Norm, you wrote that the resistance of the "primary" circuit is very
critical, suggesting silver tinned stranded wire to get resistance down
as low as possible. Have you considered using litz wire? This is used=20
extensively at lower radio frequencies to reduce "skin effect" and thus
increase "Q". I've seen the concept used in some switchmode power supply
transformers as well. You're running ultrasonic so skin effect will=20
certainly be a factor.
You also mentioned using a transformer coupled amplifier to drive the
MRA. Possibly a bad idea as the DC resistance of the output transformer
secondary will be in series with your tuned circuit, thus reducing "Q".=20
Also, the inductance of the winding would significantly alter the tuning
of the MRA. As far as I can see, you'd need an amplifier with the lowest=20
possible output resistance and, at the same time, *very* low inductance.
Thirdly, the bridge rectifier you were using. At the frequencies=20
concerned, this would have to be done with schottky or other fast =
turn-off
diodes or you would have major losses in the secondary circuit.
An idea which you may find worth looking into is to build two identical=20
MRA's with the same resonant frequency and have the piezo elements =
"coupled"=20
by placing them in a rigid tube with the spacing set up so they reinforce=
=20
each other, acoustically speaking. This *may* enhance the effect but it=20
would probably make tuning the system considerably harder.
I'm curious about the piezo transducers you used. Are you able to give =
more
details, ie manufacturer, part number and what sort of transducer they =
are=20
(piezo tweeter etc)? When I first heard about the MRA, I did some looking=
=20
around over here and none of the catalogs I checked even mentioned what=20
material was used for the piezo element. Likewise the composition of =
easily=20
available magnets.
The fact that only one in five transducers displayed the effect implies =
that
the MRA relies on some parameter which is normally considered parasitic =
and
hence able to be largely ignored by the manufacturer. Do you have any =
idea=20
as to what this key parameter(s) is/are?
Cheers,
Alan Schneider