For fun, I'm working on an audio project which calls for a differential voltage amplifier - something I don't have experience with yet. The circuit below is no good, because the differential amplifier has relatively low input impedance, and it unacceptably loads the high-output-impedance source. (Moreover, the input impedances of the two inputs are unequal.) This circuit only works well when the source's output impedance (modeled by R5 and R6 here) is very low:


simulate this circuit – Schematic created using CircuitLab

My question: is there a way to modify the differential amplifier to have very high input impedance, like a traditional unity-gain single-ended op amp circuit? I know I can simply add an additional pair of voltage follower op amps at the inputs, like this...


simulate this circuit

... but if at all possible I'd like to avoid putting an additional op amp in the signal path! Thanks in advance.

  • 4
    \$\begingroup\$ where you're going is starting to look like the classic 3-op-amp Instrumentation Amplifier ...... you could potentially use a simpler pair of followers, like a matched pair of transistors / jfet's \$\endgroup\$
    – Pete W
    Commented Sep 8, 2021 at 22:47
  • \$\begingroup\$ @PeteW Thanks, I'll study the instrumentation amplifier! If that's the way to go, which type of follower would be preferred for minimum added noise? \$\endgroup\$
    – TypeIA
    Commented Sep 8, 2021 at 22:55
  • \$\begingroup\$ JFET's can get you very low noise if you need it, but really is outside my experience. If you want a decent differential amplifier quickly for a one-off or lab project, just buying one might be easiest - there are many specialty differential amps, audio amp's, INA's and so forth. \$\endgroup\$
    – Pete W
    Commented Sep 8, 2021 at 23:38
  • \$\begingroup\$ Moreover, the input impedances of the two inputs are unequal - that is incorrect for a balanced drive signal. \$\endgroup\$
    – Andy aka
    Commented Sep 9, 2021 at 7:33
  • \$\begingroup\$ @Andyaka Interesting - then in the simulation of the first circuit above, why is the current through R5 about twice as high as the current through R6? (Also, if you're the downvoter, please let me know how I can improve the question. I tried my best and am happy to learn!) \$\endgroup\$
    – TypeIA
    Commented Sep 9, 2021 at 11:18


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.