After testing out the Mikroe EEG board I’m going to look at creating my own band from the ground up. There were a few issues with what I was doing with the Mikroe Click but the main thing comes down to the fact that it should be powered from a 5V source rather than what I was doing from a 3.7V battery. I was surprised to see how well it performed with the INA114 being underpowered and then voltage dividing the analog output, it’s a bit messy.
I’ve designed an analog front-end built around portability that can easily be powered from a standard Lipo battery – it’s still single channel and doesn’t deviate too much from OpenEEG and standard approaches. Obviously this is all pending some testing with a prototype and also chip shortages might mean having to swap parts out but on paper it’s way better than what I was doing with the Mikroe board, which worked just fine. In parallel I’ll be working on software that can process the data on a mobile phone – both Android and Iphone. The main use case will be sleep analysis so I’ll be leveraging the work I’ve already done on a Windows app that does some real-time spectral analysis and uses a neural network to break down sleep stages. In the first instance, I’ll look at using this front-end with the Adafruit Feather M0 Adalogger again but eventually I’ll want to test this with a BLE board and then ultimately design a custom board for the MCU.
The other motivating factor for this was the fact that the PC app I’m making is quite convoluted to use with the Muse band I have; It means I need to use a separate app to take the Muse Bluetooth data and stream it to LSL (Lab Stream Layer) and then from there, my app can take the LSL feed and process EEG data in real-time...it just feels like a lot of steps and not ideal having to keep a PC running overnight. I’ll still finish this app because it’s a viable way to process Muse data in real time and potentially do things like trigger a smart alarm or REM/lucid dream alarm but I’ll be prioritising tests for a standalone band over this for the moment.
More soon.
This is really great man! I may be interested to try to make one at some point if I can't find a muse s gen 2 or dreem 2 for a somewhat reasonable price :) Would love to hear if there are any updates!
It's been really extraordinary reading about your adventures building a DIY PSG. Thanks so much for sharing your research with the world. I'm really looking forward to what a more streamlined variant will look like with a 5V power source.