[[email protected]]

Ten days ago the problem appeared again just long enough to check 177A-B-C terminal was 0 volts and then it started.  Yesterday it wouldn't start so I checked TP N had 0 volts and TP M had 12VDC which meant the K31 relay wasn't passing power.  A very light tap on K31 relay and the engine started.  I had replaced K31 a while back but replaced it again.  
Hopefully it isn't a cold solder joint on the relay socket or a circuit board problem, we'll see if it fails to start again.
That will lead to testing K31 relay coil power at TP a and coil ground at TP L and a wiggle or very light tap on K31.

Nice detective work Jim! 
I was going to suggest checking the voltage to the K31 coil as a continual problem with the same relay is a sign of a potential other issue. I really do not like board mounted items like relays etc, small electronic components are fine but when you start getting into relays and larger loads I think they should be stand alone. Very similar issue to our SCS air conditioner control boards.

Are those blade style connections for the relays? If so you may make sure that the female part of the connector is a tight fit, sometimes they are known to become loose and not make good connection with the blade on the relay.