Hi Andy,

Nice of you to sacrifice your spares for science. I'm not sure what useful test that would be, to crank with the coil wire off. I would expect the tach to quiver just the same if it did with it hooked up, and neither way be affected by the fuel relay.

Now if it didn't move while cranking, it might move if the plugs are removed due to the increased cranking speed and voltage available. But still, the dead needle is not a reliable indication of anything until the car is at least idling. The tach was not built to measure cranking speed.

Here are some of my musings on the subject I mostly credit or blame a friend for, who designed high voltage ballasts for hot-restrike high intensity discharge lighting. Similar application and outdoor environment:

Insulation breakdown in a tranformer is rarely an instant death. First the arc needs to form, then the arcing chars the insulation into conductive carbon paths that grow. Lots of factors would influence the insulation resistance, like heat cycle mechanical abrasion internally and moisture ingress, making a new unit that passed hi-pot testing more robust than a 15-year old original.

I suspect the form the Rex coil takes contributes too, if voltages are equal, simply from my own experience in a small number of cars where the confirmed coil deaths for the E-I lamination (square style) I recall seem to outnumber those of the traditional cylinder form. One sticks out in my memory about 10 years ago, after helping a boss with his 80's Ford pickup where the coil looked OK on the ohmmeter, and venturing the price of a replacement from a nearby parts store was a successful gamble. I had given that square coil the evil eye even then.

This is just plain stretching it, but I think I even read somewhere another reason not to run it unloaded. Probably a repair manual warning, but I thought it was to avoid damage to the driving electronics. Whether this was supposed to arise from stray leakage currents from the tower, or reflected load into the driver transistor, was not explained. Again, this wouldn't be the kind of damage that would occur repeatably in any reliable fashion, but could be cumulative, because it essentially would be insulation breakdown in another component affected. I could well imagine on a Bosch car, damage to the AMM or ECU might occur if the coil wire was attached but unloaded at the distributor end and draped over the harness, but that is just asking for it.

So with all this babbling, if you've stuck with me this far, I'd just recommend sticking a grounded spark plug at the end of the coil wire, if for any reason you want to crank with the coil connected but without spark going to the engine.
--
Art Benstein near Baltimore