OPA2604 gain clamping by a PZT device

Changsu,

It is not clear from your schematic if you operate OPA2604 on the dual supply, +/-24V, or single supply - you show only a positive supply so please clarify it.  Since the effective bandwidth of OPA2604 is a function of the close-loop gain, please provide us with the values of all the components used.

Also, it would be helpful if you could label the vertical axis of the above waveforms, or provide us with info regarding input/output signal magnitudes, so we may make sure you do not violate the OPA2604 linear ranges (see PDS table below); this could also explain the cause of the distortion seen at the output of OPA2604.  Btw, what is the reason behind 6.8k resistor at the output?

Thanks for the reply, Marek.

RLC pair is just a duplicate. By simulation, the transmission peak is nearly 60kHz.

+/-24V is properly supplied to the op amp. Sorry for not displaying the -V in the schematic. It's just a non-inverting op amp. Due to the high output voltage and bandwidth, I chose this op amp.(+/-24V), and only one op amp was used.

In the clamper circuit. I could find, in several websites and textbook, that this scheme is typical. As far as I understand, it looks like a C-R high-pass filter for AC component, and safety load for DC component.   

It should be noted that the whole circuit worked very well when I tested it with an oscilloscope. So I was about to solder every circuit, but final test with the actual PZT component showed clamped V_pp.

So I don't think it's a problem of op amp or its operation.

Before building this circuit, I used the function generator and linear amplifier to apply the sine wave (0-15 V_offset, 0-10 V_pp) to the PZT component. It means that the PZT component also can accept that voltage ranges.

While building this circuit, I think I obeyed the op amp manual. Input voltage to the op amp is fixed as about 800mV. As I adjusted the trimpot in the amplifier, output voltage was linearly ampllified up to ~20V_pp, after that voltage it became distorted. Clamper circuit also worked well enough when I test it with the oscilloscope.

The example graphs in the figure were measured when the circuit was connected to oscilloscope. ~800mV_pp before the op amp. ~8V_pp after the op amp, ~7V_pp after the clamper, ~11V_offset after the clmaper.(slight loss along the clamper part, but reasonable) 

Every thing was normal until I connected the PZT component to the circuit output. V_pp couldn't be increased over ~4V_pp while V_offset showed normal adjustment range up to ~20V_offset.

I guess it's due to kinds of reflection by impedance/power mismatch between the circuit and PZT component. To mimic the oscilloscope, I tried after connecting 1Mohm and 10pF to the PZT component in parallel which was in vain.

I'm very confused that it works well when measured with the oscilloscope, but not with the actual component.

Maybe I miss very basic thing, or matching between the circuit and component is tricky.

looking forward to hearing easy solutions!

Changsu,

I would be really helpful for you to provide the schematic with all the components values - a picture is worth thousand words. Without the details I asked for, I have difficulties understanding how you exactly make the measurements.

I am not clear whether you connect the PZT to ground as shown on the simplified schematic below - if so, you clearly create a voltage divider between the C1 and PZT where the output voltage magnitude is close to 4V as you claim - see below.

In order to eliminate PZT loading effects, you may have to buffer it using another op amp as shown below; of course, this will require a different power supplies for the buffer (20V and 68V) to accommodate the 24V level-shifted output voltage range.

In both cases, you may be also experiencing stability problem since very few op amp, including OPA2604, is able to drive capacitive load in the range of uF.  But this is quite another issue that will require special compensation.

Attached please find the simplified Tina-TI schematic you may use for your own simulations.

Thanks Marek for the simulation.

Your simulation clearly shows why I got ~4V  at the output port.

I grounded the PZT properly, and final waveform was also a clean sinusoidal up to ~4V like your simulation. The distorted signal that I mentioned before was not at the final output, but before and after the op amp part.

According to your suggestion, I need to add a buffer at the output. I'll post again if I can get desired signals.

Thanks for the link of the simulation tools too. It may be very helpful to me.

Changsu.