Measuring Distributed Capacitance of an Inductor

Parallel resonant frequency.

Tim

-- Deep Friar: a very philosophical monk. Website:

I figure there's two way to do that. Using generator + scope.. or putting the inductor in an oscillator.

If you just connect a scope across the inductor, you'll probably see ringing noise from ambient crud, and you should be able to estimate its period to a few per cent. Then, for more accuracy, hold a lead from a signal generator nearby and explore around that frequency for a resonance.

Subtract out the scope capacitance.

John

Measure the resonant frequency, and for grins, its resistance, which will slightly alter the non-resistive resonance.

I get it... Neato! Thanks. :)

--- If the inductance doesn't change with frequency, you could measure its self-resonant frequency and then rearrange:

1 f = --------------- 2pi (sqrt LC)

to solve for C:

1 C = ---------- 2pi f² L

JF

From wirebond experiments, I'd recommend measuring impedance at several frequencies within the range of interest and fit a multi-lump equivalent. A single resonance rarely fits the whole spectrum. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |

If this is the working inductor for a switcher, the resonant frequency is what you want. The capacitance you get from that will be accurate enough.

If it is for some sort of filtering circuit where the input is more like a sine wave, you are best off to measure the impedance at the working frequency and any harmonics you care about. It is not uncommon for an inductor to have a rise in impedance a little sooner than you would expect from the simple parallel LC model. The losses reduce the inductance of the "L" part at high frequencies.

Late at night, by candle light, John Fields penned this immortal opus:

Make that

1 C = ------------------- 4 * pi^2 * f^2 * L

- YD.

--
Remove HAT if replying by mail.

Or how about:

C = (w^-2)*(L^-1)

You have been around here for long enough, provide a decent overview of the test and measurement equipment you have available to do this task.

I have a scope and generator.

I did what J. Larkin posted in this thread.

Here's the procedure I used: Put scope probe across the coil. Put a generator (on sq wave) clip near the coil. Look for ring on scope at generator transitions. Measure ring frequency. Do math to factor out the probe capacitance.

The generator doesn't load the coil too much due to the loose coupling (magnetic coupling?) between the generator and coil. If I got artsy, I would bring the generator probe to get just enough coupling to get above the noise to make a good ring frequency measurement on the scope.