1
\$\begingroup\$

Since I'm not very much aware of opamp internals. I might have a bit difficulty with expressing my question good enough but I will try with an example. This is mostly about opamp's response to sudden changes and opamp BW. So to understand better I want to ask with an example made up by me.

For example, consider a TIA as shown below and imagine the input current Iin is a current pulse and the parasitic capacitance between the non-inverting input and ground is 5nF. And current source's pulse duration is 1ns, period is 500ns, rise/fall time is 100ps and it peaks to 10mA.

TIA schematic

Source webpage: Wikipedia - Transimpedance amplifier
Picture attribution (per Wikimedia Commons licensing requirements): Zen-in at English Wikipedia, TIA simple, CC BY-SA 3.0

I have two questions (especially the second question):

1) In this case, will the current pulse shape through the feedback resistor Rf change due to the capacitance and Rf or input impedance of the opamp? Will this already act like a filter for the pulse?

2) This is more challenging to me. Now at Vout if I want to measure/monitor the average current (average of Iin), and if response time of the average is required as anything above 50ms; what should I look into as opamp parameters. Should the opamp still be fast enough (slew rate, BW etc). Or any opamp will be fine and should I just add a capacitor between non-inverting input and the ground? But then might stability be a problem?

So basically, for the TIA above and for monitoring the input pulse current's average, what path should I follow?

\$\endgroup\$
4
  • \$\begingroup\$ user4444 - Hi, Where did that image come from? To comply with the site rule on referencing, details of the original source of copied / adapted material must be provided by you, next to each item. If the original source is online (webpage, PDF, video etc.) please edit the question & add its name & link (URL) (e.g. website name + webpage title + its URL). If the original source was a book or other offline material, please edit the question & add the best reference you can e.g. title, author(s), page, edition etc. (see the linked rule for details). TY \$\endgroup\$
    – SamGibson ♦
    Commented Apr 28 at 17:04
  • \$\begingroup\$ ok how does it look now? feel free to edit. \$\endgroup\$
    – user4444
    Commented Apr 28 at 17:07
  • \$\begingroup\$ user4444 - Thanks for adding that. Just FYI, copying images from Wikipedia (actually it's from the associated "Wikimedia Commons") means that they have additional requirements for the licensing of copying from them (see here - en.wikipedia.org/wiki/File:TIA_simple.svg ). I've added the appropriate text, but just FYI, if you don't follow the licensing of the source then that could cause problems for you, even if you comply with the minimal requirements of where you're copying to i.e. Stack Exchange. Anyway, all done. Let me know you've read this and I'll delete the comments. TY \$\endgroup\$
    – SamGibson ♦
    Commented Apr 28 at 17:22
  • \$\begingroup\$ It's so simple to just use a simulator and get a very reasonable answer. Play around with the op-amp parameters and repeat. \$\endgroup\$
    – Andy aka
    Commented Apr 28 at 17:22

1 Answer 1

Reset to default
1
\$\begingroup\$

A TIA is really good for averaging current. Just put a capacitor in parallel with the feedback resistor. Choose the resistor to get the average transimpedance you want. Choose the capacitor so that RC is the time constant of the average you want. The opamp only has to be fast compared to the averaging time.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks for the answer. Is C total capacitance? When you say RC, Is C = Cf + Cj? Cj is the parasitic capacitance between non-inverting input and ground. I tried to simulate now in LTspice using LT1006 with 5V single supply. I noticed if I dont add any Cf it oscillates. \$\endgroup\$
    – user4444
    Commented Apr 28 at 18:00
  • \$\begingroup\$ @user4444 Just $C_f$. $C_j$ affects stability but not averaging time. \$\endgroup\$
    – John Doty
    Commented Apr 28 at 18:03
  • \$\begingroup\$ Ohh I see so I need to choose the value of Cf such that it should both remedy the oscillation and also will determine the averaging time along with Rf correct? Can I rely on simulation for it or should I check on breadboard for the Cf value? And does gain BW product play any role in my case? Planning to use this opamp analog.com/media/en/technical-documentation/data-sheets/… \$\endgroup\$
    – user4444
    Commented Apr 28 at 18:08
  • \$\begingroup\$ @user4444 Roughly speaking, you want a unity-gain stable amp with GBW greater than $$1/RC_f$$. \$\endgroup\$
    – John Doty
    Commented Apr 28 at 23:48

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.