Myron,

I built a set of tuned rings to couple a 4MHz signal from an ultrasound
transducer to its processing/signal generation box. I used a small
copper ring (only about 6") split at one point with a resonating
capacitor across the gap on the ultrasound box side. The coax was
connected across this gap also.

On the transducer side, use an identical loop and capacitor but make a
series resonant circuit (capacitor in series with one of the feedline
leads). Loops were separated by about 1/4 inch.

I know this configuration (parallel resonant on the transciever side,
series resonant on the "antenna" side) worked the best.

It worked well in terms of passing the signal through undistorted
(looked at amplitude and phase on both sides of the circuit, only got a
bit of a phase shift, no amplitude loss)

I don't know about power handling. If you use hefty capacitors, there
should be no problem. Bandwidth could be quite low with high-Q
circuits, but maybe you could use variable caps.

I think your idea with capacitative coupling with inductors to tune out
the reactance would work fine too, but I haven't tried that.

Anyway, the inductive thing works OK. I got the idea from a older
(1950's?) ARRL antenna book that touted it as a way to allow rotation a
parallel-line fed beam.

You'd certainly need a set of coupling rings for each band of interest,
or switched or variable capacitors.

Low E glass would get some eddy currents going. I don't know how
conductive it really is though, so ... who knows?

73,
Dan
N3OX
www.n3ox.net