Op Amp Instability When Driving Piezoelectric Patches With Capacitive Loads and Feedback

Hi all,

I'm new to the forum, I hope this is the right section. Recently I've been trying to use analog circuits to make a simple proportional controller on a beam, using one patch to sense deformation and another one on the other side (colocated) to actuate.

As suggested in some books, I used a charge amplifier to sense the charge and an inverting amplifier in series in order to obtain the correct voltage sign. However I experienced instabilities probabily due to the interaction between the capacitive load - i.e. the piezo patch - and the output op amp resistance: the system, while not actuated, started oscillating at a frequency = 1/(2pi*R0*CL), being R0 the output opamp resistance and CL the load capacitance. The problem is partially solved by adding a resistor at the end of the circuit, but this lowers the total gain by an unacceptable amount.

Plus, I suspect a similar coupling even between the input piezo patch capacitance and the input op amp resistance, because when I add more piezo patches pairs and relative circuits I have to further increase the added resistor value in order to keep it stable (but again, the circuit gains get lower and lower).

I understand that there may be available for purchase dedicated opamps to overcome this problem. Do you have any advice? Or any reference on the topic? 

Thank you all in advance,

ADDITIONAL INFO ABOUT MY SYSTEM (figure attached)


piezo capacitance = 10nF (https://www.steminc.com/PZT/en/harvesting-plate-30x30x1mm-2-mhz)
sought circuit gain = [non inverting and as high as possible, compatibly with the fact that the typical voltage measured on the sensing patch is between 0.5 to 1 Vpp and compatibly with the maximum op amp voltage supply (in my case +/-45V). Anyway, from simulations, a minimum gain of 10 should be required, and that's the one I tested]
frequency range: 100-9000 Hz approximately. Notice that I'd like to have ideally no phase delay between Vin and Vout.
op amp: OPA445


PS: keep in mind that I'm a mechanical engineer and most of this stuff is relatively new to me...

Putting aside any other considerations for what you are trying to achieve, you cannot drive piezo devices with most op amps, they will not tolerate much capacitance directly on the output before oscillating. You can find plenty of articles on the topic on the web. There are some op amps that can handle capacitive loads however, I would refer you to the following app note.
http://www.ti.com/lit/an/snoa808/snoa808.pdf.
The roll off of the transimpedance amplifier is close to 100Hz, there is quite a bit involved in designing a transimpedance amplifier, again, see this app note, there's a lot more around on the web.
http://www.ti.com/lit/an/sloa033a/sloa033a.pdf
As regards to the whole concept of what you are attempting, please give more details, since there are a lot of problems in the existing setup. Be warned that simulation is only as good as the models you have,
cheers,
Richard

Thank you for your reply, I will look into the first note (I've already read the second one but problem persist).
With respect to my work, I'm only trying to stiffen the whole structure by using active control: this can be achieved by applying the same voltage generated on the sensing patch (which is proportional to charge, ie strain and displacement) multiplied by a gain to the actuated patch. As far as I'm concerned, I don't care how or with which strategy this is accomplished, as long as it works (this just to say that I don't have necessarily to use the circuit above, on the contrary if you have better solutions please share them).
PS: when I said simulations I meant "mechanical simulation", that is no real circuit is involved, I just assumed that I was able to apply a voltage to the controlled patch directly proportional to the patch deformation; reality turned out to much more complicated however...