[TR] I give up

[Randall]

> I would think that sort of meter less 
> consistent across manufacturing samples, but maybe that's the 
> whole point, it doesn't have to be all that accurate.

Exactly.  Hard to notice even a 10% error (3 amps out of 30).

But now that I think about it, they could be calibrated by bending the wire
loop a bit closer or farther away from the vane movement.  Seems unlikely
they would bother, but maybe ...

Randall