I'm designing a trans-impedance amplifier (TIA) for a fairly fast Avalanche Photodiode (APD), to amplify pulses of 1mA that last few nanosecond. I found a solution that have just enough bandwidth with a single-output op-amp, but I want the output signal to be differential, and I haven't found any integrated TIA that meet my specifications.

I looked at several Fully Differential Amplifier (FDA), like the LMH5401, and designed a circuit close to that one:


simulate this circuit – Schematic created using CircuitLab

I could not find any example of a similar circuit. While I can pick \$Z_{f+}\$ and \$Z_{f-}\$ to achieve the performances I want, I'm not sure how I should check my design.

  1. How can I check that the circuit is properly balanced?
  2. Does \$Z_{f+}\$ and \$Z_{f-}\$ have to be equal?
  3. Is there any show stoppers that a SPICE simulation won't catch?
  • \$\begingroup\$ Are you just looking for a consistent, fast response that can be used to accurately measure time between events? Or are you looking to measure the pulse height accurately? Or what? (It may help others to know what's important to you and what is not so important.) \$\endgroup\$
    – jonk
    Aug 23, 2017 at 23:27
  • \$\begingroup\$ @jonk The main objective is to time pulses. I also need a rough amplitude estimation: The required dynamic range is large, and because of that, the TIA is followed by a programmable gain amplifier that must be adjusted. \$\endgroup\$
    – pserra
    Aug 23, 2017 at 23:32
  • \$\begingroup\$ I think the burr brown app note "photodiode monitoring with op amps" has a differential TIA in the end. \$\endgroup\$
    – DavidG25
    Aug 24, 2017 at 4:28
  • \$\begingroup\$ At nanosecond timescales, parasitic inductance and parasitic capacitance that you've neglected can easily throw a SPICE simulation's veracity in doubt. \$\endgroup\$
    – glen_geek
    Aug 24, 2017 at 12:36

2 Answers 2


You can use a FET opamp, like OPA659 and then buffer the output with another opamp with differential output. Not sure if LMH5401 is adequate for TIA.

EDIT: Have you seen the OPA857?

Maybe you'll find useful this document

  • \$\begingroup\$ I have a backup plan around the LTC6268-10, but I'm looking at other options to get a larger bandwidth and a more compact circuit. \$\endgroup\$
    – pserra
    Aug 23, 2017 at 23:33

Texas Instrument made a reference design with a very similar circuit in December of 2017, using the LMH5401. It answers all my questions.


Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.