0
\$\begingroup\$

I am designing an integrator circuit using op-amp. The circuit works very well for inputs greater than 7mVACRMS from 10Hz to 2KHz (error is inside 1%). The issue I am having now is when the AC input less than 5mV, the output the integrator has big error. About -15% error when the input is 1mVRMS sine wave. The following is the circuit I am using now. Does anyone have ideas about how to design an integrator for small input signal with high accuracy? Thanks in advanced!

The circuit is designed for time constant tau=0.0026. So at 61.45Hz, the output equals to the input value. The tau value in the following circuit is 5 times less than 0.0026, So at 61.45Hz, if the input is 1mVACRMS, the output should be 5mVACRMS. But I only measured 4.3mVACRMS. If the input is 30Hz, 1mVACRMS, the output should be 10mVACRMS, but only measured 8.6mVACRMS.

C30(0.022uf) is a 2% film cap and R16(23.7K) is 1% resistor. C33 coupling cap is 20% Tan or ceramic cap. The power for the amp is +-3.3V.

enter image description here

\$\endgroup\$
7
  • \$\begingroup\$ Calculate out your gains assuming 20% errors on your caps, and 5% errors on your resistors, both in directions conducive to your errors, and let us know how that works out. \$\endgroup\$
    – Scott Seidman
    Apr 13, 2017 at 15:18
  • \$\begingroup\$ Hi Scott, C30(0.022uf) is a 2% film cap and R16(23.7K) is 1% resistor. The circuit works well when the inputs are greater than 7mVAC. \$\endgroup\$
    – Cindy
    Apr 13, 2017 at 15:29
  • \$\begingroup\$ Is that the correct part number for the OP amp, MAX44248?? Because when i search for it shows up as a current sense amplifier with set gain... probably not the best choice for MFB filter \$\endgroup\$
    – Kvegaoro
    Apr 13, 2017 at 15:40
  • \$\begingroup\$ Doesn't R14 set the time constant of the integrator? \$\endgroup\$
    – Scott Seidman
    Apr 13, 2017 at 15:47
  • \$\begingroup\$ How about phase errors? Can you pop up a Bode? \$\endgroup\$
    – Scott Seidman
    Apr 13, 2017 at 15:48

1 Answer 1

Reset to default
1
\$\begingroup\$

About -15% error when the input is 1mVRMS sine wave

The 15% error could mean a DC error or an AC gain error. Start with a DC analysis. The input offset current for the Maxim chip is about 0.5 nA and this will produce an offset voltage error of 0.5 nA x 100k approximately. This is an error of 50 uV at the input.

You have DC gain of 22M/124k = 177 therefore, the offset voltage produced by the input offset current results in a dc output from the op-amp of about 9 mV.

Is this the 15% you refer to?

As for AC gain, you have a 22 nF capacitor in parallel with the 22 Mohm resistor and these could produce an error in gain of 10% at some frequencies. The 3dB point is 0.32 Hz so any AC signal having a frequency upwards of this could subject to the gain error you may be referring to.

Ditto the 4u7 and 23k7 input impedance. Are you using 10% capacitors by any chance?

So, you have DC error and AC gain errors to consider.

Does anyone have ideas about how to design an integrator for small input signal with high accuracy?

Given a decent enough spec, yes, plenty!

\$\endgroup\$
2
  • \$\begingroup\$ Hi Andy, The 15% error is for AC gain. We only care about the AC values from 5Hz to 1KHz and 1mVRMS to 200mVRMS. 30Hz should be far enough from 0.32Hz 3dB point. It still has 15% error. 4u7 coupling cap is 20%, is this would affect the ac gain? I have tried 22uF, but still the same result. I have Please advise me how to improve the circuit. Thanks for your help. \$\endgroup\$
    – Cindy
    Apr 13, 2017 at 15:12
  • \$\begingroup\$ The feedback capacitor needs to be a much tighter tolerance in my opinion. Try simulating it and altering valued. \$\endgroup\$
    – Andy aka
    Apr 13, 2017 at 16:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.