Author Topic: Decoder goes dead for no reason  (Read 3335 times)

0 Members and 1 Guest are viewing this topic.

mmyers

  • The Pitt
  • Crew
  • *
  • Posts: 997
  • Respect: +47
Re: Decoder goes dead for no reason
« Reply #15 on: April 08, 2017, 03:17:22 PM »
0
On the RDC's, I filed two clearance notches in those brass strips that run the length of the floor on either side. The notches are filed where the motor contacts come up. one on each side. The floor already has the squared notches, just make the brass strips match the floor where the motor contacts come up.
This lets them pass the brass strips without making contact. Gray and Orange wires are soldered to the contacts.
I also Kapton taped the four brass strips- two at each end - so that they continue to provide spring pressure on the truck pick ups. The tape insulates the four strips from the light boards so the functions are isolated from the rails.
White wire goes to both boards (W) mark. Yellow wire to both boards (Y) mark. Blue to both boards (B) mark. Screws are removed from each of the holes. IIRC, there are two to get removed in each board. One screw  goes back into each board where the holes are marked "DCC".
My installation veered from this by using a four function decoder. The extra two functions control the rear light boards so that I can have independant control of the lights when operating multiple units. No changes to resistors were needed at the time. These were the very first release RDC's. Not sure if there has actually been additional releases or KATO just selling off the boatloads of leftovers they reportedly had.

Years ago, a guy from Digitrax asked me to write this up but a couple of months later, they came out with a semi drop in replacement decoder. I never got a roundtuit. There are photos in the photo section of the Yahoo NDCC group. They are in the "KATO RDC" album.

Martin Myers
« Last Edit: April 08, 2017, 03:26:42 PM by mmyers »

OldEastRR

  • Crew
  • *
  • Posts: 3197
  • Gender: Male
  • Respect: +268
Re: Decoder goes dead for no reason
« Reply #16 on: April 09, 2017, 03:33:13 AM »
-2
When a motor contact touches a rail pickup as you describe, the train runs away at full power because it is now getting full voltage from the track. It also burns the motor output of the decoder rendering it useless. this is why isolation of the motor brushes (and the orange and gray decoder wires) is so important. Unfortunately there has never been a practical way to protect the decoder's outputs when this happens. 
Same outcome can be expected if one or more of the function wires makes contact with a rail pick up. Minimum, that function is destroyed. Maximum, whole decoder is fried.

Martin Myers

The touching motor lead to pickup makes it run away, ok , and mine does that -- except once put on the track again the decoder works fine, it's not burned out. It's an impossible flaw.  I don't think even anybody on this forum could find what's wrong and fix it.

OldEastRR

  • Crew
  • *
  • Posts: 3197
  • Gender: Male
  • Respect: +268
Re: Decoder goes dead for no reason
« Reply #17 on: April 09, 2017, 04:04:27 AM »
0
On the RDC's, I filed two clearance notches in those brass strips that run the length of the floor on either side. The notches are filed where the motor contacts come up. one on each side. The floor already has the squared notches, just make the brass strips match the floor where the motor contacts come up.
This lets them pass the brass strips without making contact. Gray and Orange wires are soldered to the contacts.
I also Kapton taped the four brass strips- two at each end - so that they continue to provide spring pressure on the truck pick ups. The tape insulates the four strips from the light boards so the functions are isolated from the rails.
White wire goes to both boards (W) mark. Yellow wire to both boards (Y) mark. Blue to both boards (B) mark. Screws are removed from each of the holes. IIRC, there are two to get removed in each board. One screw  goes back into each board where the holes are marked "DCC".
My installation veered from this by using a four function decoder. The extra two functions control the rear light boards so that I can have independant control of the lights when operating multiple units. No changes to resistors were needed at the time. These were the very first release RDC's. Not sure if there has actually been additional releases or KATO just selling off the boatloads of leftovers they reportedly had.

Years ago, a guy from Digitrax asked me to write this up but a couple of months later, they came out with a semi drop in replacement decoder. I never got a roundtuit. There are photos in the photo section of the Yahoo NDCC group. They are in the "KATO RDC" album.

Martin Myers

I bought this RDC with the TCS Z2 already installed. It seemed to run fine. But then the LED resisors started melting the plastic around them. I removed the boards entirely but that didn't end the crazy speed bursts which had also started happening. Funny, when I put the new boards on I had to reverse the LED leads to one, not make them the same like you did. That way when "forward" the wires powered the rear red light and the front white, and in reverse did the opposite. Otherwise both white lights or both red lights came on at once. But this is only a 2-function decoder. So lights work really great -- loco still goes nuts every so often.
I'd guess that the early versions of the RDC had that weak yellow LED for a headlight, because the other (non-DCC) one I got had nice bright white LEDs. I'll replace the old LEDs when -- I mean if I ever -- I put a decoder in it. Tho I guess there's a drop-in now.

davefoxx

  • Crew
  • *
  • Posts: 11675
  • Gender: Male
  • TRW Plaid Member
  • Respect: +6785
Re: Decoder goes dead for no reason
« Reply #18 on: April 09, 2017, 12:10:04 PM »
+1
Is the DCC track voltage supposed to be 22 v AC?

Maybe this has something to do with your problems.  ;)

DFF

Member: ACL/SAL Historical Society
Member: Wilmington & Western RR
A Proud HOer
BUY ALL THE TRAINS!

wcfn100

  • Crew
  • *
  • Posts: 8793
  • Respect: +1109
    • Chicago Great Western Modeler
Re: Decoder goes dead for no reason
« Reply #19 on: April 09, 2017, 12:18:07 PM »
0
Tho I guess there's a drop-in now.

Digitrax has had two different drop in decoders for the RDC.  The first was the DN122K2 in 2003 which was and discontinued in 2007 and replaced with the DN143K2.

Jason

mmyers

  • The Pitt
  • Crew
  • *
  • Posts: 997
  • Respect: +47
Re: Decoder goes dead for no reason
« Reply #20 on: April 09, 2017, 12:31:02 PM »
0
Yes, you have refreshed my memory. You do have to swap the W and Y wires on the rear light boards to get them working right.
I would reset the decoder to defaults . Write 2 in CV8. It actually appears you may have some odd entries in the speed tables or in CV's 2, 6, and 5. Resetting to defaults wil clear all of those out. Address will be 03 and you get to start fresh from there.
If the lights are too bright or hot, add resistors to the function wires. 1K is a nice place to start.

On another note: How are you measuring track voltage? I really can't figure how to get 22VAC out of a Zephyr. A power supply that would do that should make the Zephyr very hot. (just before it burns up)
With a DC volt meter, check voltage with positive on Rail A, neg on Ground. Repeat with Positive on Rail B and Neg on Ground. Add the two readings together. That is track voltage. The two readings should be the same or very close. If not, either address 00 is selected and running, or the unit is faulty.

Martin Myers
« Last Edit: April 09, 2017, 12:37:03 PM by mmyers »

Greg Elmassian

  • Crew
  • *
  • Posts: 97
  • Respect: +14
Re: Decoder goes dead for no reason
« Reply #21 on: April 09, 2017, 02:12:45 PM »
0
Umm... don't think you can read that way, unless you have a peak reading meter, as opposed to an averaging.

Better to put a full wave bridge across the rails, and measure dc from the output of the FW bridge.

a small filter cap on the output of the FW bridge is a good idea, since the square wave AC on the track is seldom perfect.

$2 in parts to get an accurate measurement.

Greg

mu26aeh

  • Crew
  • *
  • Posts: 5141
  • Gender: Male
  • Respect: +3054
Re: Decoder goes dead for no reason
« Reply #22 on: April 09, 2017, 03:58:05 PM »
+1
The touching motor lead to pickup makes it run away, ok , and mine does that -- except once put on the track again the decoder works fine, it's not burned out. It's an impossible flaw.  I don't think even anybody on this forum could find what's wrong and fix it.

Given what I've seen some people do on this forum, saying that no one could fix it is about 1 in a million.

jagged ben

  • Crew
  • *
  • Posts: 3054
  • Respect: +408
Re: Decoder goes dead for no reason
« Reply #23 on: April 09, 2017, 04:13:48 PM »
+1
Umm... don't think you can read that way, unless you have a peak reading meter, as opposed to an averaging.
...

There's nothing wrong with what Martin said, it's what Digitrax recommends.  (btw, ignore any negative signs on the results and just add them as if they were two positive numbers.  Or else keep the negative probe on ground for both measurements.)   This apparently works because 'ground' is always the negative side of the circuit compared to the rail, which for the record is not at all anything like the relationship of hot to ground in your AC house wiring. 

If you have a true RMS meter then you can get an accurate AC voltage reading rail to rail without going through those steps.  (  What this won't do is tell you if there is an imbalance due to a DC signal on the track, e.g. address 0.)   It's true that a cheaper meter that is only properly designed to read a sine wave will give an inaccurate reading. 

One thing that is inconvenient about the method that Martin described is that it is not very good for measuring the voltage accurately somewhere out on the layout where you don't have easy access to the boosters ground terminal. 
« Last Edit: April 09, 2017, 04:16:42 PM by jagged ben »

mmyers

  • The Pitt
  • Crew
  • *
  • Posts: 997
  • Respect: +47
Re: Decoder goes dead for no reason
« Reply #24 on: April 09, 2017, 08:41:03 PM »
0
I have a Fluke, a RRampmeter, and a couple of $2 Harbor Freight meters. They all read within a half a volt of each other using the AC scale. None of them is anywhere near 22VAC.

Martin Myers

peteski

  • Crew
  • *
  • Posts: 31687
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +4529
    • Coming (not so) soon...
Re: Decoder goes dead for no reason
« Reply #25 on: April 09, 2017, 09:42:24 PM »
0
There's nothing wrong with what Martin said, it's what Digitrax recommends.  (btw, ignore any negative signs on the results and just add them as if they were two positive numbers.  Or else keep the negative probe on ground for both measurements.)   This apparently works because 'ground' is always the negative side of the circuit compared to the rail, which for the record is not at all anything like the relationship of hot to ground in your AC house wiring. 

If you have a true RMS meter then you can get an accurate AC voltage reading rail to rail without going through those steps.  (  What this won't do is tell you if there is an imbalance due to a DC signal on the track, e.g. address 0.)   It's true that a cheaper meter that is only properly designed to read a sine wave will give an inaccurate reading. 

One thing that is inconvenient about the method that Martin described is that it is not very good for measuring the voltage accurately somewhere out on the layout where you don't have easy access to the boosters ground terminal.

I thought that OldEastRR was measuring the DCC voltage across the track using his meter's AC setting.  AC range on the mulltimeter has no polarity - there should never be a minus sign displayed.
. . . 42 . . .

jagged ben

  • Crew
  • *
  • Posts: 3054
  • Respect: +408
Re: Decoder goes dead for no reason
« Reply #26 on: April 09, 2017, 10:34:36 PM »
0
I thought that OldEastRR was measuring the DCC voltage across the track using his meter's AC setting.  AC range on the mulltimeter has no polarity - there should never be a minus sign displayed.

Reread the thread and note whose post I was referring to.   Measuring both rails to ground with  a DC meter is what Digitrax recommends.  They don't say anything about the polarity of the probes though, so I was pointing out that one should ignore that.  Looking into this made me realize that I don't really understand how a DCC power source is switched with reference to 'ground'.  For that matter, I'm not entirely  sure if Digitrax recommedation applies to other brand systems.

I have a Fluke, a RRampmeter, and a couple of $2 Harbor Freight meters. They all read within a half a volt of each other using the AC scale. None of them is anywhere near 22VAC.

Yeah, I think it's a bit of a myth that a typical DMM can't give an adequately accurate reading of DCC voltage on an AC setting.   Most likely it would have to both old and cheap to be that bad, and I bet most new ones will be pretty good.   I'm curious how far off your Fluke and RRampmeter are.  And also, I do think if OldEastRR is getting a consistent 22V reading he should try another meter.
« Last Edit: April 09, 2017, 10:36:39 PM by jagged ben »

mmyers

  • The Pitt
  • Crew
  • *
  • Posts: 997
  • Respect: +47
Re: Decoder goes dead for no reason
« Reply #27 on: April 09, 2017, 10:43:23 PM »
0
My Fluke and RRAmpmeter yield the same readings. They are what I used to take all readings when I were running tests for the NTRAK wiring and connector RP a dozen or so years ago.

peteski

  • Crew
  • *
  • Posts: 31687
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +4529
    • Coming (not so) soon...
Re: Decoder goes dead for no reason
« Reply #28 on: April 09, 2017, 11:06:10 PM »
0
Reread the thread and note whose post I was referring to.   Measuring both rails to ground with  a DC meter is what Digitrax recommends.  They don't say anything about the polarity of the probes though, so I was pointing out that one should ignore that.  Looking into this made me realize that I don't really understand how a DCC power source is switched with reference to 'ground'.  For that matter, I'm not entirely  sure if Digitrax recommedation applies to other brand systems.
OOPS!
With all the threads Al has going about DCC problems I got confused by this one:
Is the DCC track voltage supposed to be 22 v AC?

. . . 42 . . .

OldEastRR

  • Crew
  • *
  • Posts: 3197
  • Gender: Male
  • Respect: +268
Re: Decoder goes dead for no reason
« Reply #29 on: April 10, 2017, 04:00:24 AM »
-1
OK I'm closing out this thread, the one about the 4-6-0 got resolved so I'm going to close that, and the "programming track" one will be the only one left. Seems like there are all kinds of ways to measure the DCC current, with all kinds of equipment, and it's confusing.

PLEASE LOCK THIS THREAD