Do you have ...

No I don’t on jmri no a long

No I don’t on jmri no a long passing siding . I used my club’s tracks for this and no one is sure how do this, they usually use the same type decoders. what I did was tried adjusting them one at a time  and still can’t get close enough  in matching them

Worth a look...

Dear Casper,

Assuming each loco you're matching are already essentially tuned "standalone" for basic motor control behaviour,
and all locos are warmed-up so the comparisons are valid, this YT is worth a look...

Happy Modeling,
Aim to Improve,
Prof Klyzlr

Dear ???,

No one "on"s (presume you meant "own") JMRI.  It's free, and pretty much the best thing since sliced bread, if you want to actually manage many locos in DCC world.

Blair

You asked how

You ask how to speed match, so JMRI and a passing siding (or a yard with two empty tracks) is how*. That’s the simple “poor man’s” method for speed matching. You can also use fancy methods with a track speedometer and some loco rollers, but that takes money and some tech savvy to set up.

So get JMRI (it’s free) and learn how to use it. Also learn how to program the starting voltage, mid voltage, and max voltage in loco decoders.

Then set the two locos you want to speed match on two parallel tracks, consist them, and pick one to be your reference loco and the other loco to be your loco you will match to the reference loco (we will call this loco the TBM loco for to-be-matched).

Using JMRI and programming on the main, adjust the starting value (CV2), mid value (CV6), and the max value (CV5) of the TBM loco until it runs alongside the reference loco fairly closely. It doesn’t need to be exact, but it should be close.

*You can in a pinch also use a long empty stretch of single track. Put the two locos on the single track six feet apart. Be ready to hit emergency stop if the locos are about to collide (one is much faster than the other). Once you get them close to matched, they should start the same time, speed up similarly and stay the same distance apart throughout as you turn the throttle up from zero to max speed.

Speed matching

Thanks  guys going to put both loco’s on a single track than adjust again  . My big problem is cv 5  with  an esu and digitrax  . Digitrax is where my problem lies at I think .

CV5 = Big Problem

Dear Casper,

Pls elaborate, how is "CV5" (Max Speed) a problem?

Is it simply not-there, not obviously-responding,

or some other symtom?

Further, to properly start the diagnostic process, can you confirm the specific decoder model(s) being compared?

Want to help, but need more info in order to be helpful...

Happy Modelling, 

Aim to Improve, 

Prof Klyzlr 

Speedometer

It's well worth the $70. Set CV2 so it's just creeping, CV6 to go 30mph at SS63 and CV5 to go 60mph at SS126. Takes a couple minutes per loco to get them dead on. I set the fast inc/dec on a UWT to 63 for the speed matching, and keep DecoderPro up to change the CV values. I remove momentum and run each loco through the Accutrack II several times in each direction, once I'm happy with it, I put the momentum back on.

I really don't know why speed matching causes people so much trouble.

NMRA CVs

Sometimes depending on the decoder model CV may not be supported.  The NMRA only has four (4) "mandatory" cvs listed.  #1, primary address (short) # 7 version # (read only) #8 Manufacturer ID # (read only) and CV# 29 configuration data.  All others are recommended or optional.  JMRI is a great tool to have but not necessary to accomplish speed matching.  I recommend downloading the technical guides and instruction manuals for each decoder manufacturer and those guides will list the supported cv #s.  

4 minutes 40 seconds

Is how long it took me to speed match a locomotive, even with a few goobers (like forgetting I was in program track mode in DecoderPro). Not every one will be the same, but 5 minutes is a reasonable estimate.

EDIT: 7 minutes 15 seconds for an oddball QSI equipped Atlas locomotive.

Less than 10 minutes

I just have a oval of track I set up on my super large cutting mat that has measurements on it.  I put them a specified distance apart (generally 4 inches) and have on chase the "control" loco (the one I use to match all locos to), run it once around the track (unless it already catches up before a lap) and then make an adjustment on a separate programming track.  Once the adjustment is made, I line up the locos again and do another lap to see the results and continue on from there.  I matched seven locos in about an hour.

speed tables

You should be using speed tables.

To get close, find out minimum throttle setting to get moving.

If LOCO does not move at step one, increase it.

Do both locos for the consist.

Next do top speed, find fastest loco, lower the top speed setting down to match the 'other' loco.

Now using programmer, set all other speed in a straight line from lowest to highest.

Too complex

I don't agree, a speed table is for a very specific use case where you want a speed curve that's actually curved as in a logorithmic or an exponential curve. The need for such a speed curve is very rare.

A starting (CV2), mid (CV6), and max (CV5) voltage speed curve is a lot simpler and all you need -- just three values instead of 28. You can do a straight speed curve with just start and max, or move the middle of the speed curve up or down by setting mid to be more or less than half of max.

Putting in 3 values is a LOT simpler than putting in 28 values. And I challenge anyone to show me a use case where a loco with a three point speed curve isn't able to do 99% of what a complex 28 point speed curve will do.

If I want to do a straight curve, using start and max will do the same exact thing with just two CVs instead of 28.

Exactly, Joe

Unless you are doing a specific speed curve (like I do with the locos I use with ProtoThrottle), there is no point in doing all speed steps if you are just after a linear line. The Min, Mid and Max variables automatically create that straight line for you and you only need to match three variables.  

I don't agree Joe

Cannot use forward and reverse trim functions with Digitrax decoders to fine tune unless speed tables are used.

Load the same speed table into every decoder (best practice is to use decoders from the same manufacturer).

Designate one loco at the "standard." Adjust forward and reverse trim (if decoder has these functions) on every other loco until they are close in speed in each direction to the "standard." Done.

And I challenge anyone to

I have some Tsunami and Loksound 5 decoders that don't use CV 6... Speed tables required.

The Loksound 5 (factory installed in 2 Scaletrains SD40-2s) was a surprise to me.  Not sure why they came that way, and the 3rd SD40-2 from that run had a Loksound 5 DCC.

I was able to get the Tsunami units (factory installed in Genesis F units) to run acceptably well with Loksound 5 DCC equipped units, until I can get them changed out.

Speed curves

I can't get locomotives to behave consistently enough to the point where speed curves would matter. I'm sure there is a use case for them, like the ProtoThrottle example, but most of the time, CV2/5/6 is just fine.

I just did a QSI that didn't have a CV6,  just using CV5 gets it pretty darn close. I thought about speed tables, but they aren't necessary.

Bachmann EZ-Command decoders, OTOH, don't have a CV5 or a CV6, so they aren't going to be able to run with another brand of locos. Luckily, I have half a dozen or so of them, so they can run in their own consist. They seem to want to run 71mph at SS126, which is a bit fast for my 30/60 standard.

I really don't know why speed

Possibly because  not everybody  can wrap their gray matter around this stuff. Simply, not everybody is the same!  Everybody has their highlights in something they do through out their life, but, for how ever many people,  consisting, ( and I'm going to expand this a bit to include DCC in general,) is tuff to comprehend.  I will be the first in line to claim it's hard for me.  I have the Tsunami 2 manual, 80 pages. I have the TCS comprehensive programming guide, 30 pages.  I have an NCE PH PRO-R  system.  I have Decoder Pro. That is a lot! I can not go through each one of these, one, two or even three times and comprehend what is going on with this tuff. For those that can, I am extremely jealous! I did realize what I was getting into by going DCC, for I wanted the DCC features,  knowing full well what was ahead of me on the DCC learning curve.                         I did get several loco's consisted, (only two at a time,)  but it took me quite a while, and that was just the basics. All that other programing stuff, well, that will be a little later.....

Wrapping your head around

Speed matching just isn't that complex if you use CV 2/5/6. You're just setting a couple of CVs to match a specific speed or the "standard" locomotive. And as evidenced in this thread, there are several different approaches to doing it, some take a bit longer, but in most cases, you should end up with similar results.

Now, function mapping a Loksound v5, yeah, I could see where that could cause some brain meltage.

I haven't delved deep into the NCE manuals, as I am typically just operating, but I know with Digitrax, there is a bit of a learning curve, but the manuals spell everything out, have an index, etc. It takes 30 seconds to find anything in the manual, and it goes through the process step by step. It's not a matter of knowing everything, you never will, it's a matter of being able to use a manual or other reference material to quickly find what you want at that point in time. I keep a whole folder of manuals now for decoders and system components on my OneDrive, I don't often need to reference them, but when I do, I can find whatever I'm looking for usually in 30 seconds.

Also, DecoderPro makes programming SO much easier. You need to understand some basics, but for the most part, you no longer need to calculate CV values or look up charts and tables of CVs in manuals to do this or that.

Workflow

One of the trickier things to understand is the programming workflow. Once you realize that you can use POM with DecoderPro, you no longer need to use a throttle to program anything, and you can just let DecoderPro keep track of everything. Just don't try to use another machine to program something, as you have to re-write all the CVs, as DecoderPro doesn't keep track of what CVs were changed and saved to the roster, only what you've changed in that session before writing to the decoder.

I see a couple of ?

Disclaimer: I have not done any consisting. But have done some CV5/max on locos to get them at a close top speed. So that no one loco will run like a jack rabbit. Now for comments: First. Not all decoders have CV6. And some don't even have CV5. But I think all have CV2. And not all even have speed tables. Second. If you set CV2 at 2, CV5 at 200 and CV6 at 100 = you have a straight line. But if you change CV6 to 50 or 150, that line is not straight. So, what can be done, if you do have all 3 CV's available, you set CV2 for each loco for start; and then CV 5 on each to equalize. (Which is what I have done as mentioned above) But CV6 must be changed up and down to match for consisting. Or am I way off? And all this can be done without JMRI. In fact, it's what I originally did when all I had was the Zephyr. But as has been repeatedly suggested. Get JMRI. It's worth it. (Even if you have to buy a PR3/4, Locobuffer or SPROG)

true as far as it goes

That's entirely true as far as the statement goes, but if you do change CVs somewhere else, you can put the loco on the programming track and do Read Sheet or even Read All Sheets. That will read what is in the decoder back into DecoderPro, and then you can write that updated info into the roster.

If you do this, though, be careful! Some of the decoders will creep the loco a small amount (1/32"?) for each CV that they read. For most decoders that is no big deal. But for some of the newer ones that have lots and lots and LOTS of CVs, if you "Read All Sheets" and walk way, you may find that the loco has crept right off the end of your programming track! I learned this the hard way with a LokSound 4. My impression is that newer decoders that have approximately this behavior now alternate the creep direction, so that they stay more or less in one place. I think a LokSound 5 did that, but I'm not entirely sure. Of course, if you're reading back an Econami or some other simple decoder, there aren't enough CVs to cause any trouble.

CVs and DecoderPro

If you don't have CV6, you may be able to speed match pretty well using CV5 if the performance of the locomotive is relatively linear. If not, you'll have to break out the speed tables if they are available. My QSI doesn't have CV6, but does have tables. The EZ-Command decoders are rather hopeless, as they don't have CV5 or CV6. CV 2/5/6 is manageable without DecoderPro, but once you get into sound decoders, it becomes unmanageable really quickly, and DecoderPro is a lot faster to go through and standardize settings, at least for disabling DC operation and basic speed matching. Having a roster is great too, as it will show up on WiThrottle throttles, including function labels.

Except that I wanted to go the other direction. I had to re-write all CVs. Now I know to do it in real time. You are correct about the creeping, I have a pin solidly in the plywood at each end of my programming track. The Loksound v5 decoders supposedly have 2,035 CVs, and I think there are a few that aren't in the DecoderPro profile, so it takes about 2 hours to ingest into DP, whereas something like an EZ-Command, or even basic Digitrax decoders take anywhere from less than a minute to about 5 minutes.

Keep it simple

For speed matching, you don’t need to read all the CVs into decoder pro.  In fact if hooking up a computer, installing and learning more software is intimidating, you don’t need decoder pro at all.

What I used for quite some time back in the early days was just a pencil, a notebook, and my nce pro cab.  Whatever your dcc system is, if you can program on the main it makes things a little faster and easier, but it’s doable if you can only use the programming track too.  First step is just to read out what’s in the relevant cv’s, in this case 2, 5, 6.  Dedicate a page in the notebook to each loco.  Write down the starting point values in case you decide you made a mess of things and want to get back to where you started.

Then just tweak the values as others have described above, keeping a note about what you end up with for each value.

Next time you need to do a similar loco with the same decoder, you can do worse than just plugging those values in as a starting point.