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

[Dwight in Toronto]

Thanks Peteski, always good to hear your thoughts.  Yes, the unusual forward/reverse flip-around has not reappeared (fingers crossed).

I haven’t specifically checked my PowerCab throttle itself, but it must be set up for 28/128 because I’ve been using it for other consists, with other locos, without ever having seen that odd headlight behaviour before now.

I’m between layouts, so all of this is occurring on a 15’ oval of Kato Unitrack … no turnouts, no reversing loops, no crossovers.  This is about as simple as it gets, and it’s got me wondering if I want to continue down this sound decoder path. Things were much simpler with the less full featured microelectronics.

[peteski]

Thanks Peteski, always good to hear your thoughts.  Yes, the unusual forward/reverse flip-around has not reappeared (fingers crossed).

I haven’t specifically checked my PowerCab throttle itself, but it must be set up for 28/128 because I’ve been using it for other consists, with other locos, without ever having seen that odd headlight behaviour before now.

I’m between layouts, so all of this is occurring on a 15’ oval of Kato Unitrack … no turnouts, no reversing loops, no crossovers.  This is about as simple as it gets, and it’s got me wondering if I want to continue down this sound decoder path. Things were much simpler with the less full featured microelectronics.

I would have to check on my Power Cab, but the number of steps might not be a global throttle setting but a separate parameter per each loco in the Power Cab's stack of locos.  Like I said, I'm not sure.

As for reliability of ESU sound decoders, there are many users out there (even on this forum) who don't seem to have problems with them.  But these decoders (especially V5) are very complex and very configurable, which increases the odds of something not working quite as expected. 

If you want to replicate the headlight flashing when changing speed, just set bit 1 in CV29 for 14 speed operation, and keep the throttle set for 28/128 steps.  Actually, was the loco that had the flashing headlights the same loco that was running in the wrong direction?   If yes, both problems can be contributed to the wonky Cv29.  If bit 0 (direction) and bit 1 (speed steps) were misbehaving, that would explain the strange behavior of the loco.

[Mike C]

I would have to check on my Power Cab, but the number of steps might not be a global throttle setting but a separate parameter per each loco in the Power Cab's stack of locos.  Like I said, I'm not sure.

. Pete, it's not a global throttle setting, it's a loco by loco setting . Mike

[peteski]

. Pete, it's not a global throttle setting, it's a loco by loco setting . Mike

Thanks Mike.  That makes sense since you could be controlling locos set up for either number of speed steps.  That also likely means that when locos are consisted on the Power Cab, the consist will preserve the number of steps setting for each loco in the consist, and send the speed step commands compatible with the specific loco.

[Dwight in Toronto]

After several days of trouble-free operation with that new consist as described in the original posting, I fired up the PowerCab earlier today and that frustrating lighting problem is back … the headlights extinguish every time an odd-numbered speed step is entered. I give up.

[peteski]

After several days of trouble-free operation with that new consist as described in the original posting, I fired up the PowerCab earlier today and that frustrating lighting problem is back … the headlights extinguish every time an odd-numbered speed step is entered. I give up.

Dwigth, nto t state the obvious, but with thousands of Power Cab users and thousands of ESU decoder users, you seem to be very unlucky for some reason.

Serisously, we need to troubleshoot this logically.  The flashing headlight thing happens to a single loco, and only when consisted using the NCE (advanced) consist?  Correct? 

While still consisted, use the programming track mode (with only one loco at a time on the programming track) and read CV29 in the misbehaving loco and also CV29 of the properly working loco also from that consist. Report your findings. Don't make any other changes.

[Dwight in Toronto]

Thanks for the hand-holding Peteski!

With that most recent recurrence, I again took the loco out of consist and confirmed that the lights were fine when running solo. I put it back into (NEC Advanced) consist, and it is once again operating properly.

The next time it acts up, I will troubleshoot as you’ve instructed and will report back. Thanks again!

[Dwight in Toronto]

Peteski - several hours ago, I reported that all was fine. I did a quick check just now, and the problem is back.

I left both locos in consist. I checked them one at a time on the programming track. Both locos have CV29 = 034.

[Dwight in Toronto]

Peteski, I know that you have followed my similar topic on the io groups forum. Matt Herman of ESU said that he has seen the very same thing in their office, and asked me to check CV49 … the value was 019. I think he wants me to turn an auto-checking feature off by changing CV49 to “1”, but I am awaiting clarification.

[peteski]

Peteski, I know that you have followed my similar topic on the io groups forum. Matt Herman of ESU said that he has seen the very same thing in their office, and asked me to check CV49 … the value was 019. I think he wants me to turn an auto-checking feature off by changing CV49 to “1”, but I am awaiting clarification.

So CV29 shows that 14 speed steps mode is not enabled. Just to make sure, when you checked CV29 in both working and misbehaving locos, they were still in the consist?  That is important.  That's because the consisting process modifies CVs (POM programming) in the decoder,  so CV29 might be affected without you realizing.

Too bad I don't live closer to you. I could connect my DCC packer analyzer to the track and see exactly what packets are sent to that loco.
I thought or another thing to try:

The speed step issue is not something I have to deal with often. I brushed up on Mark Gurries website, and revisited the NCE Power Cab manual.
As it has been said on the groups.io, Power Cab does not support 14 speeds steps at at all.  Only 28 and 128.

I'm thinking of the following procedure to try.

1. Break up and delete the consist.

2. Purge the loco recall memory in Power Cab.  Not sure how many slots you have configured, but just select several non existing loco addresses until your original locos no longer show up in the recall rotation.

3.  Select (enter) both of the locos in the misbehaving consist, so they are back in the recall memory.  For each loco use the F12 button to make sure it is in 128 speed step mode.  Basically for each loco (when it is active on the throttle) press F12 and choose 128 speed steps, then press Enter.  See bottom of page 17 in the Power Cab manual for details. That way both individual locos will be sent 128 speed step packets.

4. Consist both of those locos and see if the problem persists. If the problem is gone - great!

5. If the problem persists, while the consist address is active, use the F12 key again to set the cosist to 128 speed steps.  Is the problem still present?

[Dwight in Toronto]

Thanks for hanging in there with me.  Yes, I am hereby confirming that both locos were still in consist when I individually checked their CV29 values (I had to use Programming Track mode, since POM doesn’t read back cv values).

In the meantime, I googled ESU cv49, to better familiarize myself with Matt’s suggestions. The value 19 is ESU’s default, which includes enablement of the speed step auto-checking feature.  I’m going to try cv49=3, which turns the auto check off (but retains bits 0 & 1 “on” … back emf and 40 kHz motor pulse width enabled). Will report back shortly!

[Dwight in Toronto]

Good day Peteski. If you recently checked the i/o groups topic, you will see that I tried changing cv49 from default 19 to a value of 3. This is supposed to disable the auto speed step checking feature, while maintaining back emf and 40 kHz motor pwm.

I am happy to say that the limited running time accumulated since making the change has been trouble-free, and my gut feel is that the unusual headlight problem might well have been resolved.

I’ll report back here if/when something odd crops back up.  In the meantime, thank you so much for taking such a hands-on interest, and for helping out this sound decoder noob … much much appreciated!

[peteski]

Good day Peteski. If you recently checked the i/o groups topic, you will see that I tried changing cv49 from default 19 to a value of 3. This is supposed to disable the auto speed step checking feature, while maintaining back emf and 40 kHz motor pwm.

I am happy to say that the limited running time accumulated since making the change has been trouble-free, and my gut feel is that the unusual headlight problem might well have been resolved.

I’ll report back here if/when something odd crops back up.  In the meantime, thank you so much for taking such a hands-on interest, and for helping out this sound decoder noob … much much appreciated!

I hope this is a good news.  I would still like to know what is the other loco in the consist? The one which doesn't have a problem.  What brand/model of decoder is installed?

Even if the flashing headlight issue is resolved, that still doesn't explain why that loco also had a problem with running in reverse from what the throttle indicated.  I know I chalked that to being a glitch, but I'm wondering if it is somehow related to the flashing headlight (speed step mismatch) issue.  Did the "reverse running" problem ever reappear, or was that a one-time occurrence?

[Dwight in Toronto]

The other loco is a Kato AC4400 with a drop-in, non-sound TCS decoder … the first dcc install that I ever did, maybe 12/13 years ago.

The reverse-running problem mysteriously disappeared when I tried that loco on ‘day 2’, and has never come back.

[peteski]

The other loco is a Kato AC4400 with a drop-in, non-sound TCS decoder … the first dcc install that I ever did, maybe 12/13 years ago.

The reverse-running problem mysteriously disappeared when I tried that loco on ‘day 2’, and has never come back.

Thanks for the info.  I'm glad that the problem seems resolved.  I'm still puzzled at the behavior.  There are thousands of ESU V5 decoders out there and there must be many modelers who run them on NCE Power Cab DCC system.

According to the V5 manual, the factory default value of CV49 is 19 (decimal) which results in enabling automatic speed step detection and enable BEMF. The reserved bit 1 also gets set but I'm not sure why (since it is reserved).

Default value for CV29 is 12 (decimal) which breaks down to to normal direction, 14 steps, enable Analog, enable RailCom.

The above defaults are for a blank decoder, but then I believe that the the sound projects can have its default values different than the factory default values.  ESU LokSound is a complex decoder and sometimes that can result in seemingly strange problems.  But as long as now everything works as expected, all is well in the DCC world.