The good GM engineering has tripped up many troubleshooters. They don't realize how MUCH current the pull-in windings draw and it's only for about 25 miliseconds. So if the battery terminals are dirty as in spring and fall, the solenoid won't pull in before the battery terminals "pop." The starter voltage essentially goes to 0V and then immediately returns to normal. The lights work before and after you hit the starter, so people would blame the solenoid and not bother to check the battery terminals. Lack Of Maintenance rules again.
When I worked at Fluke in the eighties, I wrote an application note of various electrical tips and techniques: one used the 20 millisecond, peak-hold function on the multimeter along with a dc amp-clamp that would capture the current peak on the starter pull-in. It turned out that pull-in peak current was higher than the total starter draw during cranking. Something like 90A IIRC. Fords were considerably higher as I recall and had problems because of undersized wiring even if the battery terminals were clean.
We gave out those application notes by the thousands, often a case at a time, to the point where paper cost became a budget item my boss was mad about. He didn't care about how may multimeters we sold (lots). Snap-On had a standing order on every print run. I think the only one I have left is a color press-proof (the printers had trouble getting the yellow right).
A good multimeter and a knowledge of ohm's law and you can diagnose a lot of things with one hand in your pocket. Jerry might have run across that Fluke App Note in his shop classes. They would have showed up about '86 or so. I still feel like I did a good turn for all those techs who were otherwise flying blind with meters that didn't even come close the the truth. Along about 1990 we developed an automotive-specific meter, the model 88. I still have one. My dc amp-clamp died though and I miss it.
