Two decoders in a single engine...I f**ked up prgramming!

[Dave V]

OK, smart people...I need a lot of help. Like all the help.

I bought a Key Imports RGS C-17 #42 in HOn3 that came with DCC and sound. Like today...I got it today. It ran awesome out of the box. Smooth, momentum, all the stuff. It didn't have a keep-alive but it didn't stall at all. Backup lighting and lit marker lamps. Perfect with only two issues...the chuff rate was way off and the whistle was some throaty Big Boy-sounding whistle.

The accompanying note mentions a Tsunami 750 in the tender, and I know Tsunami pretty well...it's the only sound system I use.

So tonight, innocently, I parked it on the programming track and attempted to program the CV 120 for the whistle to change it. Well, rather than change the whiistle voice, it apparently changed the volume to almost zero. So I tried to restore the original value for CV 120. But now the speed table changed. So I tried to change 120 again, and suddenly every sound function (F0, F1, F2, F3, etc...) was replaced with coupler crash. WTF? Even the speed table was different. All from CV 120. So in a panic, I set the decoder back to factory default. When none of the CVs made any sense, I selected CV 8 and read the value... 129. That's not Soundtraxx, that's Digitrax.

And that's when I realized...that had I read the note more thoroughly, I'd have read that there's a DZ125 in the boiler for the lights and running.

Well sh!t. I reset BOTH of them. Shortly thereafter I began to realize I couldn't even consistently read CVs back, probably because my Zephyr couldn't distinguish between the two.

You can see the locomotive and tender are connected with a single plug and drawbar.

I did the obvious and tried to program JUST the locomotive or JUST the tender, but not surprisingly, it's single-rail wired, i.e., the locomotive pick up the right rail and the tender picks up the left.

So then I tried to hold the pin from the tender to the opposite railhead on the programming track, but that yielded all kinds of nonsensical readback values.

What can I do? I don't know how to reprogram each half of the locomotive separately. I might have been able to do it had I not reset both. Sigh. How can I reprogram them individually without cutting them out, or can I?

Please help!!!

[ Guests cannot view attachments ]

[ Guests cannot view attachments ]

[ Guests cannot view attachments ]

[nickelplate759]

Ouch! 

This may be something you already tried, but it looks like the loco and tender can be separated.  There appears to be at least one wire from the tender plugged in under the right side of the cab.  The drawbar may be another conductor, and it looks like it can be unscrewed beneath the cab.

If you can separate the engine and tender, you can program them separately.  You may need to use one or two jumpers to get power to the decoders while programming.

[Dave V]

Ouch! 

This may be something you already tried, but it looks like the loco and tender can be separated.  There appears to be at least one wire from the tender plugged in under the right side of the cab.  The drawbar may be another conductor, and it looks like it can be unscrewed beneath the cab.

If you can separate the engine and tender, you can program them separately.  You may need to use one or two jumpers to get power to the decoders while programming.

As I mentioned, I did try this, at least with the tender. I was unable to get anything with connecting the locomotive drawbar to the rail either.

[ednadolski]

Seems to me the only sure way is to open the model up, physically disconnect both decoders, separately reset/reprogram each one, and then put it all back together.   (Not sure tho exactly what you mean by single-rail wired...?)

Does either/both decoder have any kind of mfr-specific programming/addressing steps, e.g. you have to set some other CVs in order to address the CV that you actually want to change?

Ed

[peteski]

Programming on programming track (service mode) is always done  in a broadcast mode (so any and all decoders on the track at that time will be programmed.  Of course, when trying to read some CV, things get complicated, as either decoder can send the acknowledgment current pulse to the command station.

I seem to recall hearing of possibly using decoder lock CVs to be able to program each decoder separately, but I don't recall the details, and I might have been dreaming it.  Plus I'm not sure if DZ125 is even capable of being locked.

But I'm surprised that programming some CV you are familiar with (like CV120) would change whatever other CV is responsible for volume on the Tsunami.  CV write packets are sent over the programming track, and the data is written, even if the decoder does not sent back acknowledgement.

The only semi easy way I see reprogramming those decoders is to separate the loco and tender, then take a test lead with alligator clips to temporarily make the "other" track connection.  Don't just hold the "pin" to the track - that is too unreliable of a connection.

As far as you not  being successful with programming the decoder in the tender (the sound decoder), I suspect that migth be because it does not have motor connected, so it does not send the acknowledgement pulse back to the command station. So as I mentioned, the data is successfully being written to the decoder but the command station shows a failure due to lack of the acknowledgement pulse.  Try programming it despite of the failure being shown by the command station.

BTW, what DCC system do you use?

[Dave V]

Programming on programming track (service mode) is always done  in a broadcast mode (so any and all decoders on the track at that time will be programmed.  Of course, when trying to read some CV, things get complicated, as either decoder can send the acknowledgment current pulse to the command station.

I seem to recall hearing of possibly using decoder lock CVs to be able to program each decoder separately, but I don't recall the details, and I might have been dreaming it.  Plus I'm not sure if DZ125 is even capable of being locked.

But I'm surprised that programming some CV you are familiar with (like CV120) would change whatever other CV is responsible for volume on the Tsunami.  CV write packets are sent over the programming track, and the data is written, even if the decoder does not sent back acknowledgement.

The only semi easy way I see reprogramming those decoders is to separate the loco and tender, then take a test lead with alligator clips to temporarily make the "other" track connection.  Don't just hold the "pin" to the track - that is too unreliable of a connection.

As far as you not  being successful with programming the decoder in the tender (the sound decoder), I suspect that migth be because it does not have motor connected, so it does not send the acknowledgement pulse back to the command station. So as I mentioned, the data is successfully being written to the decoder but the command station shows a failure due to lack of the acknowledgement pulse.  Try programming it despite of the failure being shown by the command station.

BTW, what DCC system do you use?

Digitrax.

I was able to isolate just the engine using the lead, but no such luck with the tender. I followed your advice to program the CVs without the acknowledgement signal, but no luck. It doesn't seem to have taken all of the CVs. I had the motor lead connected to the opposite rail. I heard an audible click in the tender with those changes, but that's it. The volume is louder, so that changed I guess. But it's all messed up.

[oakcreekco]

I downloaded the Decoder manual from the Digitrax website.

It goes into detail about the "lock" feature and what's involved to program two Decoders in the same loco.

If both decoders have the lock feature, you can individually program each decoder separately on the same track.

Thanks to @peteski  I've been spending a lot of time fine tuning my HO locos for my switching layout.

I haven't used the lock feature, and mine are set to off, but from what I've been reading about it, it's a very handy tool in the Digitrax decoders for some applications.

[Dave V]

Update. Thanks, @peteski ! I've been able to restore it back to how it was when it got here (other than bell ring rate). Some of the CVs still won't line up though. Changing CV 120 does not change the whistle voice and CV 144 won't change the bell ring rate. And yet, I was able to correct the whistle volume with CV 129. I'm so, so confused...

But for now, I'm calling it a night.

[Dave V]

Update 2:

Well, I was reminded on Facebook that the CV for changing whistle voice for the older Tsunamis is 115, not 120. I was able to make that change with the locomotive intact. I now have the correct CV reference table to make the fine motor adjustments.

Thanks to all who helped, especially @peteski !!!

[ednadolski]

Great!   Now you gotta post a vid so we can see (and hear) it working!   

Ed

[jdcolombo]

Before I switched everything over to ESU LokSound, I used two decoders in my steam engines - one for motor control (almost always a Zimo) and one for sound (a Tsunami).  The way I solved the "how do I program them" problem is this:

I wired in the motor decoder first, and set its 2-digit address to something OTHER THAN 3.  Say, 5.  (Hint, write it down so you remember what it is).  Then I wired in the sound decoder and left its 2-digit address at 3.

Now I did my programming by using Ops Mode programming (programming on the main).  To program the motor decoder, I'd use address 5.  To program the sound decoder, I'd use address 3.  When done, I used Ops Mode programming to change the address of each of these to the engine number (4-digit addressing).

And here's the key part: if I ever wanted to re-program the individual decoders, all I had to do is use Ops Mode to set CV29 back to 2-digit addressing mode.  When I do that, each decoder responds to its original 2-digit address ONLY (using ops mode programming).  Make your changes, then set CV29 back to 4-digit mode, and you're done.

John C.

[John]

@jdcolombo  -- that's genius  !!!

[peteski]

I'm glad that you got things back under control Dave.  From what you are mentioning, the confusion was not strange CV behavior, but that it was a different model of Tsunami decoder using different CV arrangement.  Nice to be able to explain weird unexpected behavior.


John, those are pretty handy tricks, even if things can get a bit laborious when large number of sound decoder CVs have to be programmed (especially is indexed CVs are used).  I guess it also depends on the brand of the DCC system (or even if JMRI is in play).

[jdcolombo]

Hi Pete.

Yeah, Ops Mode programming is not what you want to do if you're setting 50 different CV's on an indexed decoder like a LokSound.  But most of the time with a 2-decoder arrangement, you're just either adjusting the 3-step speed table, or momentum response on the motor decoder, and volume levels (or maybe EQ, with the TSU) on the sounds with the sound decoder.  The Tsunami didn't use indexed CV's, so it really wasn't an issue.  I'd never try to do initial setup of a LokSound this way, but I don't have to because the LokSound has enough function outputs and superior motor control, so you don't need the "two decoder" technique with it.  The old first-gen TSU had both mediocre motor control and poor adjustment of the chuff rate, so that's why I ended up with two decoders in "the old days."  A single ESU LokSound is much better!

John C.

[Dave V]

Here it is, alive and well this evening. I was even able to swap-out the steamboat whistle for something closer to a D&RGW 3-chime:

Once again, thanks @peteski ! Your reputation as a DCC guru is very much earned, and I'm in your debt.