Speed matching Digitrax and Tsunami2?

[spookshow]

I have 3 N scale Kato E8 diesels, all of which ran identically prior to decoder installation. I just installed Digitrax DN167K0A decoders in two of them and a Tsunami2 TSU-KN1 sound decoder in the third one. The Tsunami2 E8 runs noticeably faster below speed step 8 than do the Digitrax E8s (and then slower above speed step 8 ). The default settings for vstart, vmid and vhigh in the Tsunami2 are all 0's and the custom speed table is disabled in CV29 and CV25. The default settings for vstart, vmid and vhigh in the Digitrax decoders are 0, 127 and 255 (respectively). I tried setting CV2, CV6 and CV5 in the Tsunami2 to match the settings in the Digitrax decoders, but that didn't seem to make any difference .

I don't really want to speed up my Digitrax units to match the Tsunami2 unit, so is there any way to slow that puppy down?

Thanks,
-Mark

[peteski]

Are you saying that settings the 3-step table of Tsunami decoders does not seem to make any difference in their actual speed?  That's weird.  I have not messed around much with those decoders to have an answer, but those CVs are standard NMRA DCC standard defined CVs, so they should  work.  The CV values you might use in those might be different than in Digitrax, but they should have an effect on the model's speed.

I'm curious as to why didn't uou go with ESU LokSound decoders?

[spookshow]

Well, changing vmid and vhigh didn't change the starting speed, which I guess makes sense. But since vstart was already 0, I couldn't make it any lower. I was just wondering if, in this world of overbloated decoder functionality, there was some other CV that might impact any of this?

I didn't buy ESU because I have no opinions, allegiances, or "Ford vs Chevy" loyalties to decoder brands. The fact is, in this day and age of factory-equipped DCC, I rarely find myself in the position of needing to purchase and install a decoder.

-Mark

[MK]

This may sound dumb but maybe factory reset the Tsunami and try the three CVs again?  Who knows what is hanging them up.

[peteski]

Is the speed matched sufficiently at speed step 1?  Is the top speed matched well (CV5)? If so, and you want to use the 3-step table then maybe messing around CV6 (to make the speed cure non-linear) making it run slower in the lower range of speed steps will work well enough.  If not, you might have have to use the full 28 step curve to get better match with Digitrax.

I'm also curious if there is any difference if you set CV2=1 (not zero).

The other thing that makes a difference is the BEMF settings. Again, not familiar with Tsunami enough to make recommendations.

[spookshow]

At speed step 1, the Tsunami takes off like a jackrabbit and the Digitrax locos don't even move. It takes speed step 2 to start the Digitrax locos creeping along, at which point the Tsunami is already several miles down the track. Setting CV2=1 didn't result in any noticeable changes.

I tried setting CV2, 5 and 6 all to 255 just to see if they were being used at all and the loco took off at max throttle at step 1, so it does use them. But for whatever reason, the speed at step 1 with CV2=0 is much too fast.

I'd just as soon not do a decoder reset just yet as I have a lot of configurations for sound that I'd have to re-do.

-Mark

[peteski]

If the speed step 1 speed is too fast then as I see it the only option is to mess around with BEMF settings (I don't see the 28-step would making a difference if the step 1 speed with CV2=0 is already too fast).  Or get another sound decoder. 

[spookshow]

It's also OK to say "I don't know". I was actually hoping for help from someone who has experience with these decoders.

-Mark

[OldEastRR]

Yeah, different decoders in consisted engines make model railroading a headache. I've tried to match a Kato AB E8 set, each with different decoders, and after an hour I gave up with "close enough". One will be dragging/pushing the other slightly now but at least they RUN together.

[spookshow]

I don't need them to be perfectly matched, I'm just a bit annoyed that I can't make the Tsunami engine run slower. The engine itself is certainly capable of it.

-Mark

[Sokramiketes]

 

[spookshow]

If you read what you pasted, that only applies to custom speed tables, which I am not using.

-Mark

[spookshow]

I disabled BEMF (CV217=0) and put CV2, CV5 and CV6 back to 0 and that improved things quite a bit. I think the relative speeds of the 3 locos are now close enough that I should be able to run them without any wheel slip/squeak up to speed 4 (which is about as fast as I'll never need to run them).

-Mark

[peteski]

Good to know that my recommendation made a appreciable difference.  If you set CV2, 5, 5, and 6 to zero, you don't have any restrictions on the top speed.  But it you are happy with that, life is good!

[Maletrain]

I am by no means a speed matching guru, but I have recently found some things by experience that might help.

1.  For some reason, with the TCS decoders I am working with, it takes very large changes in the values of CV5 to get any change in the top speed, and then it makes big jumps.  But, finer adjustment is available with the top speed CV94 in the 28-step speed table.

2.  Programming the 28 step table and looking at the changes in speed with each step, starting with a "match ends" straight line from JMRI, I found that there was a drop in speed at a CV value of 127, no matter which speed step that occurs in.  I eventually found that BEMF cuts out at a CV value of 127 on the TCS decoders I was working with, even though their manuals say that CV10 sets the cutoff value from 1 to 127 and leaves it always on if CV10 is set to zero.  But, that is not correct, the BEMF turns off at CV value of 127 when CV10 is set to zero.  To make things simple, I decreased my top speed to in CV94 to a value of 126, which is still 120 scale mph in my loco.  Prototype max speed for that model is only 75 mph.

3.  I programmed the 28 step table in 28 step mode, and the loco did not move at step 1 of 28 with the CV67 value for the lowest speed step set at zero - the lights just flashed rapidly.  But, putting the command station into 128 step mode made the loco crawl at 4.35 inches/minute on step 1 of 128.  However, it only slightly increased up to step 3 of 128, then did not move and rapidly flashed its lights on step 4 of 128, then moved at 4.35 inches/minute on step 5 of 128.  Making the lowest step CV67 value 1 instead of zero got rid of the stall on step 4, but still gave a wonky speed graph for steps 1 through 6 or so in 128 mode.  I settled on CV67 value of 4 to make at least a monotonically increasing speed without skipping higher speed increments around steps 5 or so out of 128.

4.  Also of note is that the locos ran at different speeds on the same speed step, depending on whether they are in forward or reverse.  Because I want to run things like ABA sets with one or more engines running in opposite directions from the lead, the first thing I did was use the reverse trim CV66 and forward trim CV95 to make each run in the same speed in both directions, even though the speeds were different from unit to unit.

You could try to slow down or speed up a loco using the forward and reverse trims to match another loco bottom speed, but I am not sure that won't also be wonky at the lower end.  I have not tried to do that (yet?).

My experience with wonkiness/not-what-the-manual-says stuff is so far limited to oldish TCS decoders.  I will not be surprised if there are similar, but not identical issues with other brands.  I am trying to match some of a menagerie of Club locos with old decoders, and am concerned that trying to match locos with different quirks in decoders from different manufacturers may be exasperating if not impossible.

So, my advice if choosing the decoders to put into a set of locos is to try to stick with a single decoder brand and even the same version for locos you intend to speed match.