0
\$\begingroup\$

I'm giving an integrator a rectified and inverted sin wave. I assumed in my initial design that because you feed the input of an integrator into the negative terminal of the op amp, the signal will get inverted again. However my integrator gives practically no output, should I have given it a positive signal in order to work or does it make any difference?

The op amp used is the AD822

enter image description here

\$\endgroup\$
  • \$\begingroup\$ What are the input and output voltages? what are the power suplyy voltages? \$\endgroup\$ – jippie May 4 '14 at 8:05
2
\$\begingroup\$

You are correct about the polarity. It should work fine. Can you give some details about the signal you want to integrate? I call that circuit a leaky integrator (or low pass filter) due to R2, so check the time constant for 100nF and 220K (22mS? I'm from the ancient days of uF) versus the rate of change of your signal. This is more an IIR averager than an integrator. Also the gain, which is 1/RC or 1/10. Did you want to divide by 10?

Oh, also change + input to be 100K to ground for input bias balance.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ As the most important question: Single or dual power supply? Independent on supply: As mentioned above, you circuit will not be able to "integrate" because the phase will be 90 deg at one single frequency only (at the pole frequency). A real integrator needs a phas shift of app. 90 deg over a broader frequency range. \$\endgroup\$ – LvW May 4 '14 at 8:21
  • \$\begingroup\$ Im trying to integrate an EMG signal so between about 10-1000Hz at 1.3V, I think the time constant of 22mS is standard for EMG, however I did make a mistake with my diagram - R1 is actually 220K on order to get a gain of 1 in my design. Do I still need to place a 110K resistor at the positive terminal to bias it? The power supply is dual +-5V - I don't mind if the integration isn't perfect Im really looking for it to smooth output of the rapidly changing EMG signal for an MCU ADC. Sound good? \$\endgroup\$ – alto125 May 5 '14 at 7:44
  • \$\begingroup\$ If you are just trying to block higher frequency then the bias won't matter. To be precise, you would put the same resistance on the + and - inputs. This circuit falls in the family of Active Filters (they have feedback). How fast are you going to run the ADC? You may need a sharper low pass filter to avoid aliasing high frequencies into your band of interest. \$\endgroup\$ – C. Towne Springer May 6 '14 at 5:56
  • \$\begingroup\$ The ADC is going to sample at 14400 bits/sec to avoid aliasing. However this can be changed if it needs to be? \$\endgroup\$ – alto125 May 6 '14 at 11:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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