I followed this circuit to get the EMG signals (dual supply +-9 V). I set the Rg at 330 Ω (at 2.8 kΩ, I can't see any changes in the output, so i set the gain higher), and replaced OPA2131 with TL072 instead. After Vo is a highpass filter at 20 Hz and lowpass filter at 500 Hz.

ECG amplifier with right-leg drive circuit

Image source: Texas Instruments (Burr-Brown) INA128 datasheet Rev B. SBOS051B February 2005

The problem that I can't figure out is when RA, LA, RL electrodes are connected to my body, while sitting on a chair (legs not touching the floor). The output is at a certain level of voltage, and then when I put one or both feet on the ground, the output voltage changes, most of the time the output is saturated. Is this common mode voltage interference? Or changes in reference voltage? How do I fix it?

I also tried out this EMG circuit: https://www.charleslabs.fr/en/project-OpenEMG+Arduino+Sensor The behaviour is the same when my leg touch the floor

  • \$\begingroup\$ MonhVuv Vu - Hi, Please note the site rule which requires that when a post includes content (e.g. text, image, photo etc.) copied or adapted from elsewhere, that content must be correctly referenced. As a minimum, for online material the source webpage or PDF etc. should be named & linked (see that rule regarding references for books / articles etc.). In order to help you, I found what I believe to be the source PDF link for the copied image & added it. For the future, please remember it's your responsibility to do that :) Thanks. (Please see the tour & help center.) \$\endgroup\$
    – SamGibson
    Commented Oct 15, 2023 at 12:21
  • \$\begingroup\$ Oh thank you! First time here, i will do reference next time. \$\endgroup\$
    – MonhVuv Vu
    Commented Oct 15, 2023 at 12:24
  • \$\begingroup\$ It seems to me that the point of the right leg drive is to drain off the charge from the body to reduce the common mode voltage to zero. But it may very well be that the circuit is simply designed to work on a person who is not electrically connected to anything. In clinical usage these are used on patients sitting on (electrically insulated) exam tables. It may simply not be possible to get the signal you want from a person who is touching the floor with bare feet. \$\endgroup\$ Commented Oct 15, 2023 at 21:55
  • \$\begingroup\$ Do you have any materials to tackle this? Even with shoes on, and start walking, the output fluctuates a lot. There are some EMG devices out there that measure EMG signals of the legs while walking and still spit out reliable output, no idea how they do it. \$\endgroup\$
    – MonhVuv Vu
    Commented Oct 16, 2023 at 5:34

1 Answer 1


Short story: keep the gain low on the INA128.

Longer story: -

The INA128 Instrumentation amplifier should not have a gain that is higher than about ten (as per your diagram). The reason is that the common mode voltages are massive compared to the signal you want to measure and, if the amplification is too high you will get saturation and, then it's game over.

What you need after the INA128 is another amplifier with a gain of maybe 100. In other words, let the INA128 deal with blocking the common mode voltages (lowish gain necessary) and, once that feat is accomplished, a new amplifier after the INA128 will amplify just the wanted signal to a level you can see on the oscilloscope.

  • \$\begingroup\$ Thank you so much! I wil try that configuration now. Can you also check out this one, which also create output voltage saturation when my leg touch the floor (charleslabs.fr/projects/20190618_OpenEMG/circuit.png). This is done on the cheap lm324, which i guess at the differential amplifier stage doesn't do much common mode rejection. Therefore, at the tuneable amplifier it also amplified the noise when my leg touch the floor too? \$\endgroup\$
    – MonhVuv Vu
    Commented Oct 15, 2023 at 12:17
  • \$\begingroup\$ The input impedances are far too low for that circuit to be successful. Stick with an InAmp. \$\endgroup\$
    – Andy aka
    Commented Oct 15, 2023 at 12:26
  • \$\begingroup\$ Yeah, that's what I suspected to be the cause. Fixing it now! \$\endgroup\$
    – MonhVuv Vu
    Commented Oct 15, 2023 at 12:36

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.