Thursday, February 6, 2014

mindfex headset disassemble hacked

Posted by rinson
The TGAM1 board that is inside the MindFlex can run in three modes. One of these is 9600 bits per second normal mode which is how the MindFlex is configured by default. This outputs a packet of data approximately every second that includes signal strength, attention level, meditation level, and pre-calculated values for 8 EEG power bands.
Friendly Warning and Disclaimer: Hack your toys at your own risk. I'm not responsible for any damage you do to yourself or your belongings. This information is provided here for information and education only. If you use the original hack and just use the data available in normal mode then the MindFlex still works as a game as originally intended. If you enable RAW EEG output then you break the toy!
To enable RAW EEG you need to remove a 10K resistor from the TGAM1 board, and solder an extra connection to the board, as shown in this image:
Connection on the TGAM1
Remove the 10K resistor that appears where the red X is on the photo. You should double check with a multi-meter that you're removing the correct part. It is connected to the B1 port. This B1 port forms part of the configuration of the TGAM1 board. In the MindFlex this 10k resistor connects the B1 port to GND which configures it in normal mode at 9600 bps.
Solder a connection on to the B1 port, then you can add a new 10K resistor that connects the B1 port to VCC. This enables raw output mode at 57600 bps.
You connect to the Arduino in the same way, but you can't use the Arduino Brain Library with this configuration. I'm working on an update for this.
The TGAM1 chip still sends the summary data packets that you get in normal mode once per second, but now you will also get a packet containing the raw EEG reading 512 times per second, approximately once every 2ms.
Read More
Sharing EEG Mind Flex Arduino Hack With Programs

hellow my name is rinson n raj am from india i have done this project of controlling a chain drive robot with mind using mind flex dual and arduino i can give u all codes with a mind flex head set with tgam inside it please be free to mail me for details rinsonnraj@gmail.com

Wednesday, February 5, 2014

Dismantling Mindflex Duel Game FOR EEG ARDUINO HACK

Posted by rinson


Hi all,
Here in this video i have shown how to dismantle the Mindflex Duel Game FOR EEG ARDUINO Coding.
As you all know the casing of Mindflex Duel Game is very difficult to open. So this video will help you  to dismantle and also it will help you to know the parts inside this thing.
Read More

Monday, February 3, 2014

sharing eeg mind flex arduino hack with programs

Posted by rinson
hellow my name is rinson n raj   am from india i have done this project of controlling a chain drive robot with mind using mind flex dual and arduino i can give u all codes with a mind flex head set with tgam inside it please be free to mail me for details
rinsonnraj@gmail.com 
Read More

eeg based robot control using arduino and mind flex hack

Posted by rinson
We have done eeg based robot control using arduino and mind flex. For more help on this you can contact me directly. I will clarify all your doubts. Also you can comment below for further details on this project.


















Read More

Visualising mind activity with a hacked toy EEG

Posted by rinson

Visualising mind activity with a hacked toy EEG


The MindFlex toy includes a headset that reads EEG wave data. There's lots of information on hacking this toy to create a simple EEG machine using the limited data the headset provides by default. In this project I have extracted the full RAW EEG data using the same toy headset from the MindFlex game.
Update: I've posted an update to this blog, where I provide details of how to extract the full RAW EEG data from the MindFlex headset.
Friendly Warning and Disclaimer: Hack your toys at your own risk. I'm not responsible for any damage you do to yourself or your belongings. This information is provided here for information and education only. If you use the original hack and just use the data available in normal mode then the MindFlex still works as a game as originally intended. If you enable RAW EEG output then you break the toy!

Reading brain waves

An obvious place for us to start on this project might have been with the excellent OpenEEG project, but while being open, it still has substantial costs and complexity in getting up and running. For this project we looked for a simpler approach.
EEG technology has been incorporated into various toys and household items before. There are various commercial options available based on NeuroSky's ThinkGear chip.
The original Frontier Nerds blog post from 2010 provides a great starting point, but I've made a few changes along the way, including successfully extracting the RAW EEG values from the MindFlex toy.

Budget EEG with the TGAM1

The TGAM1 is the controller board responsible for processing the EEG signals in all of the products based on the NeuroSky ThinkGear chip. NeuroSky no longer sell the boards in small quantities (I've had confirmation of this from their business development team that this is due to support resources) and instead recommend that you buy the BrainWave starter kit. This is probably a good idea if you want easy access to EEG data, but they are still over 100 GBP to buy in the UK, and the toys are available for only 40.
Read More

MindFlex headset Arduino interfacing

Posted by rinson

Simple EEG reading

The basic data that comes out of the MindFlex headset by default actually gives us quite a lot of information for only £40. In the default configuration you have soldered on only two connections, one for a common ground and one for the serial data stream. I connect this to a SoftwareSerial pin on the Arduino and read in the packets. Here's what data you get:
  • Signal strength (0 - 200) where 0 means good signal, and 200 indicates a problem with the connection.
  • Attention and Meditation eSense values. These are a proprietary measure that is calculated by the TGAM1 using NeuroSky's own algorithm. The toys based on this technology use the attention measure.
  • EEG power band values: A measure of the various power bands that has been pre-calculated by the chip, this updates once per second.
The Arduino Brain Library already includes functionality for working with these values.

TGAM1 Region Change

If you bought the EU version and you're in EU then skip this bit.
The MindFlex, MindSet, Necomimi, etc all based on the TGAM1 that includes a notch filter on the EEG wave to remove electrical noise from the data. If you bought a MindFlex from a different region then you will need to adapt it to work for you. For example, if you've bought a headset from the US it will have a notch filter at 60Hz. In Europe our "mains hum" is at 50Hz. You can switch the notch filter to work at the correct frequency with a bit of soldering.

Getting full RAW EEG data

Now on to getting the full raw EEG data from the MindFlex headset. In addition to the two connections you soldered on before there are two extra connections required. This allows you to configure the TGAM1 board that is in the MindFlex to switch between two modes. The 'configure' connection is actually called B1 on the TGAM1 board. In the MindFlex headset B1 is connected to GND via a 10K resistor. I have marked this resistor with an X on the image.
You need to remove this 10K resistor from the circuit board, and connect B2 to VCC via a 10K resistor to enable full raw EEG data output.
The full EEG data is output at the higher baud rate of 57600. You also get the same packets as in normal mode.
You can switch back to normal mode, with just the basic data output at baud rate of 9600 by connecting B1 back to GND with the 10K resistor instead of the VCC connection.

Signal strength

The normal data and raw data modes both give you the summary packet once every second. In this is a measure of signal strength. It should be zero to indicate that the board is getting good measurements. Anything more than 0 indicates poor signal, and 200 is used to indicate a problem.
When I had the Arduino hooked up to the computer serial port I struggled to get the signal strength to 0. When running off batteries (I have an Arduino Pro mini connected to the headset's battery pack) I have had no issues with signal strength.
The issues were resolved by adding an extra GND connection from the Arduino and connecting to the body of the person being tested. I don't recommend this approach and running the headset completely off batteries is recommended.
This requires no wired connection to the computer, hence the RF link that has been implemented.

RF data link

I am using a cheap low-power RF transceiver to send the EEG packets to the computer. This makes the headset wireless and solves the signal strength problem that required the extra ground connection.

Receiving the data

The Processing Brain Grapher is a good starting point to check everything is working but it only works when the headset is in normal mode and sending the data via the serial port.
I am working on a Processing library that will work with the updated headset in full EEG mode, this giving access to the eSense meters, the EEG power band values, and the full EEG raw data. I'll be expanding this section as I make more progress with the libraries and visualisations.
Read More

Full RAW EEG data from the MindFlex headset hack

Posted by rinson
There have been many hacks posted online over the past few years that show how to build a simple EEG machine using the Mattel MindFlex headset.
It's simple to interface this with the Arduino to read the basic single signal EEG data the toy produces. The toy actually only uses one of the measures (attention) but it outputs various useful bits of information.
All of this is detailed in the original Frontier Nerds blog post and on our Brain Wave Visualisation project page.
The original hacks are all based around the preprocessed data that the TGAM1 board in the MindFlex toy produces. I have successfully modified one of these headsets to get full raw EEG data from it. I'm in the process of writing a library and updating the project page with full details, but in the meantime here's how it works.

Full raw EEG values

The TGAM1 board that is inside the MindFlex can run in three modes. One of these is 9600 bits per second normal mode which is how the MindFlex is configured by default. This outputs a packet of data approximately every second that includes signal strength, attention level, meditation level, and pre-calculated values for 8 EEG power bands.
Friendly Warning and Disclaimer: Hack your toys at your own risk. I'm not responsible for any damage you do to yourself or your belongings. This information is provided here for information and education only. If you use the original hack and just use the data available in normal mode then the MindFlex still works as a game as originally intended. If you enable RAW EEG output then you break the toy!
To enable RAW EEG you need to remove a 10K resistor from the TGAM1 board, and solder an extra connection to the board, as shown in this image:
Connection on the TGAM1
Remove the 10K resistor that appears where the red X is on the photo. You should double check with a multi-meter that you're removing the correct part. It is connected to the B1 port. This B1 port forms part of the configuration of the TGAM1 board. In the MindFlex this 10k resistor connects the B1 port to GND which configures it in normal mode at 9600 bps.
Solder a connection on to the B1 port, then you can add a new 10K resistor that connects the B1 port to VCC. This enables raw output mode at 57600 bps.
You connect to the Arduino in the same way, but you can't use the Arduino Brain Library with this configuration. I'm working on an update for this.
The TGAM1 chip still sends the summary data packets that you get in normal mode once per second, but now you will also get a packet containing the raw EEG reading 512 times per second, approximately once every 2ms.
Read More