This is the EOG (Electroencephalography circuit I'm working with) enter image description here

The output is essentially noise when I hook up the electrodes to myself, hence I'm debugging each stage, hooking up to an oscilloscope. Coming to the ina333, I expect to see some noisy signal with the electrodes hooked up, what I don't understand is why the oscilloscope output looks the same with and without the electrodes attached. Electrodes are on sides of the eyes on the temple side and reference at the mastoid.

oscilloscope output:

sorry about the glare, at the top is the signal and below that is the fft, a table shows the vrms and corresponding frequencies. I don't even understand where I'm getting MHz frequencies from in the first place. That's besides the point, this is the exact same screen that I look at with and without EOG_L, EOG R and REF attached to the circuit. The reference pin is connected to the supply ground, i'm pretty sure that doesn't have any affect to this issue. But I've seen people use virtual grounds instead.

Any idea where I might be going wrong? the components are hand soldered on a perf board, it's pretty messy ngl, but I've checked all the confections numerous times. enter image description here

Based on a comment's suggestion, I added in 0.1uF decoupling capacitor between the positive rail input of opamps and ground. Not much change tbh. In addition to using the oscilloscope, I used a voltmeter to just center on the voltages and observe how they are being manipulated through the stages.

I'm using the INA333 amplifier as my first gain stage that takes two inputs from horizontal EOG electrodes placed on the sides of the eye close to the temple. I take the common mode at the middle of two matched gain resistors, invert it and biased against the stable 1.65V reference to feed it into the body through the reference electrode. I expect to see an amplified output by a factor of 20. But none of that happens.

With the voltmeter the inputs were around .7-.8mv and the output was between .7-.8mV. NO Change to the outputs. While common mode was around .175 mv. I also thought the voltage at the reference would be around 1.65 as it's biased to that voltage but it's ~ .8mV just like the inputs.

enter image description here

As per suggestions, I replaced the schematics with familiar opamp symbols to make it easier and understood the change in REF pin of INA333, made a stupid mistake there not centering it to the Reference voltage 1.65V.

Things I have learnt:

  • Need to add low-pass filters at the INA333 input to filter all the extrinsic high-frequency noise. I understand that the differential capacitors of the low-pass filter will also help minimize effects of the high-source impedance
  • The INA333 REF will need half supply to properly amplify the signal

Things that I still don't quite understand from the comments:

  • Why body should be connected to the mid point bias (1.65V) than the common mode? It's totally different to what I've read. I'd really like to understand the logic behind that.
  • What are things to consider to the input to IC5..it seems that is quite important, can't quite understand.
  • Does the 1.65V bias to INA333 and to other opamps have to be two different supplies?
  • \$\begingroup\$ Use "decoupling" capacitor (0.1uF -- 1uF) nearest power supply pins of each opamp ... \$\endgroup\$
    – Antonio51
    Jul 7, 2023 at 7:40
  • \$\begingroup\$ is that really all it might be? lemme try it out and edit my question. \$\endgroup\$
    – roaibrain
    Jul 7, 2023 at 12:54
  • \$\begingroup\$ @Antonio51 I edited the question with new observations, do you have any input? \$\endgroup\$
    – roaibrain
    Jul 7, 2023 at 22:11
  • \$\begingroup\$ I don't know if it's related to your issues, but the output voltage of the inamp will have considerable effect on the reference voltage fed to the REF driver and output stage. I suggest adding a second RC filter just for the output filter on the inamp. \$\endgroup\$ Jul 7, 2023 at 22:44
  • \$\begingroup\$ How are you DC biasing the inputs of the INA333? Are you connecting your body to circuit common and using that for DC bias? The mid-point bias needs lots of additional capacitance or rethinking how you bias IC5 since you're feeding the output of the INA333 to the other filter stages through the mid-point bias wire. Your schematic would be a lot easier to read if you used a standard symbol for an opamp. \$\endgroup\$
    – qrk
    Jul 7, 2023 at 22:51

1 Answer 1


As pointed by @qrk ...

V+ (inputs of OP335 IC1-IC4-IC5) should be 1.65 V (?).

Note that ic6 is not "well" wired (not feedback ?)

Here is what should be done (1.65 V for Ref Input).

enter image description here

  • \$\begingroup\$ V+ is the supply, In+ are the inputs which are biased to 1.65. As you can see both the band pass and amplifier stage do that. IC6 is just a follower, how do you suggest it be wired? \$\endgroup\$
    – roaibrain
    Jul 8, 2023 at 21:06
  • \$\begingroup\$ As this one (with dual supplies) researchgate.net/profile/Hyem-Saadi/publication/252020464/… \$\endgroup\$
    – Antonio51
    Jul 8, 2023 at 21:34

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.