Awesome!
I'm going to dig through my notes (and Morgan's explanations) to hopefully bring my understanding up to the point I can ask you some reasonable questions...
By the way, I've made a bit of progress on actually understanding my resolver circuits. Funny thing is when I made my first attempt at them, they worked Yea! Except... It was one of those situations where for some reason the kludgy thing works, but you really don't UNDERSTAND why...
The main issue is the resolver driver, which I thought needed to be pretty powerful. After all, I blew up an audio amplifier capable of 3 amps... It seems if you really start to understand impedance, common mode DC, etc, you can get away with something MUCH smaller (cheaper) and reliable.
I got my hands on the data sheet for the resolver and the resolver to digital decoder used by Toyota. There is a TON of good information in those documents. The main limitation that I've found is the carrier signal is targeted at 10kHz. What happens if we drive it 5 times faster??
To find out, I ran a series of tests from 10kHz to 85kHz and measured the input voltage, output voltage and calculated the input and output impedance and transformer ratio for every data point.
The good news is that the input and output impedance actually increase proportionally to the carrier frequency. At the same time the transformer ratio stays relatively constant. Why is this good? For a given voltage, the higher frequency will require less power.
I'll post some links and pretty graphs soon.
- E*clipse