Kato NW2 hesitating/surging with Digitrax

[PanAmShay]

Hi there,

Background. I have a Kato NW2 with factory Zimo DCC Sound I bought new 4 months ago.  It runs perfectly smoothly, and other than the sound and pulling power being a little anemic compared to our Atlas S2, it is a great loco.  It creeps along at 0.6 sMPH with a SpeedStick V5, and I have made no changes to any of the motor CVs.

I just picked up a ~14 year old Kato NW2 176-4370 with Digitrax DCC from someone who hasn't been able to run trains in a while. It looks prerfectly new, but doesn't run very well. 
This is my first experience with a Digitrax equipped loco (all others ESU or Zimo).  The slowest I can get it to run is 4 sMPH, and that uneven.  The light stays on and no flickering and I don't believe it is a power pickup issue.  At mid throttle it is OK, and at high throttle speeds it is also very surgy, speeding up and slowing down.  Pickup is good on both trucks, wheels are cleaned, I fully disassembled it and pulled out the motor and gave the areas light lube with Labelle 108.  The motor seemed to spin freely, but was a little less so when the clipped on spur gear/driveshaft was attached (almost like there was no free play between the gears).  The driveshaft has plastic spacers/bushings instead of the brass ones seen on Spookshow.https://www.spookshow.net/loco/files/katonw2gear.jpg 

Are there any parameters I should try to play with on the Digitrax decoder?  I am used to being able to read CVs on the main with our ESU Cab Control/Mobile Control Pro throttle.  It seems I have to create a programming track to read any motor CVs for this Digitrax one. 

Also, any known mechanical issues or tricks I should look for to improve low speed creep?

The throttle response is VERY non linear.  With 28 speed steps, I record the following speeds
Speed Step:
1-2 No movement
3: 5.23 sMPH (10 x the slowest speed of my other NW2), and it sort of slows down and speeds up, pulsing or cogging
4: 16.2 sMPH
9: 53.8 sMPH
14: 89.1 sMPH
19: 91.8 sMPH (surging)
28: 91.3 sMPH (surging)

With all my ESU locos, we just set up a linear speed map and adjust CV53 to set the max speed.

Obviously the top speed is way too fast and no idea why it would be so non-linear.  The goal for this loco is to have it be on a nTrak module at a show and used for industry switching, so reasonable slow speed creep is desired for good switching/ops, but I don't want to spend $100+ for a Zimo sound decoder.

Thanks!

[MK]

Try disabling BEMF.  Set CV56=0.  I believe the default is 6.

[PanAmShay]

MK, thanks for the suggestion.  setting CV 56 (was 80) or 57 (was 6) to 0 didn't help. 

Vstart, Vmid, and Vmax were all at 0.  Setting these to 26, 38, and 64 respectively helped dramatically to eliminate the non linear speed and set max speed to 32 sMPH.  However it didn't address low speed surging/hesitation.

I spoke to Kato and tried running it in DC only with some clips attached to the motor cups.  It actually ran better I think in DC with smoother motor running, though I wasn't able to test the min speed.

I then removed the drive shafts to the trucks.  In DC with leads to the motor cups, the motor runs smoothly at max velocity.  However when I put the decoder in (Still with the drive shafts removed), after resetting the decoder (CV8=8) I get the surging behavior at max velocity. 

All that I know so far...

[C62-2]

I've been using digitrax decoders in various Kato locomotives, since they're easier for me to obtain. The default CVs can cause things to be rough. I typically set Vstart to 0, Vmax to whatever makes a reasonable top speed for that loco, and Vmid to halfway in between. If you're using those three, you also need to make sure CV 29 is set to use them (see https://www.digitrax.com/support/cv/calculators/). I also set kickstart (cv65) to 0, and CV 56 to 48, and CV 57 to 5 (these seem to provide somewhat decent BEMF performance, but setting CV57 to 0 should remove BEMF entirely, so that should give performance equivalent to the direct DC performance, assuming the other values are reasonable.

[peteski]

The decoder could be faulty. With BEMF disabled, in my experience there should not be any surging, since the decoder is sending straight variable PWM voltage to the motor, not trying to compensate for anything,

[PanAmShay]

C62-2 - thanks for the info and suggestions.  I will try this out.  I had checked CV29 yesterday and it was at 38, which from the calculator shows Speed Table is disabled and should allow the use of CVs 2, 5, and 6 (from my limited understanding).

I will re-try the CV57=0 just to be sure, and then your other suggestions.

Peteski - good to know.  I wouldn't have expected it could be faulty in such a nuanced way.  I'll give Digitrax a call to get their take. 

I wonder if a Lokpilot 5 Micro 59826 would fit. The 7-8mm footprint is no issue, but the published 2.4mm height looks to be more than the Digitrax on there and more than the published 1.8mm height of the Zimo sound version (dims from Streamlined Backshop)

[PanAmShay]

I spoked to Kato USA as well and got some useful info.  They rarely see motors go bad on these units.  After further full disassembly and ensuring no gears are binding and everything is turning smoothly mechanically, I used 4 alligator clip wires to connect up an ESU 58741 decoder (1 to each of the side frames and 1 to each brush cup.  It was sort of unwieldy so I couldn't get it up to speed, but it did seem to have very good slow speed performance.  So I am giving up on the Digitrax decoder and now need to figure out will a lokpilot 59826 fit or should I bite the bullet and pay the extra $75 for the Zimo with sound...

[peteski]

Which Digitrax decoder is in that loco?  One of the DZ125 or the skinny DC125T (tiny) decoder?
Where is it mounted? Under the hood or in the cab?

The paperwork included with DZ125 shows it as 3.2mm thick (I measure 3.00mm) with the heat shrink sleeve.
The LokPilot 5 micro with heat shrink measures 3.2mm thick with the heat shrink sleeve.

Do you think that the 0.2mm will make a difference between fitting and not fitting?

When I need to find out vertical clearance for decoders I take a small ball of a modeling putty (Play-Dough woudl probably work too) and stick it on the spot on top of the chassis where the decoder will be installed. I then cover the putty with a small piece of Saran wrap (or some other verythin plastic film like from a grocery bag) so o the putty does nto stick to the loco shell. Then I place the shell back on the chassis and make sure it is fully seated. Then remove the shell and the putty shows me exactly how much headroom there is.  I use my caliper to measure the clearance.

If you don't have a caliper, I highly recommend getting a digital caliper. It is an indispensable modeling tool. Harbor Freight (and likley many other online vendors) have 6" calipers for around $20 dollars (no need  for the very expensive ones - in  my experience the cheap ones work quite well).

The ZIMO replacement board sound decoder would make a nice addition to this loc, but you have to add speaker and speaker enclosure, and likely also a small keep-alive unit to get uninterrupted sound.

[PanAmShay]

Peteski, It was a Digitrax DN123K3 made for the Kato NW2, measured at 1.8mm.  I also ran into a very similar issue yesterday with a DN136 (poor low speed hesitation/cogging and surging high speed performance that was installed in a Tomix Dr Yellow Shinkansen.  Both locos worked well when bench tested with a ESU Loksound as described with alligator clips.  So I just don't think Digitrax has the same motor control capabilities as I am used to with ESU and Zimo.  I had heard this generally, but never witnessed it first hand.

On the Dr Yellow, I replaced the DN136 with a Lokpilot Micro 6 wire and it worked fantastic.  Super steady creep--2 sMPH at speed step 1 and 290 sMPH at full throttle with a linear speed curve.

Good suggestion on digital calipers.  I have 4 Mitoyo 6", a Mitutoyo 8", and a Mitutoyo micrometer and I don't know what I'd do without them!

I could put a TCS drop in decoder into the NW2, but I think I'll just spring for the Zimo designed for it.  Trying to avoid learning a new decoder brand, and we don't have any TCS at this point.

[peteski]

 I pretty much  settled on ESU and ZIMO for all my DCC decoder needs.  None of all the American brands (Digitrax, TCS, NCE, and Soundtraxx) come close to those European brands.  Several years ago the prices of the European decoders was quite a bit higher than American ones, but the European decoders came down in price while American ones got more expensive, to  the point that they are almost the same.

[PanAmShay]

We started out (a year ago) only using ESU, and added 2 Zimo as required and have been very happy with both brands for motor control and sound.

The scale speed feedback with Railcom + and the brake slider are great too when using ESU throttles.

We just happened to end up with two locos with Digitrax decoders in them, and they are now gone.  One replaced with a Lokpilot and for the NW2 I'll order a Zimo tomorrow.

The lack of an affordable DC only option to convert to ESU on the otherwise fantastic looking Bachman Via Rail SCV-42 Charger, which is very expensive and TCS only, was a big deterrent.

[peteski]

Well, if you wanted it bad enough, the Bachmann Charger could be converted to another decoder, but since the Soundtrack decoder is so highly integrated into it, that would be a major hassle and expense.