20nH-200nH range inductance tester

Three methods for measuring coils and capacitors come to my mind:

• measuring the phase shift at a known frequency
• measuring the resonance frequency with a known component (fixed capacitor in your case)
• measuring the impedance at known frequency (or several frequencies)

I would tend to the last method (personal preference), but used the first one to measure ferrite core material (in the upper uH range) some years ago.

I'm no hardware guy, google (and the TI website) have probably better suggestions for the realization ...

Thank you, f.m.,

I also saw in the Internet PWM-based solution.

I am also not an analog-inductor-guy, because I don't have specific tests equipment for 100MHz-1000MHz range, which could cost me 1000$-10000$ for a unit.

I don't know the measuring range of this LCR meter, but I had this link saved in my bookmarks:

Maybe it is interesting for you.

Thank you, Dennis, but this solution has 10kHz ceiling point.

I am also not an analog-inductor-guy, because I don't have specific tests equipment for 100MHz-1000MHz range, ...

With proper selection of the "other" element(s) of the resonant circuit (the capacitor in your case), I would try to keep the frequencies in the 10's to 100's kilohertz range. Something the ADC could measure with more than two samples per period.

To bring up another idea - I'm quite sure there are DVMs around with C/L measuring capability, and interface option. Buying might be cheaper than making here.

f. m. said:

I'm quite sure there are DVMs around with C/L measuring capability

You are right, but DVMs has 1mH lowest ceiling point it is 100000 times far away from what I need.

You are right, but DVMs has 1mH lowest ceiling point it is 100000 times far away from what I need.

I remember DVMs going down at least to the uH range. One might expect some correlation between range and price ...

A commercial LCR meter, even a used one, possibly ?

f. m. said:

One might expect some correlation between range and price ...

No, the range is quite tight: 20nH-200nH. So this is very narrow, but working frequencies for measurements are quite high.

OK, thank you for your effort. I found some very expensive and very cheap solutions on the Internet. Since I need only very cheap (I don't expect good revenue from this action) modulation-based analog conversion solution , I probably will stop on one of that.

f. m. said:

•have a resonance circuit with a known capacitor
•use a schmitt trigger to form a rectangular signal
•use a counter/divider (hardware) to scale down
•measure period (e.g. with a MSP43x), and map to inductance via frequency

Thank you, f.m. it is a very good idea.

Sounds perhaps simpler as it is.
Designing an amplifier in the MHz range with a stable output amplitude is not trivial, neither is factoring in the parasitic capacitances/inductances and temperature drift. Don't ask me about that ...

f. m. said:

parasitic capacitances/inductances and temperature drift. Don't ask me about that ...

Don't worry, based on NASA promotion and prospective, even children from primary school can do that. ;)

I already had thoughts about problems in implementation of a schmitt-trigger in 0.1-1.0GHz range.

Thank you, Ilmars, for such a precise answer.

Ilmars said:

I would pick "resonance circuit with a known capacitor" too.

I know. This is a challenge to test capacitance lower than 10pF (lower than 2pF is needed for this situation) because self capacitance by touching the isolated wires is around the same.

>lower than 2pF is needed for this situation
In your case (LC tank oscillator) 100pF would result 113MHz at 20nH and 35.6 MHz at 200nH. Quite sweet spot I think. Just be sure to make proper fixture and get calibration inductor set. I would find someone with hi-end LCR bridge to measure for me couple of calibration inductors, perhaps regularily. Then I can get away with stable but not so precision capacitor. Oscillator will have parasitics anyway, so calibration of such instrument is must.