Driving capacitor with class-D

[highfieldrebel]

I've asked this already in Pass Labs, but then thought this might be a more natural home...

I need to drive a piezo actuator (part of an optical filter) with a 30V pk-pk sine wave at between 20 kHz and 50 kHz. The spec sheet says the piezo will show as a capacitance of 0.4 uF. I can envisage one issue with the low-pass filter class-D amps have to filter away the PWM frequencies. Other than that, would a class-D be able to drive a load like this?

I calculate a reactance of 8 ohms for 50 kHz.

Thanks for your help

[Bone]

You'd have a job to find a Class D kit amp to amplify up to 50kHz. The sampling frequency would be too low.

[Saturnus]

A piezo is actually electrically identical to a variable resistor in series with a fixed capacitor. There are also some parasitic parallel capacitances and resistances but they are very minor.

The variable resistor part is the loss over the piezo which depends on the output.

[highfieldrebel]

Originally Posted by Saturnus

A piezo is actually electrically identical to a variable resistor in series with a fixed capacitor. There are also some parasitic parallel capacitances and resistances but they are very minor.

The variable resistor part is the loss over the piezo which depends on the output.

interesting, so when you say the variable resistor part depends on the output, you mean it depends on the output frequency or the output amplitude? Or both?

Is is possible to roughly estimate the size of the variable resistance? It might help the amplifier

Thanks
Steve

[Saturnus]

It varies with output amplitude but that will also vary with frequency as a fixed voltage will generate different output amplitude at different frequencies but it's pretty linear and the impedance from the capacitance will be the main effect except at the frequencies you're looking into above the audible range.

For example, a piezo like an old Motorola KSN1001 is electrically identical to a 33ohm resistor in series with a 0.3uF capacitor, so at the frequncies you're looking into the resistance part will be the main part of the impedance.

[highfieldrebel]

Originally Posted by Saturnus

It varies with output amplitude but that will also vary with frequency as a fixed voltage will generate different output amplitude at different frequencies but it's pretty linear and the impedance from the capacitance will be the main effect except at the frequencies you're looking into above the audible range.

For example, a piezo like an old Motorola KSN1001 is electrically identical to a 33ohm resistor in series with a 0.3uF capacitor, so at the frequncies you're looking into the resistance part will be the main part of the impedance.

excellent, thanks Saturnus, that should help my search for an amp, but maybe I'll need to go higher in voltage also

[trevmar]

I used to design Class E amplifiers to drive capacitive and inductive loads (RF, mainly). It is possible to use Class D to drive a reactive load, but difficult.

Maybe this page can help you. Scale the frequencies and inductor to match your capacitance:
http://www.tonnesoftware.com/classe.html