Author Topic: That KATO RDC problem....? Well, funny thing ....  (Read 3384 times)

0 Members and 1 Guest are viewing this topic.

Greg Elmassian

  • Crew
  • *
  • Posts: 97
  • Respect: +14
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #15 on: April 09, 2017, 02:24:57 PM »
+3
I think that many of your issues are no problem to solve, once the FULL story is revealed.

The very first thing I heard about decoders is to make absolutely sure that the motor wiring can never come in contact with the track pickups, or you destroy a decoder.

Seems that's part of your problem in the beginning of the thread.

The next thing I was told about DCC 10 years ago, if a loco takes off unexpectedly, turn off Analog (DC) mode.

I've seen these 2 things repeated over and over in many places, but you seem to be not only surprised, but irritated about just learning now.

So, I wonder if you are not listening to advice, or not reading the volumes of information available? Blaming your misfortunes on everyone else won't reduce the problems you are encountering.

This is NOT rocket science, but it does involve spending some time to "get" some of the basic concepts.

Not trying to give you a hard time, but being a newcomer to the forum give me perhaps a "unbiased look" at this situation. Also I'm a little surprised at some of the, well, insults to the people attempting to help you.

Greg

OldEastRR

  • Crew
  • *
  • Posts: 3412
  • Gender: Male
  • Respect: +311
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #16 on: April 10, 2017, 04:10:05 AM »
0
I'm also calling the correctly working RDC LEDs as a win here, and will have to live with the random wild fullspeed runs of the thing, I guess. The original problem was the resistors getting so hot they melted the plastic, and that seems to be fixed at least. I can run the chassis now with its shell and other parts on. It was not very prototypical running around with just the frame and huge LEDs on the ends. So I'm closing this thread.
For the record, I did have a friend turn off the DC running CV setting, which I reported earlier. No fix.

Greg Elmassian

  • Crew
  • *
  • Posts: 97
  • Respect: +14
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #17 on: April 10, 2017, 03:05:51 PM »
0
Good deal, a win is a good thing!

Perhaps, the seemingly high voltage on your setup could cause this.

I'm still not believing the 22 volts, but if it is true, that can scramble many a decoder.

Anyway, congrats!

Greg

djconway

  • Crew
  • *
  • Posts: 482
  • Respect: +74
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #18 on: April 10, 2017, 04:17:09 PM »
0
To turn Analog (DC) operation off
1  read the value in cv29
2  subtract 4 from the cv29 value
3  write the new value to cv29


peteski

  • Crew
  • *
  • Posts: 32957
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5341
    • Coming (not so) soon...
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #19 on: April 10, 2017, 05:36:01 PM »
0
To turn Analog (DC) operation off
1  read the value in cv29
2  subtract 4 from the cv29 value
3  write the new value to cv29

Not so fast.  That works only when the DC bit is already turned on.  If not, then this will not have the desired result.   For example if CV29 is set to 34 (decimal) which is long address and 28/128 steps then subtracting 4 will mess things up.

34-4=30. That will configure the loco  to use: 28/128 steps, enable DC operation, turn on RailCom, and enable custom speed curve.  That is a totally undesired result!

The safest way to do this is to read the value from CV29, determine what functions are enabled, then decide whether the 4 needs to be subtracted or not.
. . . 42 . . .

djconway

  • Crew
  • *
  • Posts: 482
  • Respect: +74
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #20 on: April 10, 2017, 06:50:06 PM »
0
Not so fast.  That works only when the DC bit is already turned on.  If not, then this will not have the desired result.   For example if CV29 is set to 34 (decimal) which is long address and 28/128 steps then subtracting 4 will mess things up.

34-4=30. That will configure the loco  to use: 28/128 steps, enable DC operation, turn on RailCom, and enable custom speed curve.  That is a totally undesired result!

The safest way to do this is to read the value from CV29, determine what functions are enabled, then decide whether the 4 needs to be subtracted or not.

In 21 years of operating DCC I have only seen locos that have the 4 bit set have the runaway issue.  Has it been known in other cases?

peteski

  • Crew
  • *
  • Posts: 32957
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5341
    • Coming (not so) soon...
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #21 on: April 10, 2017, 07:37:51 PM »
0
In 21 years of operating DCC I have only seen locos that have the 4 bit set have the runaway issue.  Has it been known in other cases?

That is irrelevant (and yes, a decoder could go wild even with DC mode disabled). I was just saying that make sure that bit 2 (with value of 4) is set before subtracting 4 from the value of CV29.  Because if it's not set, you will not get the result you expect.
« Last Edit: April 10, 2017, 09:35:43 PM by peteski »
. . . 42 . . .

Greg Elmassian

  • Crew
  • *
  • Posts: 97
  • Respect: +14
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #22 on: April 10, 2017, 08:59:51 PM »
0
Yes I have seen decoders go nuts with analog mode off... I often get them in Z scale with CT Elektronik and TCS decoders, especially on motors that have a lot of current draw, in Z most often the older 3 pole motors.

In G scale I see this when there is a very noisy signal, poor wiring, and older type decoders like the NCE ones.

Greg

OldEastRR

  • Crew
  • *
  • Posts: 3412
  • Gender: Male
  • Respect: +311
Re: That KATO RDC problem....? Well, funny thing ....
« Reply #23 on: April 11, 2017, 01:56:15 AM »
-2
Not so fast.  That works only when the DC bit is already turned on.  If not, then this will not have the desired result.   For example if CV29 is set to 34 (decimal) which is long address and 28/128 steps then subtracting 4 will mess things up.

34-4=30. That will configure the loco  to use: 28/128 steps, enable DC operation, turn on RailCom, and enable custom speed curve.  That is a totally undesired result!

The safest way to do this is to read the value from CV29, determine what functions are enabled, then decide whether the 4 needs to be subtracted or not.

See, this is the confusion. "Determine the magic number, then subtract 4 and you're fine!" So every decoder from different companies has a different value for CV29 to turn off DC. It's not just something like "Enter THIS FIXED NUMBER to shut off the DC". I must calculate something from somewhere with some kind of variations and then subtract a number (at least they tell me the number) to get the result I need. Why did they make it so complicated? A set number for EVERY brand of decoder to reset factory values (that I know), a set number for EVERY brand of decoder to shut off DC, and so on. That's not possible? Or the decoder makers don't want to use anybody else's values?

 As far as DCC goes, I have no problems with the drop-ins or factory-installed decoders (except MRC, which I will never get again). I do have locos with sound and I use them, they work fine. DCC is great for all those engines and I get no frustrations. But I'm going to have to avoid doing or buying from someone any hard-wired decoder locos, as they are too much trouble. Which will be disappointing because nobody makes decoder-ready DL109s, or FA-2s, or 5-axle C-liners, all of which were signature NH power of the early '50s (and I have models of them). FL-9s are after my cut-off date so my NH layout passenger trains of Osgood Bradley and (hopefully soon) single window coaches will have only PAs or RS-3s to pull them, otherwise it's E8s or a Ten Wheeler. There seems to be a drop-in for RDCs so at least I can salvage them. For freight locos I'll have to go with more E8s and a U-30 or Geep. Sigh. That may explain why I'm really out of sorts about this.

So again, I'm calling just working LEDs for the RDC as a win here, so we can close this discussion. My other DCC problems will be in the other thread.

PLEASE LOCK THIS THREAD

« Last Edit: April 11, 2017, 02:45:18 AM by OldEastRR »