Below is a link to a .zip file (387K) that you can download. The file contains :
Click here to download the .zip file
NOTES :
In the board.jpg, the scale is centimeters. The top of the board contains the analog ground plane (top board in image). Unfortunately I placed the boards upside-down in the image; the inputs are on the right-hand side of the boards.
The values on the schematic are the original values as per the BYTE magazine article (I believe) . The latest current values are defined in the matlab file EEGfilterII.m
There are two gifs of the pcb layout, one shows the top side connections (red), the other superimposes the connections on the bottom side (blue lines). I fabed a double side board, and placed a ground plane on the top side.
In the schematic (eegamp21.sch, schematic21.gif) there are a few 'Fake' parts; this is because I didn't figure out how to represent inputs and output connectors in Protel, and I needed to represent the part for pcb layout purposes, so I represented the connectors as resistors. 'FIN1', 'FIN2', 'FPWR', 'FOUT12' really just represent 2 connectors each.
Also, I represented the 2.048V referecnce source (REG1,REG2, an ADR290 device) as a transistor because it's a 3 pin regulator (the base of the transistor represents the reference output voltage).
Finally, although the the opamps are listed as TL084, I experimented with some other opamps which were micropowered (TLC27L9CN). The 14 Hz noise is much greater when I use the micropower devices. (I think the gain-bandwidth of these devices is high enough that it should not be unstable, but maybe I'm wrong.)