• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 46 subscribers
  • Views 1056 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 Amp with 1 GOhm Gain and 50 kHz Bandwidth

Andy Bartfay
Prodigy 150 points
Other Parts Discussed in Thread: INA116

Hello.  I'm trying to assess the feasibility of building a transimpedance amplifier with a gain of 1 GOhm and a bandwidth of 50 kHz.   I need to amplify sub-picoampere currents (500 fA - 100 fA) generated by an ion detector that has about 25 pF of capacitance.  At first glance, this seems possible given an op amp with high GBW.  However, compensation may be an issue.

The equations in http://www.ti.com/lit/an/sboa122/sboa122.pdf suggest a minimum GBW of approximately 400 MHz given my requirements.  But the respective feedback (compensation) capacitance is in the range of a few femtofarads, which is likely even less than the parasitic capacitance of the feedback resistor.  A higher GBW would help but not much.

Am I interpreting the equations correctly?

Thanks.

  • 0
    Mastermind 7395 points

    Maybe you could achieve those results with several stages. For the first stage, you could use the INA116 which has an extremely low offset current.

  • 0 in reply to Avi Chami1
    Guru 21975 points

    Andy,

    Albert's suggestion on the INA116 will not work as it is an instrumentation amp and cannot be configured as a transimpedance amp.

    You have a very difficult target and I have serious doubts that it can be achieved. To be honest, I did not check the equations that you cite. I'm confident that you would need a very high bandwidth amplifier. But these equations are not likely to adequately describe the complexity of the situation. The problem will be a combination of achieving extremely low parallel capacitance while avoiding distributed capacitance along the resistor to ground. This distributed capacitance forms a delay line that greatly complicates the stability issues. The parallel capacitance can be reduced with multiple feedback resistors in series but this complicates the issue of distributed capacitance. In monolithic circuits that are affected by similar issues we use a guard drive resistor under the feedback resistor to bootstrap the distributed capacitance. The guard resistor is a comparatively low value resistor from output of the amp to ground placed under the feedback resistor.

    I have little idea of what can be practically achieved in this area but I fear it would fall well short of your goal.

    Regards, Bruce.