Naked Science Forum

On the Lighter Side => Science Experiments => Topic started by: R0derick on 10/10/2013 13:10:48

Title: EEG Costs
Post by: R0derick on 10/10/2013 13:10:48
I have a science exhibition coming up in my high school in December 2013. I'm guessing that controlling another creature with my mind is enough to sway the judges. I was planning on combining the (in?)famous 'RoboRoach' experiment with a brain-computer interface. Specifically, replacing the normal button remote with an EEG device. That last part is the problem. There's guides out there on how to create mind-controlled multi-function robots, but the one part they seem to gloss over is the EEG device itself. I can't buy a Neurosky or Emotiv device, if they were even available in my area. The cost of making a device such as the one suggested by the OpenEEG project, apparently ranges from 200 to 400 Euros  xx( I only need a device that can give 2 or preferably 3 output signals for 'Left', 'Right' and 'No Output'. Can someone explain how I can cut costs, which components are the most expensive and why? I'm far from an expert and any help is appreciated  ;D If what I'm suggesting is even possible in the first place and I'm not just blowing hot air :-\
Title: Re: EEG Costs
Post by: alancalverd on 10/10/2013 20:46:01
You need to be able to detect voltages of around 10 microvolts in the low audio frequency range. Not a big problem as you can buy differential amplifier chips that will do the job admirably, but you won't get much change from EUR100 for a 4-channel kit with a safe power supply. The practical problem is getting a decent, noise-free contact to the skin, and you will probably need to buy medical grade electrodes and connectors, which may be the most expensive part of the job.

On the other hand if you can stick gold needles into your scalp, you will get a lot more signal - and a lot more pain!
Title: Re: EEG Costs
Post by: RD on 10/10/2013 22:33:59
If you're sticking electrodes to your scalp my suggestion would be make the device battery powered to prevent the possibility of it going from DIY-EEG to DIY-ECT (http://en.wikipedia.org/wiki/Electroconvulsive_therapy).

Because of the very high gain involved, notch-filters are required to remove ubiquitous mains hum ...
http://www.instructables.com/id/DIY-EEG-and-ECG-Circuit/step4/Stage-2-60-Hz-Notch-Filter/  [50Hz in Europe ]
Title: Re: EEG Costs
Post by: CliffordK on 14/10/2013 00:16:27
I have a science exhibition coming up in my high school in December 2013.

You have a lot of design work and experimenting to do in a 2 month period.

I'd encourage choosing either making an EEG, or RoboRoach, for now, not both.  You could try combining technology at a later date.

There are commercial products available such as the emotive system (http://www.emotiv.com/), and you could still write your own software interface.

I only need a device that can give 2 or preferably 3 output signals for 'Left', 'Right' and 'No Output'.

Your headset will likely have at least a half a dozen outputs.  The output will essentially be the data to create a graph of time vs voltage, from which you will have design the computer to recognize certain features & differences in the EEG from which you will make your simplified commands, all in real time. 

I don't believe your brain output will look at all like say hitting a button with your finger.