How to make an 8mH inductor which can handle 13V peak square wave without saturation

It is an artificial horizon called KI256. I need to attenuate the roll output a bit.

I have the circuits of the driver and circuits of the KI256. I measured the KI256 LVDT primary inductance, etc.

Some other data points are that another inductor but a small one, just did nothing and presumably saturated.

Crazy to have such a high drive level; I agree.

It sounds kind of like the LVDT wanted to be a resolver.

Well that's a shame. You'll have to use a method that's suited to your application.

Just an update.

I tested that 8mH coil from Mouser. Verified at 8mH with an LCR meter and with the $500 micro tweezer thingy.

Doesn't work!! It behaves like it was 20mH or something like that. The attenuation is way too high. But 8mH was the perfect value with that funny milspec 0-10mH box I got off Ebay, which was heavy enough to contain pretty big inductors.

So what is happening?

Must be something in the waveform which is buggering up the way these inductors behave, in conjunction with the demodulation scheme used to "decode" the LVDT position.

Maybe it is saturating, but saturation has the opposite effect: it

So I bought a 2.5mH one from Mouser, again one which can do 1.5A or so, and will try that, after the potting compound has gone off :)

An inductor in series with the excitation coils of an LVDT will shift the phase of the current going through the LVDT. If the output of the LVDT is being demodulated with a phase sensitive detector you'd need a matching shift in the phase of the drive to the demodulator to get a sensible (or useful) output.

A precision rectifier wouldn't have that problem, but they weren't all that precise decades ago, and not all that popular. They also don't reject noise and out-of-phase pick-up, so nobody sensible would have used one.

Worse - it could be saturating on part of the cycle and giving you a very funny current waveform going through the LVDT - actually a rotary variable transformer, but we've been through that,

Looking at the current waveforms is always a good idea, but not always all that easy.

More looking a what's actually going on might be a good idea - buying random parts before you had worked out exactly why the last one didn't work isn't good policy.