The total for the board is less than 3 amps, and the PA is not switched.? There is not more than 500 ma being switched by the main T/R relay K1. K3 is handling the antenna switching between the PA and the receive front end.? That could have high RF induced currents that would be difficult to measure.? The other band switching relays are also handling the RF power.
With the above, we need to know which relay is sticking.
There is one other possibility, that is RF feedback into the rig causing the Nano to sense that a key is pressed. I had this issue with an off center fed dipole feeding back into the rig and causing it to lock up.? Have there been any changes to the shack right before the relay started sticking???
Above are just some ideas.? Please take with a grain of salt and research before using.
73
Evan
AC9TU
