New Loco - Minor Out-Of-The-Box Problem

[Dwight in Toronto]

I have posted this on a couple of sites, so apologies if you’re seeing it again.

My first experience with sound decoders was with two Rapido N scale FP9A ESU-equipped locos, acquired five months ago.

I just bought a new Kato N scale ES44AC with “Kobo” ESU LokSound, factory installed. 

My NCE Powercab display shows direction as FOR: but the loco moves in reverse (and if HEADLIGHT is pressed, the rear lights come on). And, of course the opposite also occurs … change direction to REV: and the loco moves forward with the front light on.

I checked CV29 thinking that perhaps bit 0 “reverse running” might be on, but it shows a value of 34 (bit 1 and bit 5 are “on”, meaning that only 28/128 speed step, and long address, are enabled).

Might anyone care to help troubleshoot what appears to be a minor mixup, please & thanks?

[peteski]

Set the reverse direction bit in CV29.  That will fix the problem.
If both the headlight and motor are reversed then it seems like the only possibility is that the decoder was wired in backwards.
I assume the loco is not in any consist, and that you are addressing it directly by its address.
What is the value of CV29?

The person who does the Kobo factory installs for Kato is a member of this forum.  I suspect that if you changed the subject line of your post to be more descriptive (something like  "New Kato Kobo DCC Loco - runs backwards"), will like get him to look in this thread (and likely reply).

[eja]

Peteski ...

Your contributions to this web site are invaluable.. This is yet another example..

Thank you .

eja

[Dwight in Toronto]

Thanks Peteski, I was hoping that you would chime in on this. I see that you kindly responded on the ESU io forum as well, so thanks for that.  I was also hoping that Kelley might see this, but never thought about making the topic heading more ‘click-bait’ for his eyes, as you suggested.

This also means that you have probably seen that the problem just as mysteriously vanished overnight. The only difference in ‘operating sessions’ on my test loop of Kato Unitrack was that I flipped the loco around on Day 2 (and not because I was curious if that would help … it was just, well, because!).

I can see no rhyme or reason why it would just as inexplicably correct itself.

Fwiw, CV29 = 34 (only bit 1 and bit 5 are “on”, meaning that only “28/128 speed step”, and “long address”, are enabled).

[reinhardtjh]

The person who does used to do the Kobo factory installs for Kato is a member of this forum.  I suspect that if you changed the subject line of your post to be more descriptive (something like  "New Kato Kobo DCC Loco - runs backwards"), will like get him to look in this thread (and likely reply).

I was also hoping that Kelley might see this, but never thought about making the topic heading more ‘click-bait’ for his eyes, as you suggested.

Well, you can always  do the obvious thing like putting an @nightmare0331 in the post which will set a "mention" flag when he logs in next.

And @peteski  I fixed your comment.  Kelly left Kato some time ago.  He is now a Scaletrains employee (IIRC) though I don't know his responsibilities there.  He may not have even done this ES44 depending on how long ago Kato released it.  Unless Dwight bought it through Lombard Hobbies, who Kelly has been doing a lot of add-on work for from what I see on his Facebook page.

[peteski]

Thanks Peteski, I was hoping that you would chime in on this. I see that you kindly responded on the ESU io forum as well, so thanks for that.  I was also hoping that Kelley might see this, but never thought about making the topic heading more ‘click-bait’ for his eyes, as you suggested.

Funny how we think differently.
To me your current "vague"  subject line is a click-bait.  It doen't mention the brand of the loco, decoder, or the problem.  TO me the more descriptive the subject line is, the more on-target replies will be received.  At least that is how I format my subject lines and I have good luck with them.

This also means that you have probably seen that the problem just as mysteriously vanished overnight. The only difference in ‘operating sessions’ on my test loop of Kato Unitrack was that I flipped the loco around on Day 2 (and not because I was curious if that would help … it was just, well, because!).

I can see no rhyme or reason why it would just as inexplicably correct itself.

Fwiw, CV29 = 34 (only bit 1 and bit 5 are “on”, meaning that only “28/128 speed step”, and “long address”, are enabled).

Of course it goes without saying that this was the value of CV29 when you were experiencing the problem?

Just like I mentioned in the LokSound forum: The ""overnight fix" seems to indicate that the decoder experienced a "glitch". Stuff like this unfortunately happens in fairly complex consumer grade electronic computer circuits.  Could be blamed on cosmic rays. Seriously!  Here is an example: https://www.eesemi.com/soft-error.htm  While this is somethign that happens to dynamic RAM, the fact is that "glitches" of all sorts can affect any highly integrated electronic circuit.  it is very rare, but it can happen.

I've been dealing with electronic circuits (on professional and hobby levels) for several decades, and I know for a fact that unexplained does happen.

[peteski]

Well, you can always  do the obvious thing like putting an @nightmare0331 in the post which will set a "mention" flag when he logs in next.

And @peteski  I fixed your comment.  Kelly left Kato some time ago.  He is now a Scaletrains employee (IIRC) though I don't know his responsibilities there.  He may not have even done this ES44 depending on how long ago Kato released it.  Unless Dwight bought it through Lombard Hobbies, who Kelly has been doing a lot of add-on work for from what I see on his Facebook page.

Yes, I know that Kelly left Kato, and that was why I didn't want to blatantly call him out here.  Regardless of whether he does the installs for Kato or Lombard, I would think that he was still familiar with the installs.  But the problem magically resolved itself, so all is well in the DCC universe again.

[Steveruger45]

If the cause was a spurious glitch having a snubber or two installed on the layout  could help.
I have two (one at each end) on my layout after something similar happened to me a couple or more years ago. Had no further spurious issues since.

[Dwight in Toronto]

Guess I’m not out of the woods with this new loco yet - weird things happened again today.

Was running in consist with a Kato AC4400 (TCS non-sound decoder).  Every time I hit on an odd numbered speed step, the lights would go out. Example - lights on at idle … accelerate with speed 1 … lights go off. Toggle ‘Headlight’ - lights illuminate momentarily but won’t stay on. Accelerate to 2 …. toggle Headlight … lights come on and stay on. Accelerate to 3 … lights go out and stay out. 

If I use the Fast accel button, where the increments are 4 steps each time, the lights stay on, since the multiples are all even numbers (4, 8, 12, 16). If I then try a single speed step, either faster or slower, the speed is now an odd number and the lights go out/stay out.

I cleared the consist and the loco ran fine in solo mode. I reformed the consist, and everything worked properly.  Late in the day I turned on the PowerCab once more, and the goofy lighting problem was back.

I just don’t get it guys!

[peteski]

Guess I’m not out of the woods with this new loco yet - weird things happened again today.

Was running in consist with a Kato AC4400 (TCS non-sound decoder).  Every time I hit on an odd numbered speed step, the lights would go out. Example - lights on at idle … accelerate with speed 1 … lights go off. Toggle ‘Headlight’ - lights illuminate momentarily but won’t stay on. Accelerate to 2 …. toggle Headlight … lights come on and stay on. Accelerate to 3 … lights go out and stay out. 

If I use the Fast accel button, where the increments are 4 steps each time, the lights stay on, since the multiples are all even numbers (4, 8, 12, 16). If I then try a single speed step, either faster or slower, the speed is now an odd number and the lights go out/stay out.

I cleared the consist and the loco ran fine in solo mode. I reformed the consist, and everything worked properly.  Late in the day I turned on the PowerCab once more, and the goofy lighting problem was back.

I just don’t get it guys!

Welcome to the wonderful world of electronic (computer) train control.   
As I mentioned in the LokSound forum:
https://sites.google.com/site/markgurries/dcc-welcome-page/dcc-problems/decoder-problems/headlight-blinks-with-speed-changes
DCC is working as designed - this is a user error/misconfiguration. 

[eja]

If the cause was a spurious glitch having a snubber or two installed on the layout  could help.
I have two (one at each end) on my layout after something similar happened to me a couple or more years ago. Had no further spurious issues since.

Can someone with a EE degree, or significant experience, please explain how  this circuit might eliminate " spurious issues ". 

I recall several years ago someone had a similar circuit at each end of a helix for similar reason, but with no explanation why or how it would help.

Curious minds .....

[peteski]

Can someone with a EE degree, or significant experience, please explain how  this circuit might eliminate " spurious issues ". 

I recall several years ago someone had a similar circuit at each end of a helix for similar reason, but with no explanation why or how it would help.

Curious minds .....

Mostly it absorbs signal reflections and spikes at the raising and falling edges of the DCC signal. Those can confuse the decoders "brain".
For slightly more technical explanation see
https://www.arrow.com/en/research-and-events/articles/what-is-a-snubber
https://dccwiki.com/Bus_Termination 
https://ncedcc.zendesk.com/hc/en-us/articles/204128735-Snubber-RC-Filter
https://sites.google.com/site/markgurries/dcc-welcome-page/dcc-best-practices/new-large-layout-wiring/snubberrc-filter

[eja]

Mostly it absorbs signal reflections and spikes at the raising and falling edges of the DCC signal. Those can confuse the decoders "brain".
For slightly more technical explanation see
https://www.arrow.com/en/research-and-events/articles/what-is-a-snubber
https://dccwiki.com/Bus_Termination 
https://ncedcc.zendesk.com/hc/en-us/articles/204128735-Snubber-RC-Filter
https://sites.google.com/site/markgurries/dcc-welcome-page/dcc-best-practices/new-large-layout-wiring/snubberrc-filter

Thank you for the technical references .....

[Dwight in Toronto]

Although the 14/28/128 mismatch symptoms perfectly describe what I was seeing with my new Kato/Kobo N scale loco lights, I don’t think that is the problem. I checked both locos, and CV29 = 34 for both … bit 1 and bit 5 are ‘on’ (28/128 speedstep enabled, and long address active).

I cleared the consist, confirmed proper solo operation, reformed the consist, and things are working properly once more.  Experience suggests this is temporary.  Very strange.

Also, when I push F8, the loco always initiates the startup sequence. However, pushing F8 to shut down the engine sounds rarely results in the shutdown sequence … 9 times out of 10, the sound just immediately turns off.  Suggestions?
 

[peteski]

I looked over this thread and we are now onto troubleshooting the 2nd problem (as the first problem of reversed direction fixed itself), correct?

The 14/28-128 speed step selection has to match both on the decoder and on the throttle.  Are you sure the throttle speed step setting is correct .  Not sure how this is done in the consist (whether the consist inherits the throttle setting for each individual loco, or is there a consist speed step setting). I have not done much of NCE consisting (as I almost exclusively use the system for testing single locos).

As for F8 (it is a "silent" or "mute" function in American decoders), that is a latching function in the throttle.  In my experience when the F8 is off, the decoder is silent, and when it is  on, the decoder makes engine noises.  The transition from off to on will cause startup, and from on to off, shutdown.  Or maybe I have no clue what I'm talking about.   

And attempting some humor I do have to say that in a short time you have come up with a series of some unusual DCC issues.  I hope you're not jinxed.