I agree that the OVP relay is suspect. But I disagree on diagnostic procedures. There is NO WAY to test an OVP, at least no way using readily available tools. That includes the procedure in the shop manual, as well as any silly stuff you read here. If the fuse is blown, you can assume the relay is done. If the fuse isn't blown, you can't assume anything....the relay may still be bad. The reason is that the zener or MOV that provides overvoltage protection can be failed, even as the relay continues to appear OK. The "shake method" is a new one, I'm chuckling.

The thing to do is to substitute another OVP. If that isn't the problem, then I'd suspect the crank position sensor.