5
\$\begingroup\$

In the following configuration on this operation amplifier (OPA838), which is transimpedance configuration, can I connected the non-inverting input directly to GND, or should I use this filter? Not using this filter can cause instability? this can just improve the DC offset? in general why should I use this filter on the non-inverting input? what would be the benefit?

enter image description here

Reference: https://www.ti.com/lit/ds/symlink/opa838.pdf?ts=1713854959449&ref_url=https%253A%252F%252Fwww.ti.com%252Flit%252Fgpn%252F

Thanks

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Remember to consider acoustic coupling in capacitors; if 100 nF is not available as NP0 dielectric in the size you want, a 10 nF or 1 nF NP0 is probably better than 100 nF X7R. Otherwise mechanical noises can couple into the signal. \$\endgroup\$
    – jpa
    Apr 23 at 18:23
  • \$\begingroup\$ Hello, thank you for the nice comment, my application is ultrasound and there is possibly acoustic coupling in capacitors?! this topic "acoustic coupling in capacitors" is very new to me, would you please explain more on this? \$\endgroup\$
    – Andromeda
    Apr 24 at 11:11
  • 1
    \$\begingroup\$ "Capacitor microphonic effect" is a useful search term. Here are some resources: TI blog post, Question NP0 in loud environment? \$\endgroup\$
    – jpa
    Apr 24 at 12:28

2 Answers 2

10
\$\begingroup\$

in general why should I use this filter on the non-inverting input? what would be the benefit?

The OPA838 doesn't need the RC to work but, if you don't have the 100 kΩ, the input bias currents can produce an undesirable offset voltage on the output. Then, if you do use the 100 kΩ it becomes desirable to reduce its thermal noise by shunting it with the capacitor.

The typical input bias current for the OPA838 is 1.5 μA and, that would introduce a 150 mV offset at the output if the feedback and non-inverting resistors were not balanced. There will still be a slight offset due to the 20 nA difference between the two input currents (input offset current) but, that is surely much more acceptable.

\$\endgroup\$
1
\$\begingroup\$

This schema has disadvantage - gain depends on an input impedance of a sensor that is depends on temperature and lightind. In my opinion - it would be better to use 1.25 ref voltage bias like shunt regulator.A sensor connect between ref and positive input of op amp.Negative feedback resistor (divider must have two resistors and maybe RC compensation network for stability) also connects to ref. Ratio of resisrors depends gain and output voltage without input (when dark). In this case would be better use non inverting amplifier (gain is fixed because optosensor connects to high inpedance positive input)

\$\endgroup\$

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.