Decoder failure after short circuit

[Airbuspilot]

I made my first attempt at a consist today, one loco has an LGB decoder the other a Trix decoder. I am using a ZIMO MX 32 as the CAB. The consist worked correctly and I ran for about an hour with no issues. When I returned the consist to its parking slot one loco derailed crossing a turnout and shorted the system, the MX 32 shut the power off normally.

Powering up again one of the locos (Trix decoders) ran continuously at a low speed, I was unable to control it with the MX 32 and the only way to stop was power off. Power ON and again the loco ran with uncontrolled low speed. I moved to the programming track and with the MX 32 in programming mode asked for the loco address, it stopped moving immediately but failed to respond to the address request and is now completely unresponsive.

I find it hard to believe a short would damage a decoder, I could understand a short causing the decoder to become locked in some way. The loco was completely de-powered for about an hour to ensure any stray voltages were dissipated and still no response to the loco driving on the main or CV Requests on the programming track. I presume it is only possible to reset a decoder if it is responding normally to the MX 32?

Any thoughts appreciated.

Robin

Robin

 

[PhilP]

I suspect one of the output transistors has blown. Turned from a transistor into a resistor, so is leaking current, rather than being turned off.

PhilP

 

[Airbuspilot]

I suspect one of the output transistors has blown. Turned from a transistor into a resistor, so is leaking current, rather than being turned off.

PhilP

Trains derail and short all the time without damaging the decoders I guess this time I was unlucky. I hoped not to be in dismantle mode but I think you are probably correct.
Thanks Phil

 

[Dan]

I find that LGB input bridge diodes fail and have repaired several LGB decoders. Just measure with a meter and the diode should have infinate resistance in one direction and a small resistance in the other direction.

 

[Greg Elmassian]

That is weird Dan, that means excessive current drawn through the decoder. One would think the only way to draw that much current is short a motor output. Shorting a function output should blow the ASIC / processor first one would think.

Do you have any insight on how this happens?

Greg

 

[Airbuspilot]

I find that LGB input bridge diodes fail and have repaired several LGB decoders. Just measure with a meter and the diode should have infinate resistance in one direction and a small resistance in the other direction.

Thanks Dan

I have replaced the faulty decoder with a new one, didn’t think a repair was realistic. I will test the new decoder first, just to confirm the other is in fact faulty and then have a look at the diodes in the old one. If a repair is possible I have a second faulty decoder which might be reparable as well.

Robin