# Impedance matching for piezo element

I am trying to boost audio signal from around 3V to around 60V for a piezo element with a component as small as possible.

I found audio transformer MET-23 that reflects 1.6 kOhm impedance on primary when there is 3.2 Ohm load on seconday (http://www.tamuracorp.com/clientuploads/pdfs/engineeringdocs/MET-23.pdf). Music player says that the impedance should be 32 Ohm.

This is my train of thought: I was planning to hook up the transformer backwards to boost the voltage. Piezo element which acts like a capacitor produces capacitive reactance which is xc=1/(2*pifC) where f=1kHz (sound average frequency). Secondary impedance should be then |sec coil inductance - piezo capacitance| (if I don't count the resistance). That impedance should reflect on the primary so primary impedance is sec impedance*(1/22,4)^2. If the primary impedance is below 32 Ohm then I can just add a capacitor to primary.

Is this correct way to do it?

• What bandwidth do you need for your audio signal? Apr 28 '16 at 22:53
• Sound will be human voice so the frequency will be 300 Hz to 3400 Hz. Apr 28 '16 at 22:56

The transformer has the specified (rather large) bandwidth when it is loaded to some active load (resistor). This mode is assumed in its datasheet.

Your load is a capacitor. It is not an active load. So the combination of transformer and piezo element capacitance works as a resonant circuit. We do not know the transformer's inductance, I do not know piezo element capacitance. Probably it is specified on its datasheet. So it is not possible to predict the system performance. May be it would make one frequency only.

This problem has to be solved by introduction of active losses to some place. The first idea is: to add a resistor in parallel with piezo element. The starting value is about 100 kOhm. You have to check the acoustic bandwidth, if it is good, you can increase this resistor. If it is bad - you decrease it.