• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 542 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Transimpedance Amplifier problem

james gannon
Prodigy 20 points


I have a circuit with very high gain, 10 GOhm resistors. The photodiode has a very large  cap. value 380 pf. I have compensation across the feedback resistor and I am using a OPA129 amplifier. When I put a step into the system there is a slight overshoot/ringing at the beginning and then a very long slow slope. The input signal has a time constant of 1us.

I was wondering if there where any suggestions for attacking this problem. We need the output to settle to < 0.5% within a second, the risetime looks fine, it is this low level ringing and drift that is the problem, on the order of 1%

 

Regards,

Jim Gannon

  • 0
    TI__Genius 12345 points

    James,

    One of the best methods I've seen to deal with this problem is to bootstrap the photodiode capacitance using a wide bandwidth, low noise JFET (The BF862 from NXP is a good choice). There is a very short app note on this here: http://cds.linear.com/docs/en/design-note/dn399f.pdf. The JFET reduces the effective capacitance of the photodiode, allowing for a smaller compensation capacitor across the feedback resistance. 

  • 0 in reply to John Caldwell
    Prodigy 20 points

    Thanks John,

     

    But I have even used that circuit before, but with a smaller feedback resistor. I think I would have problems with leakage currents if I were to put it on the board, since my feedback resistor is so big. If it weren't for that I think I would be okay, since I have a DAC providing an offset current which could be used to dial out the Igs current of a few pA.

     

    Regards,

    Jim Gannon

  • 0 in reply to james gannon
    TI__Genius 12345 points

    Jim,

    The only other option I can think of at the moment would be to partition the gain between two stages instead of one. For example you could reduce the gain of the transimpedance stage to 100M (and get the associated reduction in the feedback capacitance), and then follow it with a voltage amplifier with a gain of 100. The downside here is increased noise but it would help address the settling time issue.