If I try impedance matching the transformer with an ultrasonic transducer . The transducer is approx 20ohms ( set my manufacture) and set the turns ratio of the transformer to match. When a transducer is connected across it my output voltage goes from approx200V to 1000V (transducer ringing) , this seems to make the impednace matching calcs invalid.
So imedance matching doesn't seem to help.
Can someone please confirm my thoughts ?
Most ultrasonic transducers operate at resonance which makes them difficult to match. With a transistor power stage loaded with the transducer, as the voltage at resonance goes quite high, it is the safest way to use a switchable-ratio transformer, also tuned to resonance, for best results.
It was easier with vacuum-tube power stages- they have a high output impedance and are not damaged by overvoltage like transistors...
I used a 2N3055 transistor with a "pulse" transformer, ratio 1: 10approx. with 24 and 40 kHz transducers. THe optimum adjustment was done with another transducer connected to an oscilloscope. (I cannot hear so high ) )
- - - Updated - - -
I refer to ceramic transducers. With magneto-striction ones I used several turns of heavy (2 mm dia.) copper wire, and discharged a 0.1 uF capacitor by a triggered spark-gap. The capacitor was charged to 5 kV. If you want to try...
Make sure that the impedance of the transducer is really 20 ohms and not 20 -j (some number). If it is really 20 ohms then replacing the transducer with a 20 ohm resistor on your transformer should give the same results.
Frank
One problem is to understand the electrical behaviour of a piezo transducer that exposes both a parallel and a series resonance. The below equivalent circuit shows the basic elements. It's very similar to a crystal, but with lower Q. All "mechanical" parameters are influenced by the transducer mounting.