Page 1 of 1

timeout on PT acknowledge

Posted: 21.10.2010, 19:40
by joernr
Hi rocrail-team,
I have no central-station. I' have only a Ubuntu PC with DDX and a booster from railway-lauf (http://www.railway-lauf.de/BST-375KF.htm). The acknowledge detection I have copied from the ord-1. I can write to the Decoder (Lenz Gold/Silver mini) but I couldn't read from it. I don't know if have a hardware or a software problem. On the oscilloscipe I can see ack pulses on RI. Is it possible to trace the RS232 Ports RTS and RI?
I have switched on debugging. You can see logs as attachment.

After 30 seconds I got this error: timeout on PT acknowledge
The value of cv is -1.

Last messages in the logs:
0101021.191913.911 r9999c cmdr08B5 impl/ddx 1398 PT: ack = 0
20101021.191913.911 r9999D cmdr08B5 OApp 0081 broadcasting a message: PT: ack = 0
20101021.191913.911 r9999D cmdr08B5 OClntCon 0460 Broadcast event...exception
20101021.191913.911 r9999D cmdr08B5 OThread 0215 queue[broadcast][ClientCon Broadcaster] items 0
20101021.191913.911 r9999c cmdr08B5 impl/ddx 1400 PT: disable booster output
20101021.191913.912 r9999D cmdr08B5 OApp 0081 broadcasting a message: PT: disable booster output
20101021.191913.912 r9999D cmdr08B5 OClntCon 0460 Broadcast event...exception
20101021.191913.912 r9999D cmdr08B5 OThread 0215 queue[broadcast][ClientCon Broadcaster] items 1
[ ][ ][RTS][ ][ ][DTR][RI ][ ][4086]
20101021.191913.912 r9999I cmdr08B5 OControl 0182 Program event...value=-1
20101021.191913.912 r9999D cmdr08B5 OClntCon 0460 Broadcast event...program
20101021.191913.912 r9999D cmdr08B5 OThread 0215 queue[broadcast][ClientCon Broadcaster] items 2

What is the best way to troubleshoot this problem?
Rocrail 2.0 rev 2084

Best regards

Joern

Posted: 23.10.2010, 22:11
by RainerK
Hi Joern,

The ORD-1 has symmetric power supply and the output driver is only a half bridge.
Therefor is one rail stable on ground potential and the other rail changes the potential between plus und minus voltage.
This kind of operating is urgent required to use the unmodified acknowledge circuit like ORD-1.

Please consider that Your booster has a symmetrical output to the rails with a full H-bridge and also if You use it for programming.
Therefor is it not correct to use an unmodified acknowledge circuit from ORD-1.
Connecting the 39Ω resistor in this case at one rail You get a mismatch that deform the acknowledge pulses.

In my opinion is it necesary in case of programming (only temporary in this case :!:)
the current measurement resistor (0,05Ω) in the "sense" circuit in the H-bridge of Your booster to separate
and inserting the acknowledge circuit (such as that from ORD-1) at this place.
In many other H-bridge boosters with programming function is that a usual proceed.

Best Regards
RainerK

PS: Please Admins, it will be better to move this thread to "All other Hardware"

Posted: 30.10.2010, 09:28
by joernr
Hi Rainer,

I know that I have a symmetric output on my booster. So I have modified my layout a little bit. You can see my layout as a new attachment. My problem is, that I don't know, how can I check, that I have a correct acknowledge. Can you explain me, what happens if I'm reading CV's from the decoder? When send's the decoder ACK? Can I check this in rocrail logs or with unix commands? On my Oscilloscope I see one peak at RI when reading starts and a second after ca. 10-20s. But this is not the of the read process.

Best regards

Joern

Posted: 30.10.2010, 14:17
by RainerK
Hi Joern,

where did You take the +5V supply for the acknowledge circuit ?
I think it must be galvanic separated from the booster circuits and also the ground from other ground must isolated.

Best Regards
RainerK

Posted: 31.10.2010, 11:11
by joernr
Hi Rainer,

the 5V from the 7805 of the booster and the Ground is form the booster to. I will try to use a separate Powersupply.