G Scale Model Train Forum banner
1 - 15 of 15 Posts

·
Registered
Joined
·
60 Posts
Discussion Starter · #1 ·
I have a set of USA Trains R22271 Santa Fe F3 ABBA that I am trying to consist. I have converted them all to DCC with ESU decoders, Loksound 5xl in the A units (in which I installed speakers) and Lokpilot 4xl in the B units, which have no sound.

My high level question is should I just turn back EMF off with a 4 loco consist, or is it worth it to try and speed match and fine tune back EMF.

I am having issues with the consisting. All the locos run well independently but I am getting frequent issues when I consist them all together. With larger loads, the consists are frequently tripping my DCC Specialties DCC circuit breakers (which are all set to trip at 10 amps).

I think it is likely a speed matching/back EMF issue. I looked at the motor and driving setting for all decoders on my Lokprogrammer. Both As and both Bs are the same, but there are slight differences between the As and Bs on both driving and back emf settings. Also note there are some differences between the available tuning parameters between the ESU 4.0 and 5 xl decoders.

I haven’t yet run the motor auto tune function or harmonized the drive settings. Before I start all that, wondering if I should just turn off back EMF? Any thoughts welcome.
 

·
Super Modulator
Joined
·
22,477 Posts
I have not had the issues with BEMF on in consisting as the "common wisdom" seems to always point out.

Speed match is your best buck.

Before you continue to guess, what does your DCC ampmeter read? (if you don't have one, you need one if you run anywhere near 10 amps) Get the RRAMPMETER, and get the best one (at one time they offered 4 models)

You should run the "auto tune" which helps map the BEMF parameters.

Also remember that under load, the locomotives will buck less... running several locos light will seem worse that it really is.

Do the auto tune
do the speed match
measure the current (double check your circuit breaker setting, are you using jumpers?)

greg

 

·
Registered
Joined
·
60 Posts
Discussion Starter · #3 ·
I have an rrampmeter on the line and it is clearly spiking. I have set the circuit breakers with cvs I didn’t trust the jumpers. Going to try auto tube and speed matching. Thx
 

·
Super Modulator
Joined
·
22,477 Posts
What level did you set the trip current? Also you might test individual locos to see if there is one culprit.

All locos in my video have 2 speakers, one in the fuel tank and one shooting out the top grill.

If you don't have a damaged loco, your grades must be wild, or you are pulling a lot of cars. My 55 car train went up a 3.4% grade and did not pull 10 amps.

I wonder if there is a fault somewhere.. Also, did you put the heat sinks on your circuit breaker? You need them if you go over 10 amps.

no heatsinks:



With proper heatsinks:


Greg
 

·
Registered
Joined
·
60 Posts
Discussion Starter · #5 ·
CV 49 set to 08 so 10.2 amp trip. I don’t have the heat sinks yet, I really don’t expect to be running consistently that hot - I have 23 separate power districts/circuit breakers, so I wont have any more than one loco in a power district at any one time.

Grades are all 2% or less (except on my mountain run approach, which is 4% but only services smaller trains).

I think it has to be speed matching/back EMF. I pulled together the consist without doing any speed matching or back emf adjustment at all. I did a quick pursuit test and the B units (with the Lokpilot 4xls) were clearly outrunning the A units (with the Loksound 5XLs).

To be honest, I dodn’t even know you had to speed match or tune back EMF for consists. Dumb newbie. I was spoiled because i had to date been running several consists with no problem, but they were all AAs using the same locos from the same manufacturers with the exact same decoders and decoder settings. Likely were fine ‘out of the box’ with no adjustment.

So I have spent a good amount of time today getting educated on speed matching and back emf tuning. I have a scale speedometer on order. Looking forward (somewhat) to doing it the old fashioned way.

I love your idea of having throttle setting (1-128) tie directly to smph. I’m pointing to that. Standing on the shoulders of giants…
 

·
Super Modulator
Joined
·
22,477 Posts
Trip setting looks fine, but you should try to get the heat sinks, you are on the cusp of needing them. I assume what is tripping is the circuit breaker version, not the autoreverser.

The B units should be identical speeds to the A units, so something is strange. I would set cv5 to 255 in all units, but if it was me, I would read that value first from all locos. Also set momentum to 0 in 3 & 4 on all locos... just to keep confusion out of it.

Run the auto BEMF on all units.

One theory here is that perhaps different firmware loads/versions have different default settings, should not be, but force them all the same.

"common wisdom" will tell you all kinds of things about BEMF and consisting. I have never had to disable BEMF to consist. That said, messed up BEMF settings can cause issues. (really only extreme settings)

But all in all somehow something does not make complete sense in terms of current drain. The consist you see in the video does not trip 10 amps and will pull you off your feet if you try to hold the coupler.

Let's keep going and investigating.

Greg
 

·
Registered
Joined
·
60 Posts
Discussion Starter · #7 ·
I think the main issue is the Bs have Lokpilot 4xl decoders and the the As have Loksound 5xl. I can see just looking at the CVs that most of the motor and BEMF settings are different. My mistake for not starting with BEMF optimization and speed matching.
 

·
Super Modulator
Joined
·
22,477 Posts
Yeah, it's not uncommon for the default BEMF parameters and even momentum settings to be all over the map at stock. When you have an optimization routine in the decoder, I have found that the default BEMF parameters are even "weirder".

I'll be interested in your results, would suggest you check operations after each step to see what makes the largest difference.

Greg
 

·
Super Modulator
Joined
·
22,477 Posts
You know, it seems that NO manufacturer has consistent defaults for all these performance parameters.

At first I thought that it would be easiest to run no momentum, but then most sound decoders depend on some momentum to give you the best sound characteristics.

So that is yet another tuning parameter, and the best advice I have is when testing, let the speed "settle" at various speeds so the momentum is "done"...

This is all sort of unwritten learning stuff, and it's that double edged sword, the great thing about DCC is there are so many parameters to tune/customize... it is also one of the worst things about DCC. ;-)

Greg
 

·
Registered
Joined
·
60 Posts
Discussion Starter · #11 ·
But the learning is the fun part…

Some progress yesterday. It does appear to be the difference between the A and B units. I ran the same load with just an AA combo and it worked fine. I need to wait to get my speedometer to speed match the Bs.

The motor autotune function for BEMF on the Loksound 5XLs was no help at all. I tried tuning two LGB Amtrak Genesis and two USA Trains F7As and for each, the tune made it worse. I went from smooth low speed operation with the default settings to terrible stuttering at slow speeds on all 4 locos. I went back to the default settings, and they were fine again.

I can’t find much online about the autotune function. Most say it works well. Steve Weeks says he gets different results every time and doesn’t use it. I saw one place that the engines had to be really warmed up, which I didn’t do, but not sure after my first experience that I will try it again.
 

·
Super Modulator
Joined
·
22,477 Posts
Well that is disappointing, but you had to go through the exercise. I fail to see how "warmed up" would make a large difference, only if you were in subzero temps where the gearbox resistance was increased.

I have tried various ones over time and have likewise had mixed results. One thing I have learned is that manufacturers that make mostly HO equipment do not realize the "scale" of large scale, and often the current draw is beyond the limits of "auto tune" procedures.

I had similar issues with Pittman motors in Accucraft/AML locos, since they can draw huge stall currents.

The stuttering you saw was clearly the BEMF tuning, what I encountered when first using the large scale QSI decoders. I dug in and tweaked all 3 parameters and found a set that really improved the stock parameters. Reported back to the MFG and found my parameters in the next manual ha ha!

Anyway the 3 parameters PID, take a little understanding, but you can tweak them. Right now, you might indeed turn BEMF off and proceed with your speed matching.

Have you turned it off and seen how the locos run?

Greg
 

·
Registered
Joined
·
60 Posts
Discussion Starter · #13 ·
I haven’t but may if I continue to see issues. I’d like to keep it on if possible because my main lines all have two grades (it’s a folded double dog bone) and I’d like constant speed.

What did you set your PID to? I’d like to compare. Separately, do you know what motors are in the USAT Trains F3s? ESU has a table that gives settings but you have to know the motor type. I can’t find it online.

Here is the article from Ric Weeks, helped me a lot to understand what the parameters do: https://www.southwestdigital.co.uk/PDF/SWD LSV4 Motor back EMF Setup.pdf
 

·
Super Modulator
Joined
·
22,477 Posts
Ahh, you will most likely NOT get constant speed. That is an extreme case of the BEMF PID parameters, and not all decoders can accomplish this.

AirWire is one of the few systems that made it possible.

So, like many manufacturers, ESU has renamed the controls from the proper standard reference, and also appears to have only given you control of 2 of the 3 parameters in BEMF. This is typical unfortunately.

The first thing I would do is relate the ESU parameters to the proper PID parameters in BEMF.

I would only play with the higher speed parameters now, leave your low speed ones alone, so you should not re-encounter the shuddering/jerking.

I would encourage you to give up on the constant speed goal right now, it will be at odds with better running, in my experience.... and with a long enough train, it won't go much faster downhill, I have a 5.5% downgrade at one point and I am able to pick a speed and leave it running unattended, 3.4% up. This was with a 50 car freight.

Let's get your consists and speed matched first. I think the first CV is the "amount" of BEMF influence, and I would uncheck that first box to disable BEMF and see if things run.

Then set CV2 and your custom speed tables to match smph

Then try a consist and see how it runs.

Greg
 
1 - 15 of 15 Posts
Top