Society of Robots - Robot Forum
Electronics => Electronics => Topic started by: Jdog on March 12, 2009, 09:41:36 PM
-
One of my goals over the next few years is to develop an electrode cap so I can measure tiny electrical signals coming off of my brain. This has been done before by many researchers and now I want to try. From what I understand from my readings is that basically all it is is a small electrode pressed against the head attached to an amplifier attached to an mcu with a computer interface. I am considering purchasing voltage amplifiers and starting my research but the only problem is that I have no idea what voltage range the signals coming off my brain are to fill in the equation in the voltage amplifier tutorial (Vout-Vin)/Vin = Rb/Ra . I then want to convert the Vout of the amplifier to a value that can be read by the adc on my axon. Does anyone have any ideas/suggestions.
-
Hi,
Don't want to be pessimistic on your behalf, but plain and simple - forget it.
We're talking µV's here, buried in volts of noise, mains hum etc. and there's a reason biometrics amplifiers are expensive as a certain hot place.
What you want to create is an EEG (Electro Encefalos Graph) amplifier and EEG voltages is much lower than eg. ECG (cardio=heart) and EMG (myo=muscles), which would be quite a challenge as well.
Besides the trouble making the amplifiers (and the filters needed), it won't do you any good without knowledge of the brain and the different patterns from the different placements of the electrodes.
You simply need to know what you get where to have even the tiniest chance of being able to interpret what signals you get.
Even the smartest scientists is a long way from controlling anything reliably with EEG.
Now if you were only to discriminate between two "states", had the right biomedical interface and trained yourself for some time, you could (perhaps) learn to create a signal called P300 (positive, 300 ms) reliably. This is fairly easy to discriminate (given the right equipment), but it's sort of a binary interpretation, either it's there or it's not, you don't get a better "resolution" than that.
Save a lot of time and even more grief, by using your hands a little while longer ;D
-
Years ago, Radio Shack used to sell a cheap skin-resistivity amplifier that was used for biofeedback. Like $25. It was fairly easy to train yourself to mentally control the output signal from this device, but the signal is slowly time-varying, and it would only give you one dimension of control. There are several other variables used for biofeedback, such as temperature, muscle potential at the skin, heart rate, etc.
A brief search shows there is a lot of equipment available for biofeedback, including for EEG, most of it in the thousands of $. Some of the sensors are in the $300 range.
-
Hack this
http://www.engadget.com/2009/01/09/video-mattel-mind-flex-hands-and-heads-on/ (http://www.engadget.com/2009/01/09/video-mattel-mind-flex-hands-and-heads-on/)
-
Hack this
http://www.engadget.com/2009/01/09/video-mattel-mind-flex-hands-and-heads-on/ (http://www.engadget.com/2009/01/09/video-mattel-mind-flex-hands-and-heads-on/)
Thats a possibility
EDIT:
I found this video of prototype this when they made a thought controlled car. Unfortunately I didn't see that entire episode and I can't find somewhere where I can. The headset they're using looks commercially made, I wonder where they got it. Has anyone seen this episode?
http://dsc.discovery.com/videos/prototype-this-mind-controlled-car.html (http://dsc.discovery.com/videos/prototype-this-mind-controlled-car.html)
-
Hi,
Sorry again, but the Mattel set is nothing but what we build in the late seventies for control of ultra simple computer games (on home brewed Personal Computers).
It's good for controlling one parameter only and it's slow reacting, however good you get.
The EMG mentioned by 4x4 would be a much easier* solution and with different electrodes on different muscle groups, a reasonably fast acting control can be executed by simply flexing and relaxing the relavant muscle groups.
* Depending on your level of understanding of electronics, both theoretically and practically, which I have no precise clue about.
-
I found the company that made the headset in my previous post but it's not consumer yet :-[ and starts at $300 :-[
[url=http://emotiv.com/]http://emotiv.com/ (http://emotiv.com/)[/url]
-
http://www.newegg.com/Product/Product.aspx?Item=N82E16826100006 (http://www.newegg.com/Product/Product.aspx?Item=N82E16826100006)
-
http://www.newegg.com/Product/Product.aspx?Item=N82E16826100006 (http://www.newegg.com/Product/Product.aspx?Item=N82E16826100006)
That looks nice but I read the reviews and it seems like a piece of crap. Also, if I didn't have to, I'd rather use one like the emotiv that registers muscle movement rather than emotions, because the former is much easier to control.
-
From what I understand, the devices mentioned above work a lot like this... http://www.instructables.com/id/Build-a-Lie-Detector (http://www.instructables.com/id/Build-a-Lie-Detector)!/
They measure the resistance of you skin to see if you are how your are feeling. If your nervous or stressed you sweat more and the resistance of your skin is lowered. I think with some practice and a pot to get the ideal resistance in the circuit, you might get very good at controlling the circuit. And, as a bonus the circuit would only cost a few bucks.
EDIT-From what I understand, the devices mentioned above work a lot like this
I looked at the last two those are a bit more complicate.
-
Hi,
If you really want to try and interface your brain waves you have two options:
1) Buy something expensive if you want it to look good.
2) If you are in it with the true experimenters spirit, buy a swim cap with a tight fit (made in as strong material as you can find) and mount snap-locks (the part with a curved metal plate on the inside) where you want electrodes. Then connect your wires to the other halves of the snap-locks and get some electrode paste to smear at the electrodes (you can get it from a fitness shop, it's used for "electro-fitness-machines").
The last option won't be something to impress the ladies with, but it's cheap.
For an extra good contact, shave your head :D
-
Hi,
If you really want to try and interface your brain waves you have two options:
1) Buy something expensive if you want it to look good.
2) If you are in it with the true experimenters spirit, buy a swim cap with a tight fit (made in as strong material as you can find) and mount snap-locks (the part with a curved metal plate on the inside) where you want electrodes. Then connect your wires to the other halves of the snap-locks and get some electrode paste to smear at the electrodes (you can get it from a fitness shop, it's used for "electro-fitness-machines").
The last option won't be something to impress the ladies with, but it's cheap.
For an extra good contact, shave your head :D
I figured that part out except for snap-locks, what are they. I was just planning on taking some small copper plates and sewing them to the cap. My problem is I need amplifiers strong enough that will take a microvolt signal and amplify it to something that an mcu can read, and I don't want it to cost a million dollars either. Also, I'm not sure about the placement around the head, but I would probably try and mimic the placement of the emotiv.
-
Hi,
I figured that part out except for snap-locks, what are they.
This is what I had in mind (discovered how many different things are called snap locks when finding this pic :-\).
(http://www.sz-wholesale.com/uploadFiles/UserImg%5Csmall_41600.jpg)
The part on the left goes towards your skin.
Placement depends on what you want to interface (brain stem, cortex, etc.).
This site: http://www.ebme.co.uk/arts/eegintro/index.htm (http://www.ebme.co.uk/arts/eegintro/index.htm) might be worth a read (I just skimmed it very lightly).
This book: http://www.moberg.com/files/products/EEG-placement.htm (http://www.moberg.com/files/products/EEG-placement.htm) might be word buying if you're serious about it (again, I just googled the link - didn't read the book).