Driving Piezoelectric Transducers

Hello to everybody, as a part of my thesis work I'm trying to design a device for driving piezoelectric transducers, to be used as a portable non-destructive testing device.

I have lots of troubles in finding some good technical documentation about probes. I'm quite confused so I think it's a good idea to ask for help, before I'll go brain-dead.

I tried looking on several web sites of some serious producers of piezo-ultrasonic transducers (like Panametrics-ndt) but I haven't found detailed *electrical* information regarding a model I can use to represent the device as bipole. Asking by email was hopeless too.

In a more detailed way:

The transducer is a wafer of piezoelectric material between two metalized layers used as electrode for polarizing the active one (so applying a voltage we can produce a thickness-deformation, in this case.)

The first-order equivalent model could be a capacitor; if we want a more detailed characterization, we have to assume some RLC-series branch in parallel with this capacitor to account for the multiple resonance frequecies that the device shows.

The transducer will absorb a very low power from the driving device. I read somewere (I can't remember) about 250mW. Piezoelectric properties could be lost forever if the device dissipate too much power in heat (If the temperature is over the "Curie temperature" for too much time, we'll assist to a irreversible depolarization of the piezoelectric layer.)

I wasn't able to find these informations and characteristics parameters on the datasheet (ehr, I wasn't able to find the datasheet too) of the probes I found, because producers usually assume that you'll plug these device in yet-built pulser/receivers. So you'll not need these information.

I tried to do some "reverse-engineering work" to discover something regarding those devices, looking at the characteristics of some pulser/receiver (like those about Ritec or Panametrics-ndt) or power amplifiers usually used to drive transducers.

I discovered that all those devices are assumed to be closed on 50 ohm impedance.

Here is the origin of all my doubt!

Why they assume to close the pulser on a 50 ohm bipole, if the transducer is modeled as a capacitor?

Have I to worry myself about that or it's just a normal "loading condition?"

Perhaps in the transducer is included a matching network so the impedance we see looking in the bnc is just 50 ohm?

But it's quite strange, to me, because if this is the case applying a

200 V peak voltage, the power absorbed by the transducer is

P = (V^2)/R = 200*200/50=800 W

a rather huge power...

Perhaps the 50 ohm load is just a standard way to describing the characteristics of the pulser, so I haven't any matching network? But in this way we haven't any matching condition realized (nor maximum power transfer, nor minimum distortion)

I'm really confused...

Is there anybody that knows links or articles about the information I need?

Thanks in advance for your help,

Pierpaolo

Reply to
George Gabriel
Loading thread data ...

information.

condition?"

need?

Your simple capacitor model is no good at all when you are driving at high power and want reasonable efficiency since this will usually involve operating the piezo transducer at or near it mechanical resonant frequency.

There is a lot of tutorial stuff on

formatting link

start at

formatting link

which may help you.

Good luck.

Michael Kellett

formatting link

Reply to
MK

--
http://www.ctscorp.com/techpapers/piezotechprimer.pdf
Reply to
John Fields

Here's a link to an Apex app note that might help.

formatting link

Don Cleveland

Reply to
Don Cleveland

Hello George,

They need to be driven like transducers in a medical ultrasound system. Basically there are two schemes, unipolar or bipolar. The lower cost version is unipolar and it works well. A FET that can withstand the required pulse voltage has its drain connected to a large positive voltage via a choke. The drain connects to the transducer. Another LC (or LR) combination can be used to get rid of the DC component on the transducer. Now a brief pulse or a burst of pulses is applied to the gate. It is important to drive this gate hard. The pulse length is roughly compliant with the expected center frequency of the piezo. In case of materials such as PZT-5 you can calculate that from the thickness.

As for voltage level it is not uncommon to pulse PZT-5 in excess of

100V. Please be cautious with these voltage levels. Especially at DC these voltages can be very hazardous or even lethal.

Also, when sending bursts you have to mind the dissipation in the FET. While Rdson might be low enough much of the losses there will be caused by the slopes of the drive signal where the FET has to transit through its linear region.

Regards, Joerg

formatting link

Reply to
Joerg

Thanks to all people answered me, now I'm reading the material with much interest. Expect new questions soon! :)

Reply to
George Gabriel

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.