Impedance Measurement and an Experimental LCR Meter

RE: Impedance Measurement and an Experimental LCR Meter

shabaz — Mon, 27 Jun 2022 17:31:26 GMT

test comment

RE: Impedance Measurement and an Experimental LCR Meter

shabaz — Mon, 09 May 2022 02:18:28 GMT

I tried getting higher-res measurements, I think this method works!

Basically, the Pi instructs the DSP to hold a measurement, and then the integer portion is subtracted. That leaves the fraction portion, which gets boosted by X10 or X100 and the best value is read. I don't think the DSP likes me making so many changes on the fly, otherwise I would have tried something more complicated. As it is, the measurement fails on average 1-in-10 times, and I don't know why. It works 100% of the time after the DSP is loaded. I could load the DSP for each measurement, but then I can't easily speed up measurements : ( Anyway, that's a different problem to resolve! (I believe it will be resolvable, I need to put some thought into it still).

With this high-res scheme, the measurements are looking very sensible for capacitance, whereas inductance still looks wrong at the 100uH level, I've not tried higher inductance.

I measured the 2.2uF and 10uF capacitors (with nothing in parallel) with an EUCOL U2832 meter, and it reported:

2.2uf Capacitor:

Z: 76.6 ohm

Cp: 2.067 uF

Rp: 880 ohm

10uF Capacitor:

Z: 17.08 ohm

Cp: 9.28 uF

Rp: 175 ohm

And this is what this project reports for the two capacitors, but with 10 ohm in parallel, so that I can get it into a range I can measure:

2.2uF Capacitor // 10 ohm:

10 uF Capacitor // 10 ohm:

I think this is as good as I can hope with the current hardware, the results are pretty accurate!, and the measurements are stable (except when it fails 1-in-10 times or so).

RE: Impedance Measurement and an Experimental LCR Meter

shabaz — Sun, 08 May 2022 19:06:25 GMT

To test if the values are realistic or not, I had a shot at creating the simulation of the potential divider with LTSpice.

This should hopefully solve any glaring errors, but it doesn't simulate everything, i.e. phase offsets and scaling etc. Based on this I think the parallel calculation is ok, so perhaps the wild discrepancy with inductance is because the measurement is too small with the component values, and that is introducing the large error. Or the way I'm getting rid of the of the phase offset is wrong.

RE: Impedance Measurement and an Experimental LCR Meter

michaelkellett — Sun, 08 May 2022 15:08:25 GMT

Interesting stuff.

A long time ago I did a lot of work with DSPs (programmed in assembler at first and then C).

The AD block diagram tool looks OK but it would be nice to able to get right into the chip when you want to.

The little board is enticlngly cheap..

In the last few years all the projects that have needed DSP have been OK on an ARM or else needed an FPGA.

RE: Impedance Measurement and an Experimental LCR Meter

scottiebabe — Sun, 08 May 2022 12:47:51 GMT

Fun little project. Just a few observations:

Z = R + jX => so |Z| > { R, X } in some of your examples |Z| < R (or X)

The impedance you calculate is the series combination of a resistor and a reactive component. If you want to measure components in parallel you will need to apply a series to parallel impedance conversion.

I wish the laser marking of chips always looked this good:

It almost looks too good to be true! Nice photography skills [emoticon:c4563cd7d5574777a71c318021cbbcc8]