Sorry - you cannot acquire an EEG signal (or any biopotential signal) directly with NI data acquisition products. You must have an isolated differential amplifier input with high common-mode rejection to successfully (and safely) acquire signals from an electrode attached to an animal or human. The EEG signal is typically about 100 microvolts measured at the scalp, but is riding on top of 200 millivolts or more of DC generated by the electrode offset potential (like a half-cell battery), plus there is typically a lot of common-mode line frequency (50 or 60Hz) interference that can swamp the small differential EEG signal.
The easiest way to manage all of this is to purchase or borrow a commercial EEG amplifier (or more general biopotential amplifer) that has an analog output that is high level (+/-1V or more) and ready to connect to an NI data acquisition board for digitizing, processing, and storing on the PC.
As noted earlier, there are real safety concerns with attaching electrodes directly to people (or animals) and then plugging those electrodes into system that is then connected to wall power. Even battery voltage has the potential to create a nasty burn if there is a fault of some sort. Read up on these safety issues - this could be a good part of your project.
As for sampling rates, you must decide what you want to do with the data. For simple visual review, 200 samples/sec is sufficient to capture the waveform with good fidelity (the signal conditioners are usually setup for 0.1 to 70Hz bandpass, but most of the energy in the signal is below 30Hz). If you are hoping to do spectral analysis, a higher sampling rate is often required. If you are doing anything that requires tight synchronization between channels, then you may need a A/D board that has simultaneous sampling to prevent channel-to-channel sampling skew.
Hope this helps.
Steve
