Crosstalk? Some Cobalt motors operating without being switched!

[Edward]

A very recent problem (yesterday in fact) at a weekend show in our MPD layout.   Some fiddle yard Cobalts kept operating points without being ordered.  Details of our system:  NCE powerCab plus 2nd cab as PowerPro, using NCE Mini Panel to call macros setting routes with 1,2 or 4 Cobalts being activated.  Inputs in the mini panel triggered by SPST/DPDT switches: a total of 8 three way points in use.  The 5 fiddle yard points are just single ones and are controlled directly by SPST switches.  (We have MERG circuit breaker to give layout protection)
The Mini Panel inputs worked faultlessly but when 5 of the 9 inputs were grounded at least 3 of the quite separate FY points changed. It seemed to happen occasionally at first but careful checking revealed it was happening each time a particular input was grounded.  The point seems to change in the one direction only.   I'm wondering if there is a "memory" left in the cobalt's setting circuit which is responding to the macro?  Is there a possibility that interference is there? It seems unlikely as it's always the same motor operating?
As you can imagine, it took sometime before we cottoned on to what was actually happened! I've had a quick look at the panel wiring  but it all seems to be in order.
I would be very pleased any help and/or for some advice on solving this very annoying problem or at least what/how to check. 

[Nigelcliffe]

First question would be what is being sent to the track from the MiniPanel.   ie. are there commands you are not expecting appearing from the MiniPanel ?  

 

I don't do enough NCE stuff to know what diagnostics are possible with tools such as JMRI connected to the PowerPro, but that's one place to start.   Or a DCC Packet Sniffer to see what is actually being sent to the track.  

Or, if lacking either of those, program another stationary decoder (eg. a new Cobalt-Digital) with the address of the ones which seem to randomly change and see if that also changes.  

 

 

- Nigel

[Edward]

Thanks Nigel - I've now been in touch with the helpful people at Dcc Concepts/.  The solution is to fit diodes in line in the circuit of each Cobalt motor .  Apparently is quite common when there are long lengths of wire to the switch. 

[DCCconcepts]

3 hours ago, Edward said:

Thanks Nigel - I've now been in touch with the helpful people at Dcc Concepts/.  The solution is to fit diodes in line in the circuit of each Cobalt motor .  Apparently is quite common when there are long lengths of wire to the switch. 

 

Hi Edward,

 For the benefit of others that may read this thread.

 

It doesn't mention in your original description that you also have manual control via the PBS terminals.

We sometimes see interference via the PBS terminals, which can cause ghost operation.

The solution is quite simple.

Fit diodes in two of the PBS leads - they can be 1N4001 or similar.

The diagram shows an S-Lever, but the same applies to two pushbuttons or a spring-to-centre-off toggle switch

 

https://www.dccconcepts.com/manual/cobalt-ip-digital-long-wire-and-diodes/

 

If using a single pushbutton across 7 and 9 to toggle the motor, use a single diode in series with the pushbutton with the band end pointing towards terminal 7.

 

Twisting the wiring between buttons and PBS terminals will also help.

 

Best Regards.

The DCCconcepts Team.