Last Updated on September 13, 2024 by pg@petergamma.org
To build a multi-channel EEG device with as many EEG channels as required which is free from noise? With the current architecture of OpenBCI it is not possible to sample more than 16 channels at the same time. This, and also to use the ADS1299 with InfluxDB, requires that all the channels aer sampled at exactly the same time. PiEEG is a new architecture, and the ADS1299 should in principle be controllable from the Raspberry Pi. But the signals we have seen from PiEEG aren’t very convincing to this date, they look noisy:
We suppose the noise could be caused by the single board computer which is close to the PiEEG shiel. OpenBCI Cyton uses an Arduino compatible PIC32MX250F128B 32 bit microcontroller which has only 0.5 mA/MHz current consumption instead of a single board computer and is free from noise.
Aditya Mukherjee, Electronics Engineer from Bombay, India has built a mobile EEG system based on the ADS1299 EEG chip in combination with a Raspberry Pi RP2040:
https://www.linkedin.com/in/adityamukherjee11
The Raspberry Pi RP2040 consumes about 24 mA on average, without any code running. It has 40 GPIO pins, 20 on each site. The connector is different from the connector of PiEEG, so PiEEG cannot be connected directly to it. But the device Aditya Mukherjee has built is in principle a Raspberry Pi RP2040 with a PiEEG shield. Aditya Mukherjee will start selling his device once the firmware is tested. You can email him at aditya@mukherjee.tech with your requirements. It would be desirable that PiEEG takes this device over, sould it be free from noise.
With a new architecture which allows full control over the ADS1299 EEG chip, multiple ADS1299 EEG chips from different Raspberry Pi’s could in principle be sampled at exactly the same time, to aquire as many EEG channels as required at the same time:
https://forums.raspberrypi.com/viewtopic.php?p=2252319#p2252319
This is an architecture which does not require InfluxDB. To build devices based on InfluxDB is time consuming:
And this is an alternative from which we think it is worth studying. And also for instance to solder a Raspberry Pi RP2040 to a PiEEG shield and test it.
Comments are closed.