More DCC woes

[mmyers]

Motor contacts/brushes shorted to the frame at some point would make the engine run off at full speed. common problem with Atlas engines but it is possible with other brands as well. Old decoders shouldn't be an issue. I have some early DN93's from the mid 90's that work very well.
Martin Myers

[signalmaintainer]

"2"? "6"? Not a good advice.  Blindly setting CV29 is not helpful.  We are assuming things.

Bull. Assumption is on your end. Go to the Digitrax CV29 calculator and do the math yourself.

https://www.digitrax.com/support/cv/calculators/

[mecgp7]

"2"? "6"? Not a good advice.  Blindly setting CV29 is not helpful.  We are assuming things.

LIRR, factory defaults will leave the DC mode enabled (so the decoder will universally run on DC or DCC).  If you don't even want to deal with modifying CV values, I get the feeling that you really don't really want mess with DCC.  I'll mention again:  there is a learning curve to use DCC.

I get the same advice from you each time I post  issues with DCC. I have REPEATEDLY said all I want to do is test DCC locos period. I don't want to run DCC. I don't want to check CVs, etc. There are a total of three pages in the Digitrax manual on how to read an address and then run said loco with that address. Step by step instructions. Even I, a DCC illiterate like myself, can manage that much. It seems simple enough so where is the learning curve? The learning curve would come if I WANTED to run DCC. I don't. So, either my system is subpar or the decoders or both. I have tested a number of locos over the years  and they work fine when tested for address and running on the same address. I know how to do that much which is all I want. I learned to add the resistor to the programming track among other things so what part of the curve for TESTING DCC locos have I not learned?

I am looking for helpful suggestions and I have received some more tips through this post. I truly appreciate those who have reached out.

I also mentioned earlier that I am testing older, used DCC locos. My guess would be that if I plopped a brandy new recently manufactured DCC loco that things would work wonderfully.

[Bendtracker1]

Sounds to me that the motor is shorting out like others have said.
Take a close look at the motor tabs and make sure they are NOT touching the frame somewhere.

If that happens, the motor will see full track voltage and will do as a straight DC motor would always do, take off like a rocket.

Check the motor tabs first.
And yes, the chip could be bad as well.

[nickelplate759]

Sounds to me that the motor is shorting out like others have said.
Take a close look at the motor tabs and make sure they are touching the frame somewhere.

If that happens, the motor will see full track voltage and will do as a straight DC motor would always do, take off like a rocket.

Check the motor tabs first.
And yes, the chip could be bad as well.

Pretty sure you meant "make sure they are not touching the frame"

[Bendtracker1]

Pretty sure you meant "make sure they are not touching the frame"

Yup!   
Thanks!

[peteski]

Mecgp7,  testing only?  Even for testing purposes reading/writing CVs is part of the process.  When you write the address to the decoder, you are writing CV1, 17, 18 and 29.  Whether the command station writes them for you or you access them manually, they are being accessed and written to. 

Granted, if all you want to do is to perform a decoder reset (by writing some value like "8" to CV8), then just address it using short address "3" to  run it back and forth, then you can get by without any deeper knowledge of accessing CVs.  But when a problem crops up like in this example, having some basic knowledge of few basic CVs is helpful in troubleshooting.

Sure, the problem you described could be caused by hardware, but I doubt the motor leads are shorted t the track pickups.  DCC signal is AC and that signal would just cause the motor to buzz (not run at full speed).    But the alternative troubleshooting step provided to you was to disable the DC mode by changing the value of CV29. That could possibly take care of the problem if you were willing to do that.

I would also be curious as to why you are testing them (if as you mentioned, you are not planning on running them on DCC)?

[MK]

He's testing them to sell.  To make sure they work in DCC so he can price them accordingly.

[Bendtracker1]

Pete, you are correct about AC voltage, but....

If the chip happens to be a "dual mode" decoder and there is a short, a spike or some glitch, then there's a possibility that it'll fake the chip into thinking it's running on DC power.   At which point, since there's a constant 12 volts to the rails, it'll take off down the track with full power.
I've had this happen a few times, which is why I always set "Bit 2" in "CV29" to off, thus preventing it from seeing any DC power.

[peteski]

Allen, I agree that some glitch in the DCC track power can cause the decoder to misbehave as you described.  We have been describing this possibility since the 3rd post of this thread.  However I have doubts that shorting the decoder's motor leads  to track power pickups would cause this behavior. If anything, that might actually damage the decoder (like blow out the H-bridge transistors which drive the motor).

I guess the question that has not yet asked about that "Speedy Gonzales loco" is whether that is its permanent behavior, or it only happened once.  If the behavior is permanent, then I suspect that decoder is toast.  It would also be interesting to find out how it behaves under DC control (powered by variable voltage DC throttle).

As for decoders being dual-mode, that feature is part of the NMRA DCC specifications, and I have yet to encounter a decoder (of any brand) which would not be dual-mode (run under DCC and DC control).  And I have dealt with some rather old Digitrax decoders.

[John]

Before we get too far down a rabbit hole ..

Decoder reset values for various manufacturers

https://www.nmra.org/sites/default/files/sr201402_dcc.pdf

The value in ID in CV8 tells you manufacturer - the CV for reset + value has the CV and value to program for a decoder reset.

A more comprehensive list of instructions and hints may be found at

https://dccwiki.com/Decoder_Reset

And finally

https://dccwiki.com/Analog_Conversion


Personally, I think analog operation should be removed from decoders completely -- IMO it runs like crap -- if you want to run DC - then don't install a decoder .. or remove the one the one in the system ..

[mecgp7]

Personally, I think analog operation should be removed from decoders completely -- IMO it runs like crap -- if you want to run DC - then don't install a decoder .. or remove the one the one in the system ..

Amen! Not only do they run poorly, but I am paying a lot of money for a decoder I will never use.

Thanks to all who have offered suggestions. The misaligned motor contact was indeed an issue on one loco. My guess is these particular locos sat for a long time unused. I ran them in analog for a few trips and they were much more responsive to DCC. Perhaps oxidation was part of the issue and why some locos worked then didn't and then did again.

I have a Digitrax Zephyr DCS50. Is this part of the problem too? Do I need to upgrade?

[signalmaintainer]

https://dccwiki.com/Analog_Conversion
Personally, I think analog operation should be removed from decoders completely -- IMO it runs like crap -- if you want to run DC - then don't install a decoder .. or remove the one the one in the system ..

Right on!

[MK]

I have a Digitrax Zephyr DCS50. Is this part of the problem too? Do I need to upgrade?

No!  I still run mine and it has been flawless.  It's the only command station that I have.  No need to upgrade for what you are doing.

[peteski]

IMO, there is no need  to remove the analog operating mode. It is a very handy feature (for operating on foreign layout or even more so for quick pass/fail decoder testing - I use 9V battery).  Why remove it when it can be easily disabled by . . . changing a setting in . . . CV29.  As simple as that.

I'm also glad that (for the earlier mentioned testing reasons) it is enabled by default.   Think about it - removing useful functionality is never good.