My initial hardware design involved pulling the TGAM1 board out of the MindFlex Headset, placing it in a separate enclosure with a Bluetooth transmitter and using a separate power supply (with batteries and voltage regulator). While this was ended up being perfectly functional, I found there were other issues – mainly the reach for the EEG electrode. The lead wire is limited to a max length of 12” (according to the data-sheet for the TGAM1 board – the original lead wire was only about 3” long with a fixed electrode). With a max length of 12 inches, the TGAM1 board needs to be fairly close to the the head to allow the EEG electrode to reach most of the scalp.

The main limitations I see with the MindFlex headset are:

1. Electrode placement is fixed. (No choice on placement)
2. Comfort, EEG electrode contact uncomfortable after short period of time.
3. Poor conduction, EEG electrode is just metal to skin contact (simple dry electrode) – no electrolyte used to improve signal conduction and no adhesive used to ensure good contact.

Electrode Specs (per the TGAM1 Datasheet, page 3):

• Maximum surface area of ~150mm2 (but less surface area is optimal)
• Ag/AgCl, Stainless Steel, Gold, or/and Silver (both solid and plated material works)
• EEG electrode located above the left or right eye on the forehead
• Ground and reference electrodes located behind the ear or at the earlobe
• Have enough pressure to prevent movement, with a minimum of 0.8 PSI
• Length of less than 12 inches, the longer the higher the susceptibility to noise
• Shielding (not necessary for the ground)
• Thinner than AWG28

For the purposes of prototyping this project I decided to go back to using the MindFlex headset and fitting it with a different electrode and lead wire. I've finished fabricating a new electrode out of some thin stainless steel sheet metal, heat shrink tubing, 0.1” Dupont female crimp connector, and shielded 28AWG lead wire (just under 12 inches). I pulled out the original electrode, in the headstrap, and removed the lead connected to the TGAM1 board. Following that I soldered my new shielded cable to the board and then drilled a 1/8th inch hole in the MindFlex headset for the EEG lead wire to exit the headset.

My, newly fabricated, test electrode is a 10mm diameter disk (surface area of 78.5mm2 – well under the max of 150mm2). Testing of the new electrode went fine (good signal conduction). In addition, I just ordered some Ten20 electrode paste (conductive adhesive to attach electrode to scalp and provide better signal conduction).

I may attempt to fabricate some 10mm cup electrodes from some 416 stainless steel bar stock that I have (left over from another project).

I continue to inch my way along, teaching myself Java coding. Learning Java is my most challenging part of this project.