples).

I don't quite follow the diodes idea.  Wouldn't resistors be the component to use?  How do you know how much the voltage will drop for the components you use?   I think I'd like to either better understand the electrical math there, or maybe I will just try to find a DB150 on eBay since I know the voltage from those can be adjusted.

...On ESU decoders, this is CV53, and it should be set at roughly 10 times the track voltage to get appropriate throttle response across the entire range of the throttle.  Move engines set correctly for 13v to a layout that is putting out 11 or 15v, and the throttle response of the decoders will be different. 

John C.

This is interesting, too, possibly a different kind of workaround.  It seems like you're saying I could mitigate what I'm experiencing by setting the speed at home, then adjusting this CV upward proportionally to the voltage difference?   Unfortunately it seems like Digitrax decoders do not have this CV.  CV53 is not used in FX3 decoders, and I don't see anything else in their list that resembles this.  One possible solution: stop buying Digitrax decoders.   

[John]

It's odd that your DCS-52 is putting 18V on the track ..   please disconnect it from the track, then follow the procedure above for measuring the voltage . use address 00 .. then measure between ground and rail a and b using a DC voltage meter ..

The voltage drop on a diode depends on the type .. I use 1N5400s .. they have a forward voltage of 1.2V .. ( which is the voltage drop across the diode )   https://www.infobloom.com/what-is-the-diode-forward-voltage.htm

[peteski]

I don't quite follow the diodes idea.  Wouldn't resistors be the component to use?  How do you know how much the voltage will drop for the components you use?   I think I'd like to either better understand the electrical math there, or maybe I will just try to find a DB150 on eBay since I know the voltage from those can be adjusted.

In this instance diodes are the correct components to use for reducing the voltage.  Just like diodes were used in DC days to drop 1.4V going to the motor (to be a source of 1.4V for the constant lighting headlights).

Resistors are linear devices, so the voltage drop across them is proportional to the current flowing through them. So the voltage drop would vary depending on how many engines are on the track (consuming current). 

Diodes are non-linear, so the voltage across their P-N junction will be fairly constant, regardless of the current passing through them.  Perfect for dropping a known quantity of voltage, regardless of how many locos are in the track.

[jagged ben]

Okay, using the measure-DC-from-'ground'-to-each rail method I get 14V on the home test track.  Disappointing to find out my home multi-meter isn't that good with the AC square wave.  Anyway, I'll have to get more measurements at the club because now I don't know where the voltage difference is, or its real magnitude.

I don't get the diodes setup.  I put them in series with the track circuit?  Won't that drop the voltage from one rail to 'ground' more than the other?   I think I need a diagram or a more complete description of the solution.

In this instance diodes are the correct components to use for reducing the voltage.  Just like diodes were used in DC days to drop 1.4V going to the motor (to be a source of 1.4V for the constant lighting headlights).

Resistors are linear devices, so the voltage drop across them is proportional to the current flowing through them. So the voltage drop would vary depending on how many engines are on the track (consuming current). 

Diodes are non-linear, so the voltage across their P-N junction will be fairly constant, regardless of the current passing through them.  Perfect for dropping a known quantity of voltage, regardless of how many locos are in the track.

Okay, but what if it turns out the issue is increased voltage drop at the club due to long wiring runs and such, i.e. that the issue is dependent on the locos current draw?    Then I would want to simulate that at home, and resistors would be the appropriate component, maybe?   I intend to get voltage measurements at the club with and without locos running and also compare problem loco by itself vs with a consist sharing the track.  I think I need to shed more light on the problem before assuming the solution. 

[peteski]

I don't get the diodes setup.  I put them in series with the track circuit?  Won't that drop the voltage from one rail to 'ground' more than the other?   I think I need a diagram or a more complete description of the solution.

Yes, you can install them in-series on either lead of the DCC bus.   The booster's ground is not involved in this (the decoder in the loco has no connection to the ground).

Here is a diagram showing few example of what the diode circuit would look like.

Okay, but what if it turns out the issue is increased voltage drop at the club due to long wiring runs and such, i.e. that the issue is dependent on the locos current draw?    Then I would want to simulate that at home, and resistors would be the appropriate component, maybe?   I intend to get voltage measurements at the club with and without locos running and also compare problem loco by itself vs with a consist sharing the track.  I think I need to shed more light on the problem before assuming the solution.

Well, I didn't get the impression that at the club layout the speed mismatch occurs at different severity in various locations of the layout.  Plus, I assume that when you are testing the speed, the locos being tested are running fairly close to each other, so any resistive voltage drop from the DCC bus, feeders, and track, would affect both locos (both locos would see the same voltage at that section of track).  Honestly, N scale locos, running light, don't consume all that much current. I would expect your club's layout wiring to be robust enough for the drop between the booster and the locos to be negligible.

And yes, you are correct, if you wanted to simulate voltage drop that occurs in the bus/feeders/track, resistors would be the proper components to use, since bus/feeders/track can be considered low-value resistors.

[jagged ben]

Yes, you can install them in-series on either lead of the DCC bus.   The booster's ground is not involved in this (the decoder in the loco has no connection to the ground.

Here is a diagram showing few example of what the diode circuit would look like.

Aha, so the key is they go in both directions, paralleled with each other but in series with the rest of the circuit.  That was a key point.  Thanks!  Makes so much more sense now.

Well, I didn't get the impression that at the club layout the speed mismatch occurs at different severity in various locations of the layout.  Plus, I assume that when you are testing the speed, the locos being tested are running fairly close to each other, so any resistive voltage drop from the DCC bus, feeders, and track, would affect both locos (both locos would see the same voltage at that section of track).  Honestly, N scale locos, running light, don't consume all that much current. I would expect your club's layout wiring to be robust enough for the drop between the booster and the locos to be negligible.

And yes, you are correct, if you wanted to simulate voltage drop that occurs in the bus/feeders/track, resistors would be the proper components to use, since bus/feeders/track can be considered low-value resistors.

Well, the problem is ... who knows.  For a minute I thought it was a source voltage difference as discussed abive, but having remeasured the Zephyr voltage I now have no real idea and I'll have to get more empirical evidence to. It could be that the club wiring is just not as good as it should be.  (For one thing a lot of it is pretty old, and splices and solder joints may be going bad.)  Or it could be something else.

[John]

At the end of the day -- the only value that matters what the locomotive picks up .. the diodes  just help to adjust in a standard way .. if you know your Zephyr is putting X volts on the track and thats what you used to calibrate them .. on my layout -- I see a few places were the speed changes just because my wiring at that spot is inferior .. some day I will fix that

[mmyers]

Normal track voltage for a Zephyr is about 13.8 volts. Any other Digitrax booster set on the N scale setting will be putting 11.5 to 12 volts on the rails. That is the reason your speed matching at home doesn't give you the same results on the club layout. A club member had this same problem several years ago and the diode bridge solution worked to get his Zephyr's track voltage down close to the club's track voltage. Next time he ran his trains, they ran just like he had set them up them at home.

Martin Myers

[jagged ben]

Normal track voltage for a Zephyr is about 13.8 volts. Any other Digitrax booster set on the N scale setting will be putting 11.5 to 12 volts on the rails. That is the reason your speed matching at home doesn't give you the same results on the club layout. A club member had this same problem several years ago and the diode bridge solution worked to get his Zephyr's track voltage down close to the club's track voltage. Next time he ran his trains, they ran just like he had set them up them at home.

Martin Myers

Thanks.  The thing is, I'm pretty sure our club boosters are on the HO setting.  Also I measured it at around 14 volts, but then again that was on the AC setting with a different multimeter than the one at home.  So that's where I need to get more data from different methods and meters.  Haven't had a chance this week to go by the club.

I'm hoping that the issue is exactly as you describe, but we shall see.  Good to know the diode bridge worked for your friend.