Strange Loco Address 63 behaviour

[gilbo2]

Programmed my latest loco to address 63 tonight. Was really surprised though that when controlling the loco, or switching its lights on/off, that I could hear some of my points trying to change!? They are all in point number range 1 to 24.

I therefore tried going up another loco address to 64 and everything back to normal. Seems really odd that loco address 63 has this effect on the points.

Any ideas other than just avoiding address 63?

(Using eLink and RailMaster)

[St1ngr4y]

Interesting that 63 in binary is 111111. I wonder if any other addresses which happen to be (a power of 2) - 1 give equally odd behaviour - addresses such as 31 or 127 for example.

Ray

[gilbo2]

Thanks for your reply Ray.

Have just carried out some further tests this morning on the following addresses;

31 - No problems

63 - Switching lights on/off causes points to change

64 - No problems

127 - Switching lights on/off causes points to change

I also noticed that when points changed above that I could hear a very fast "ticking" which comes from I believe the loco motors on the track (all of them!).

All really odd, but I have two "blacklisted" loco addresses now never to use - 63 & 127.

Would be interesting if others can replicate these findings too.

[St1ngr4y]

I'll give it a try later today.

Ray

[St1ngr4y]

I tried a few different addresses using locos which don't actually exist. Like you I found perculiarities with addresses 63 and 127. By the way, I have Elite + RM. I set up functions F0 Lights on/off and F1 Sound on/off, and I found that clicking F1 for either loco 63 or 127, caused one point to change. This point has an address of 33 and is one of the four ports on my one remaining R8247 decoder. What I found in addition, however, was that the point would only move once. Clicking on the RM screen on ANY of the four ports on this accessory decoder (33, 34, 35 or 36) on the RM track diagram would not change the points. All other accessory decoders and all locos still worked ok. I found that I could restore functionality to the accessory decoder by resetting the Elite. It could be that all that was required was for the power from the DCC bus to be switched off then on again, but for me, resetting the Elite was the easiest.

When the accessory decoder stopped working, it was as if the CDU had not recharged, because you could still hear it trying to throw the points, but not having enough power.

Since you discovered the problem originally by using address 63 for your loco, do you want to report it through the system? Or should we wait to see if Admin view this thread and comment?

By the way, I didn't notice any noises from the locos on the track when I was experimenting. Was it F0 or F1 which you were using? What kind of accessory decoders are you using for your points?

Ray

[gilbo2]

Thanks for doing this test too Ray. I was using the standard F0, lights on/off.

I suggest we both report this through Railmaster so they can see its not just an isolated case.

As for the clicking noise, one of my locos was making more noise than the others, so I guess its down to the motor characteristics as to whether it can be heard or not.

[Fishmanoz]

How weird. 

HRMS, any comment from you on this?

[gilbo2]

It seems to be address 127 that triggers the "clicking" I've mentioned above when operating f0 lights.

[gilbo2]

Sorry, one more post to say have sent an email to HRMS. The help request through RailMaster would not send/connect, even with no firewall enabled at all.

[St1ngr4y]

Another interesting co-incidence - about a week ago I tried to report another fault through the RM system, and it got stuck during the sending of the message. In the end I gave up and posted a message on the forum instead.

I wonder if they have deliberately switched off the fault reporting system, so that they can concentrate on the preparation of v1.56?

Ray

[Fishmanoz]

I was just able to send an email through RM Help with no problem although yet to receive the auto confirmation.

I asked about this issue and the Decoder Compatibility thread and inability to progam some decoders through RM/elink 

[Fishmanoz]

Have had a reply from HRMS saying:

We are already looking into the issue regarding the throwing of points when turning lights on/off for loco addresses 63, 127 and others.

 

 

[gilbo2]

I had the same reply today too.  Be interesting to see what the outcome is.

[gilbo2]

I'm not sure if there is a 1.56 connection to the fault report system issue, perhaps we're hoping :-)

As a matter of interest, did you see the same problems with address 63 or 127 if using your Elite direct, ie without RailMaster?

[St1ngr4y]

I thought maybe they didn't want any new fault reports until after 1.56 has been released.

I didn't think of using Elite directly. Presumably all I need to do is key in 63 or 127 as the loco (even though it doesn't exist), then try functions 0 and 1 to see if points are thrown.

I'll try that tomorrow and report back.

Ray

[gilbo2]

Yes, thats all you would need to do on the Elite to try these addresses. Thanks.

[St1ngr4y]

I thought it was later than it was, so I tried it out just now. It seems to be a feature of the Elite. When I used loco address 63, I got similar results to those when using RM i.e. point 33 would try to throw in response to F0 for loco 63. When I tried F0 with loco address 127, point 33 went berserk - it seemed to be trying to throw rapidly, 2 or 3 times a second, but again not having enough kick to throw. I left it for a few seconds and it calmed down and stopped, as if the CDU charge had completely gone. I pulled the Elite power input plug at that point in case I did damage to either the point motor or the R8247. Thankfully, when I powered it up again, point 33 operated normally using the Elite.

At least the problem doesn't appear to be with RM itself, more of a controller firmware problem I would have thought. I did these experiments without the pc being switched on.

Ray

[gilbo2]

Really useful information Ray.

So in summary the following set ups show this problem;

Elite on its own

Elite with RM

eLink with RM

 

So it is looking like either the firmware in the elite and elink or perhaps a bug in the NMRA standard itself? At least now this issue is known it's just a case of staying away from 63 and 127. I wonder though if there are others to avoid?

[St1ngr4y]

Another element is the accessory decoders. I don't think you mentioned which accessory decoders you use, and which points/port address(es) are affected. I have one R8247 and five ADS8s and it only seems to be the one port on the one R8247 which reacts. I have a spare R8247 which I can attach to my programming track and re-configure it to any address. I also have a spare p*co point motor rigged up to plug into any one of the four ports on the R8247. I'll see if this decoder reacts to loco 63 or 127, and I'll try it on a variety of accessory decoder addresses.

Ray

[St1ngr4y]

I plugged in my spare R8247 decoder, which already had port addresses of 97, 98, 99, 100. I plugged my point motor into port 97. On the Elite I selected loco 127. When I used F0 and F1 I got the constant pulses on the point motor on port 97 and the same with the point on port 33 of the other R8247. None of the points attached to the 5 ADS8s reacted at all. I put a multimeter across the output from port 97 to the point and it registered 2 or 3 volts, not constant but varying up/down by about 0.2v. I tried the same on ports 98, 99, and 100 but they did not produce any output.

So, dg, are your decoders of the R8247 variety? The problem could lie only in the R8247, reacting eratically to loco addresses 63 and 127.

Ray

[gilbo2]

Ray, you have been busy :-)

I have a mixture of R8247 and Tawcraft PAD's. Only the R8247 operate BUT the Tawcraft have latching outputs so not possible to detect if they are being pulsed as we see on the R8247's.

Personally, as 127 was causing my locos to "click" I suspect it is not just R8247 related, but could be wrong. To me its as if 63 & 127 are like a skeleton keys, able to operate anything on the bus in various forms.

Will be interesting to see what HRMS come back with. I think this could be a tough one?!