What drives mechanical piezo device ultrasound maximum (sound) power output - electrical power or voltage?

Question

Many definitions of piezo devices say that they convert mechanical energy to electrical (voltage) potential energy. This gives the impression that you want a high voltage to drop onto piezo device when driving it from an amplifier.

How do I drive the piezo device electrically to deliver maximum (sound) mechanical energy in a medium such as water?

If my amplifier is the source and the piezo device the load, do I want a low source impedance and high load impedance to get maximum voltage to drop on the load?

From my RF background you always want maximum power transfer from source to the load, thus you would like to have the source and load impedance to be matched.

Therefore, I will always want to use matching circuits to match 12Ω source impedance of the amplifier to coaxial cable characteristic impedance and the load (150Ω +-) to cable characteristic impedance.

My tranducers ranges from 0.1 MHz to 15MHz and cable lengths are maximum 3.5 meter coaxial.

In the last scenario I will always maximise the sound power delivered in to the medium, or am I missing something here?

Answer 1

Transducers for use in water as projectors have a sensitivity specification called Transmit Voltage Response (TVR) which relates the rms voltage applied to give a Sound Pressure Level (SPL) at 1 meter distance. TVR is usually given in \$ dB \; re \: 1\mu Pa /V \; @ \; 1m \$. TVR changes over frequency. If the transducer is driven at 2.3 V_rms and the TVR = 188, then the SPL at 1 meter from the transducer will be \$ 20 log(2.3) + 188 = 195.2 dB// \mu Pa \$.

How much power you need depends on the impedance of the transducer, which is generally complex, and the drive voltage. You can use a tuning inductor to remove the reactive part of the impedance so you are driving a pure resistance (only valid about one frequency).

The maximum source level attainable depends on the cavitation threshold and heating of the transducer.

At lower frequencies, the length of the cable isn't an issue. Thus, use a low impedance amplifier to drive the transducer. As mentioned above, a tuning inductor can be used to remove the reactive portion of the impedance. This will present a kinder load to the amplifier and also give voltage gain if there is a significant reactive component.

At higher frequencies, the 3.5 m (I assume "me" means meters in your post) cable length will be close to 1/3 wavelength at 15 MHz which can cause matching issues. You may need to make matching networks (tuning inductor & transformer) to protect the amplifier and/or reduce nulls in your output voltage. At a minimum, match the amplifier output impedance to the cable to absorb the return reflected signal.

Answer 2

From my RF background you always want maximum power transfer from source to the load, thus you would like to have the source and load impedance to be matched.

There are two answers, yes, and no.

If you have a fixed source impedance, then yes. A conjugate matched load impedance will give you the largest power transfer.

If you have a fixed load impedance, then no. The lower the source impedance, the more power you can put into the load.

However, there are other reasons for wanting matching between source and load if there is any significant electrical length between them, and that is to have a flat frequency response to the system.

A piezo device tends to be relatively high capacitance, which makes it difficult to drive well. If you were operating at a fixed frequency, then the answer is obviously to resonate it with an inductor, to make it look resistive at the operating frequency.

The trick to operating it wideband is to design a low-pass filter, to work between a finite input impedance and an output open circuit (most online filter design resources will allow you to do this), which will have a shunt capacitor on the open circuit end. Obviously the corner frequency of the filter must be above the top of your operating frequency range. Choose the input impedance of the filter to give an output capacitor value equal to or greater than your transducer's capacitance. Add extra shunt capacitance to the transducer to bring it up to the filter's required capacitance if the latter.

Now design your source and transmission line with the filter input impedance. Using a lower source impedance will get you more power, but the filter won't be properly terminated, and the frequency response will be non-flat. If you need a lower impedance transmission line to match the filter, you can connect several in parallel.

Power or voltage? Of course you need both. Although voltage equates to displacement on the piezo, getting that voltage there needs a certain amount of power.

Answer 3

Exactly, the condition for maximum power transfer is matched source and load impedance.

(the cable impedance probably won't matter - at 100 kHz, the wavelength is km, and even at 15 MHz, a wavelength is ca. 20 m in the cable, and my guess is that at \$Z_{load} \gg Z_{cable}\$, the mismatch won't matter, especially since sink impedance probably becomes quite a bit imaginary at higher frequencies).

Note that just like an antenna is an impedance matcher from waveguide impedance to free-space impedance, your Piezo actuator is an impedance converter from waveguide to free-space/free-water impedance, and its impedance depends on the medium. So actuator-moving-air impedance is probably significantly different from actuator-moving-water impedance. I don't know too much about Piezo oscillators, but my guess is you'll be the one in charge of figuring out a matching! So, ideally, you build appropriate matching networks and experimentally minimize the standing wave voltage on the coax.