Building a master-slave amp to drive a piezoelectric driver for a sonoluminescence experiment. Having a problem with the amplification.

The rails are at +/- 45 V (sorted that problem out, thanks to all who helped). Straightforward design from the data sheet of the OPA445 (45 V) op amp, with capacitive stabilization added in.

The chip: http://www.ti.com/lit/ds/symlink/opa445.pdf

And the actual circuit I'm using. The buffer is necessary because the cheap frequency source will feed back if the input impedance isn't very high. All op-amps have trimmer pots attached per the data sheet's instructions, and have been properly nulled.

Circuit Schematic

Circuit is driven around 25 kHz, don't think that is a problem for this chip and circuit - it's designed to drive piezos around those frequencies. At low outputs, it seems to work fine. But once the output voltage between the master and the slave goes to about 70 V peak to peak, increasing the input doesn't increase the output, it just clips at the top of the waveform. I can take some pics of the o-scope output if that would help. As far as I can tell, the output is symmetric to ground (until the clipping starts.)

Any help or info appreciated, will answer any questions I can, but will be on the road over the weekend. Thanks in advance.


  • \$\begingroup\$ The master output begins to clip ? or the slave output begins to clip ? Or both ? What is Load ? The spec sheet shows about 15 mA capability. \$\endgroup\$ – Marla Aug 4 '16 at 3:22
  • \$\begingroup\$ are C2 and C4 4.7 nf ? If so, the impedance of a 4.7nf @ 25kHz is 1354 ohms. At 70 v peak to peak (35v peak, 25 VRMS), current in capacitors would be 0.018mA (18 mA), WITHOUT any further load. Could be a problem. \$\endgroup\$ – Marla Aug 4 '16 at 3:38
  • \$\begingroup\$ A scope shot would be a good idea, removing any ambiguity about what you mean when you say "clipping" and also showing the input waveform (assuming sine). I notice the minimum Full Power Bandwidth is 23kHz, so you could be running into slew rate limiting. If the wave form is looking more like a triangle wave than a sine that's a pretty sure sign. Trying at 10kHz drive for comparison would also likely provide insight. \$\endgroup\$ – scanny Aug 4 '16 at 4:10
  • \$\begingroup\$ Also, as @Marla mentions, those 4.7nF caps on the output are massive. Are those a chosen component or meant to model the load somehow? If chosen, how did you arrive at that value? \$\endgroup\$ – scanny Aug 4 '16 at 4:20
  • \$\begingroup\$ Values for 4.7 nF caps were taken from the data sheet example for capacitive coupling. I'll test it out at 10 kHz tomorrow and see what I get. \$\endgroup\$ – aeronaut Aug 4 '16 at 6:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.