Measuring voice-coil self-inductance

[Klimon]

Hi,

Does one need special measuring equipment to get ahold of a woofer's voice-coil inductance (for adding a zobel)? I've tried on several different woofers with my cheap multimeter and the value is always false (uncredibly high e.g. 10mH). The multimeter only goes berzerk on voice-coils, caps and crossover coil readings are always correct.....

Cheers for any inspiration,

Simon

[Geoff H]

My guess is your digi MM is being tricked by the complex Z of the driver.

If you can determine the Z of the driver well above resonance, you can then calculate the inductive reactance at a higher freq. Say 3 Khz
The capacitive component will be negligible.
Z squared = R sq'd + X sq'd. or
Xsq'd = Zsqd - Rsq'd.

From there calculate L

Xl=3.1416 F L

You probably know the formula, after all you are talking Zobel networks, that's for benefit of others.

Regards

[Svante]

The problem is even worse. Not only do the mechanical properties of the driver make measurment hard, but so does also the voice coil inductance itself.

The inductance varies with frequency. I have written a bit about this in both the tech doc and user's guide of Basta! . The best way to my knowledgeto model the inductive behaviour is to throw out the idea of an ideal coil, and use the behaviour described in the link above.

The voice coil inductance numbers in datasheets are typically useless.

[Klimon]

Guess I'll have to trust the manufacturer's specs for now. Thanks

Simon

[Ron E]

Your meter goes haywire because you need to restrict cone movement to measure the inductance. The coil should be "locked".

Inductance in a voice coil varies with frequency(it is not a pure inductance), your meter probably measures at 400Hz, so if you are using the measured value at 400Hz for some other frequency, you will not get the expected results.

If you are designing a Zobel, a single frequency measurement is not likely to help much, you need to measure the entire impedance curve, then model in a program to get the flattest impedance. You will find that once you can model you probably won't use "ideal" Zobel values anyway.

Believe me, designing zobels and crossovers with cookbook formulas is not as deterministic as it sounds, there are many more variables you need to consider to do it right.