2

I'm designing a transimpedance amplifier but I'm having trouble getting higher gain. I started with Rf being 16.5 kOhms but then I realized I will need higher gain, so I simply switched it out for 50 kOhm, 100 kOhm, and 1 Mohm, but after the 50 kOhm I stopped getting higher gain.

I basically want to take measurements from a photodiode, and I didn't want to mess around with AC so there is Cf to ensure stability and filter any oscillations. So the main question is is there any limiting factor for the opamp that I'm not seeing?

enter image description here

Notes:

  • Vcc = single supply 3.3V
  • Opamp = rail-to-rail MCP617
  • Photodiode = sfh 206 k
JYelton
  • 32,302
  • 33
  • 134
  • 249
OnHai
  • 23
  • 6
  • 1
    Gain at what frequency? What is the application? That's quite a big capacitor there. What type is the capacitor (CF1)? – Tim Williams Aug 30 '23 at 16:55
  • 2
    It looks like you've got a dual op amp. Why not use the second op amp as another gain stage? – Adam Haun Aug 30 '23 at 17:01
  • 1
    There's a good reason why op-amp symbols were invented = clarity in your circuit. How did you measure the circuit gain? Please link a data sheet for the op-amp and optical device. What is your required 3 dB frequency? – Andy aka Aug 30 '23 at 17:13
  • 1
    Please give more information. What is the output voltage with zero light input and with the intended light input with the different feedback resistors. – Kevin White Aug 30 '23 at 17:13
  • Thank you for your comments, frequvency would ideally be under 1Hz, because I dont need high data rate and I want to avoid problems associated with higer frequvencies. The Cf1 is a MLC cap, and thought is that it would filter out any potential oscilations. And regarding the dual opamp, yes it is, but would be better than just getting Rf1 of higher value? – OnHai Aug 30 '23 at 18:09
  • And regarding the way i measured it, the output is directly connected to ADC, and I measure every couple seconds. – OnHai Aug 30 '23 at 18:16
  • And regarding the application, its the sensing circuit for a simple air particle detector that uses IR LED and diffracted light goes to the phodiode. – OnHai Aug 30 '23 at 18:18
  • And regarding the voltage/current input/output, Im not able to measure the photodiode current as I dont have equipment, so im just testing. And ideally it would be that at dark cca. 0.1V and at max cca. 3.3V. – OnHai Aug 30 '23 at 18:21
  • @TimWilliams sorry I dont know if you get a responce without tag. – OnHai Aug 30 '23 at 18:51
  • @Andyaka sorry I dont know if you get a responce without tag. – OnHai Aug 30 '23 at 18:52
  • @AdamHaun sorry I dont know if you get a responce without tag. – OnHai Aug 30 '23 at 18:52
  • @KevinWhite sorry I dont know if you get a responce without tag. – OnHai Aug 30 '23 at 18:52
  • 1
    In such a system I would always take at least 2 measurements, one without IR LED and one with active LED, and process the difference. This removes all sort of side effects like temperature drift and OpAmp DC offset error. And it allows higher gain in an AC coupled second stage. – Jens Aug 30 '23 at 21:03
  • @Jens thank you, im currently taking two measurements and subtracting them. But Im not sure if I understand the AC coupled second stage tho, because the subtracting takes place digitally so its not possible for me to then just amplify the result. But do you think this would be equivalent if i put in high-pass filter and took measurements at about 5kHz? – OnHai Aug 31 '23 at 07:08
  • @OnHai - It would be useful to put any measurements in the question, not in the comments. How do you know the gain is not what you expect? – Kevin White Aug 31 '23 at 17:54
  • @Kevin White - I understand, well I'm simply monitoring output voltage but at the same lightning conditions for the photodiode I got the same output voltage at 50kOhms and even 3MOhms. – OnHai Sep 01 '23 at 07:56
  • @OnHai - Put the ACTUAL measurements in the question. Don't just say "the same output voltage". What output voltage? – Kevin White Sep 01 '23 at 16:56

1 Answers1

2

Cf is normally pico-Farads. Assume leakage current through this massive MLCC cap is what is limiting the gain. Replace it with a C0G-class ceramic cap (even 100pF) and the gain will likely improve drastically.

A good analysis of the TIA is Digikey's Article about them.

rdtsc
  • 15,913
  • 4
  • 30
  • 67
  • Great point. Here are some measurements of MLCC, from them the series resistance of that huge cap could be as low as a few tens of kiloohms. https://www.silabs.com/documents/public/application-notes/an1415-current_leakage_through_ceramic_capacitors.pdf – user1850479 Sep 01 '23 at 17:14
  • Thank you so much I had no idea about this, I will try replacing it :) – OnHai Sep 02 '23 at 13:58