Consist - matching speed for the same models

[Arras88]

Hi, I searched forum and read a lot about matching speed for models which should run in a consist.
I understand the problem when one wants to match different models from different producers. There are few reasons why engines might run with different speed (decoder, motor, gears).

But, where is a problem when I have exact the same models?
I bought four NS Heritage SD70Ace Athearn models. My goal was to run them in a consist. This is what I like in American Railway most (to see multiple locomotives pulling a long set of cars).

Unfortunatelly,  EVERY model which I bought runs with different speed. It is impossible to run them together because differences are so huge that faster models derail slowest one. I put them on track in the order of speed (the fastest first, the slowest last). It works but only if I run forward. In reverse at least two models derail on the first curve.

All these engines should be the same. The only difference is the painting. All components are the same: motor, decoder, gears. I programmed the same CV to all engines.

Is the only solution to programm speed table to each model? This is a very time consuming process. I did it ones for two models and spent a whole day. I can imagine how much time I will spend for four models

I have all Union Pacific Heritage Aces which I bought 2 years ago. These models are from MTH and there is no problem at all. I can run 8 models together. Sometime I run a very long train where 6 engines are on the head pulling cars and 2 are at the end helping to push cars. Everything is ok because all engines travel with the same speed.

This time I wanted to try Athearn because they look much better and are very detailed. However, I am not able to run them in a consist.

Any suggestion from experienced users?

 

[richhotrain]

Yep, you would think that identical locos would at least run at similar speeds, if not identical speeds, but they don't.

I have a set of four Athearn Genesis Santa Fe Warbonnets, F7LABC.  They all run at different speeds.

This is where JMRI and Decoder Pro come in handy.

Rich

[rrinker]

 Are they ones with sound or not? If they are the sound ones, reset all the decoders first. I've never had two of anything, even different models, that would actually derail others because the speeds were so off (like putting an S1 switcher with a GP7 - without speed matching, one will spin its wheels, but they don't derail on me).

 Also do not trust the response right out of the box, they will need to run time to break in properly.

                --Randy

[mlehman]

DCS locos run OK but DCC ones not?

I know virtually nothing abiut MTH's DCS. I understand it's now supposedly somewhat "compatible" with DCC. Don't know what version of DCS you have, though.

Are you programming the DCC locos have trouble with the DCS system? Or with an actual DCC-spec system?

I agree with Randy. Reset those decoders and start over. Somehwere in there the CVs disagree. There's no way that the locos themselves have that much speed difference. And presuming the locos are running the same decoder, there's no way there's that much difference if the CVs are really set all the same.

[Arras88]

I don't use DCS. I have the Digitrax Super Chief set and use DCC. As far as I know, MTH models have something that is called "speed control" and it really works. I am talking about models with ProtoSound 3.0 decoders, not dcc ready.
I can start all my 10 MTH Aces with 1 inch gap between them and after running one loop on my layout (about 80 ft) they arrive with almost same gaps between models.

These new Aces are Athearn Genesis Sound equipped units with Tsunami decoders. The difference in the speed between the fastest and the slowest model is so huge that after running half a loop with 50% speed on the Digitrax throttle there is a 3 ft gap between them. When all models are together, faster models push so hard on the next models that trucks are being pushed out of the track on curves. As I wrote the only option for now is to run them in the order of speed (the fastest first, the slowest last). In this case each models pull others and they are able to run in the consist. But it works only for one direction because in reverse each model is being pushed by following faster models.
Even one of plastic couplers broke because of it.

I will reset all decoder tomorrow. Maybe I will be able to record a movie to show you how it looks.

Thanks,

[mlehman]

Tsunamis should be much, much closer than that if applied to the same loco and with the same CVs loaded. I've never had any perform like that and can only guess it's some CV setting difference issue.

Also note, if you have BEMF enabled in a consist, it can sometimes cause issues as one loco or another tries to match the speeds of the others in consist. I doubt the variation is as large as you describe, but that's another possibility here.

[Arras88]

All models after reset. None of CVs changed. Models run on the same default address 3.

The fasters #1070 Wabash and the slowest #1030 NS.

The initial gap is about 2". First, they run with speed 15% (Digitrax has no steeps but percentage). With this speed #1030 is a little faster. But when I increase speed over 30% #1070 run away.
On the movie they run with 50% speed. After one loop there is an over 3 feet distance between models.

http://www.youtube.com/watch?v=_SD9PC-2cqg

Second movie shows how #1070 gets derailed. In forward direction everything is ok. #1070 pull #1030.
In reverse, #1070 is much quicker and stronger. It push #1030 but because it is too heavy #1070's trucks jump out of the track. It happens mostly on curves.  Sometimes only two axles are dereiled, sometimes all truck.
If I add two more locos to the "train" and #1030 is in the middle also #1030 gets derailed.

http://www.youtube.com/watch?v=gDqbblaKTsk

It is quite annoying and I have no idea how to deal with it.

[rrinker]

Time to start playing with speed tables. At low speeds, you need to speed up the slower one, and at higher speeds, slow down the faster one. Tsunamis don't support the simple method of adjusting just 3 CVs, so you'll have to use a full speed table. Much easier using JMRI, if you have an interface for your system, either a PR3 or Locobuffer.

              --Randy

 

[peahrens]

I haven't graduated to DecoderPro yet, so I can imaging tinkering with the twenty-some speed table entries would be a pain.  Another case of Tsunami lacking CV5&CV6 being a bummer.

If the speed offsets between engines were linear with speed (I gather they are not), would not a simpler way (less CV settings) be to adjust all locos top speed to the slowest one using forward and reverse "trim" CVs 66 & 95?  Then see if the mid speeds were close enough.  To use these, I believe it is necessary to set CVs 29 (appropriate setting to "enable" speed tables) and CV25 (set to 2 to use straight line speed table) first.  The straight line speed table eliminates having to enter all the table values.   

I presume you have downloaded the Tsunami Diesel Users Guide Part 3 explanations.  Actually you should first turn off momentum, then set CV2s to get a reasonable match, then do the above with (in order) CV 29, 25, then trims 66 & 95.  Again, this will only match top speeds (each direction) and if the variation is too non-linear across the speed range it will not be adequate.  Might be worth a try.   

The folloing thread (last entry) describes how I used the above trim adjustments on a GP9 to match it to another type loco (a Kato RS2 with TCS motor decoder). 

http://cs.trains.com/mrr/f/744/t/218127.aspx

Disclaimer:  I'm pretty new to this stuff so hope I'm describing it correctly.  Considering a 2014 goal of tackling Decoderpro.

And, sorry the link above is not clickable (paste it into your browser).

[Arras88]

Thanks all for suggestions. I just watched a movie about Decoder Pro and JMRI. It seems to be a powerful tool. I will buy an interface and try to learn something new.

Yes, I downloaded manuals from Soundtraxx and I know the procedure for programming user speed table. I just wanted to avoid this time consuming procedure. Trim function wil not help here because difference in models speed is not linear. Until 30% the #1030 is even faster than #1070.

Until inteface arrive from US I will put tracks into a circle and connect prodigy advance with two handhelds for quicker adjusting 28 CV for user speed table. It will be fun

Thanks,