bci2000.org BBS

Hi

I am trying to acquire data using ModularEEG boards but the signals I appear to be getting do not look normal. I have attached a screenshot which shows a lot of square peaks within the signal. The screenshot was taken with the electrodes connected to a subject performing no movement.
My setup is as follows:
- 3 modularEEG amplifier boards (6 channels)
- 1 modularEEG digital board
- Since there is no serial port on my laptop I am communicating via a serial to USB cable which is set up as COM port 2

When I connect the same setup to Brainbay instead of BCI2000, the signals appear normal - I have attached another screenshot showing this.

If anyone has had similar problems I would greatly appreciate some help.

Leandra

Hi,

I am not familiar with the ModularEEG boards, but understand that one board only does 2 channels. I am not sure how the data from the 3 boards are integrated, but perhaps the protocol handles 3 boards differently than just one. In other words, it is possible that the current BCI2000 implementation of the ModularEEG support only supports one board. Thus, I suggest as a first step to just try one board, and then to see whether this will work.

Gerv

Thank you for the reply.

I tried putting a signal generator into just one channel and the signal still does not look right.

I then tried the same thing but with connecting the board to another computer directly into via its serial port (not via a serial to USB cable). The signal does seem a lot better as the sine wave I am putting in through the signal generator is clearer, but there are still some spikes - I'm not sure how serious these are.
So it seems that the serial to USB cable is definitely part of the problem since the signal gets better, while in Brainbay the signals are the same whether it is directly via serial or via USB. I'm just not sure if the problem is completely solved or if the spikes are still a concern?
I've attached a document with some screenshots.

Leandra

It seems like the BCI2000 "driver" for ModularEEG is not very robust. The best thing I can suggest without changing the code is to increase the priority of the ModularEEGSource.exe module to "High" using the Task Manager in Windows. Try this and see if the # spikes reduces.

Gerv

Hi Gerv

I agree, I think the problem is that the ModularEEG module is not very robust.
I have since changed to instead use a National Instruments DAQ and seem to be getting better results so will stick with that.

Thanks
Leandra