measuring inductors: expert advice wanted

[1audio]

Sencore used to build a tester for flybacks that might work. It was designed to spot shorted turns.

 

[DF96]

A Q meter is routinely used to check for shorted turns

A shorted turn may degrade Q somewhat, but it will also reduce inductance significantly. It depends on the 'quality' of the short.

 

[jneutron]

These coils are alternator rotor poles 🙂

A Q meter is routinely used to check for shorted turns

However, i am not certain i should include more detail (deliberately attempting to keep this vague for IP reasons)

Thanks for the advice, i think perhaps i should leave this idea.

No prob. You didn't mention if you are testing with metal involved.

The very high Rs you get is a combination of the current crowding within the conductor, and the eddy dissipation of the rotor metal. Unfortunately, there is no method to distinguish between the metal eddy losses and the copper ones using an inductance meter, as both losses are part of the real part of the response, whereas the energy storage (inductance) is part of the imaginary response. The meter simply looks at the in phase and out of phase response to the stimulus.

One distinction to be made is the flux direction and the direction of the rate of change. Self excitation of the coils by their terminals at a frequency will produce eddy currents can be significantly different from that which occurs when the coils are rotating within a magnetic field. The laminations break conduction in only one plane, so are designed to minimize eddies in rotation. The end result is an anisotropic flux/dissipation system. So, while Q and R are very good QA checks in a production line, actual final performance may not be accurately described by self excitation of the rotor. edit: The Q/L check will also look at the health of the rotor laminations, specifically the insulation integrity of the lam to lam. Stabolite (IIRC the trade name of the glue) tends to be uniform thickness from the vendors, as long as the rotors aren't overcompressed at curing. Trying to get above 98% packing factor without lamination to lamination conduction problems is getting into problem territory.

If you wanted to predict the copper losses a bit better, you could put a powdered ferrite rotor together, but you'd take a big hit in permeability, and that would also change the proximity effect in the conductors.

jn