Paul,

I probably should have included more details. A fellow I was working with
was trying to extract a clock signal from a synchronous data stream, and was
getting nowhere. We swept the frequency back and forth to be sure he was
tuned to resonance - he was. Changed the ceramic cap to mica, and everything
worked like a charm. We never analyzed why the ceramic did not work, but I
suspect it was because of the capacitance vs. applied voltage dependence. If
the cap had 6VDC on it, and he had a few mv of RF, I expect it would have
worked. Instead, he had 0 bias, and a couple of Volts p-p signal.

BTW, somebody mentioned powdered iron toroids. Sounds like a good idea.

Tam