1

I am building a hobby EEG and I am interested in quantifying how effective are the various noise reduction techniques.

The basic circuit would be two electrodes going to an instrumentation amplifier. Then the signal goes to various filtering stages and then to a final "mundane" amplifier. I am interested in noise reduction before the filtering. Here are the steps that I am considering (hopefully this can evolve into a community wiki).

  • Adding a right leg electrode directly linked to the ground of the circuit. (against 60Hz hum)
  • Adding a driven right leg electrode for common mode rejection. (against 60Hz hum)
  • Adding passive low pass filters in front of the inputs of the instrumentation amplifier (see here). (against DC difference between the electrodes and bias current from the inputs)
  • Faraday cage (against 60Hz hum)
  • Shielded cables for the electrodes (against 60Hz hum)

Are there any technical manuals discussing these approaches? Could you share your knowledge about the effectiveness of each approach? Could you suggest other approaches?

Scott Seidman
  • 29,274
  • 4
  • 44
  • 109
Krastanov
  • 207
  • 1
  • 9

1 Answers1

3

You want to look into all the documentation on the ModularEEG Design that is here. This is an open source/hardware project to build an EEG and has a wealth of information

You also want to take a look at this app note by TI that discusses dealing with noise in ECG.

Sidenote: Before you build anything taking bio-potential feedback please ensure that your circuit has sufficient isolation (via optocouplers, transformers or something else). Also try to design your circuit to run off two 9V batteries (+9V and -9V with current draw <100mA on the rails). These are good precautions when starting out to make sure you don't hurt yourself. Nerves are very sensitive and can be damaged by low voltage high frequency pulses that you won't even feel as a shock.

EasyOhm
  • 4,609
  • 26
  • 35