[Dave Moll]

I'm glad it wasn't just me failing to understand the diagram. I agree that Dag's modified version would make more sense. Perhaps what is needed is some resistance measurements between the contacts on the magneto - both at rest and off-normal (i.e. with the magneto handle away from its rest position but not turning!) The latter may or may not work depending on how the magneto switch operates.

Incidentally, the switch of the GPO magneto 26N/26P works to make the magneto electrically invisible when at rest, but with the handle off-normal (I don't think it needs to be turning) disconnects the local end and connects itself across the line to send its ringing current.