Audio input to Cyton's D11 from Mac's aux out ... interfering with EEG
Hi, I have an openBCI Cyton with Daisy, and I want to run audio event-related potential experiments.
I have a mac which runs experiments using PsychoPy, and outputs audio triggers through the 3.5mm AUX output. I have crocodile clips attached to the tip, and sleeve of the aux output, to a full-wave rectifier, with 2 capacitors attached, which subsequently goes into the cyton's D11 analogRead.
I can see that audio triggers appear, under analogRead D11, but then if I look at the FFT graph I can see that there's electrical leakage into the EEG channels (very minute, on the order a few micro Voltage) However, if I'm going to aggregate over the signals, this might interfere with any results I might find. Does anyone have any idea how can I reduce the electrical interference of the D11 audio input from the EEG channel signals?
Comments
The link attached above is to a video showing the wave forms I'm getting
It's possible this is happening because the electrodes are not connected to a user.
Yunda, hi.
It's possible you may have some issues related to the 'direct' connection of your audio system to the Cyton, with no 'isolation'. There are small isolators like this that may help. The idea is to use a transformer to decouple the output and input systems.
https://www.amazon.com/dp/B06XQYN77L/
Also, why are you pre-rectifying the audio? Square waves have many high frequency components because of the sharp on and off transitions. The square wave pulse on the wires to the Cyton, may be coupling into the very sensitive (microvolt level) EEG channel inputs. Just send the plain audio sine wave from your PsychoPy to the D11.
Regards, William
Instead of the pre-rectification, you may instead need a simple two resistor voltage divider, to get the audio output level down to what the PIC32 onboard ADC can handle.
https://www.google.com/search?q=voltage+divider
But the ADC should be capable of 3.3V max levels.
Since you still have the coupling potential of incoming audio, millions of times 'louder' than the microvolt EEG cables -- you may want to get the audio levels down as low as possible. Even possible to use one of the EEG channels if you get it into say the 50 microvolt range.
Using the PIC32 D11, I believe the PIC32 ADC is 10 bit, 1024 values. 3.3 V / 1024 = ~ 3 millivolts per one count ADC increment, finest possible resolution.