Hi all,

I'm in the midst of winding a bunch of little coils on ferrite toroids,
and was pondering how to measure their inductance and adjust it (by
changing turns spacing around the toroid) to the desired values. I
have an HP4274A LCR meter, but its test frequency only goes up to
100kHz, and I wanted to measure the coils closer to their ~10MHz
operating frequency. It occured to me that I could make the
measurement with a one-port network analyzer measurement. If I
resonate the coil with a capacitor of known value (C1) with one end of
that tank grounded, and put another capacitor (C2) in series from the
top of the tank to the VNA port, and a third capacitor (C3) shunt
across the VNA port, by adjusting C2 I can get "perfect" return loss at
some frequency. When I do that, the values of C1, C2 and C3, and the
frequency at which the return loss is infinite, will tell me not only
the inductance but also the Q at that frequency.

I guess I can write a general expression for the impedance the VNA
sees, in terms of C1, C2, C3, L and Rseries, and the frequency, and set
that equal to 50 ohms, plug in the capacitor values, and calculate L
and R. But for now I'm just using an easy RFSim99 model, with the
inductor "physical model" selected so I can set Q and frequency at
which Q is applied. I just set the capacitances as they are in my
measurement circuit (with C2 estimated from the rotation angle of the
little air trimmer I'm using there), and adjust the coil inductance and
Q in the simulation till the return loss null is deep and the frequency
matches the observed VNA frequency for that null.

My coils are on T-25-6 cores, around 1uH, and I'm using 180pF for C1,
150pF for C3, and can vary C2 from about 12 to 30pF. Even with my
crude reading of C2, I can determine the Q within 10% tolerance, which
is plenty accurate enough for my work. This same method should work
well for coils over a pretty wide range of Q, inductance and operating
frequency, so long as you have capacitors with Q much higher than the
coil (or known capacitor Qs), and the parasitic reactances around the
circuit are not out of control.

Lest you think a VNA is out of your reach, check out ebay auctions and
the very nice homebrew VNA described at
http://users.adelphia.net/~n2pk/. It can be a very useful piece of
equipment for an RF homebrewer.

Cheers,
Tom