EEG not detecting

Connection, ECG, EMG, EDA, etc.
MrFloppy
Posts: 10
Joined: Tue Oct 11, 2016 12:59 pm

EEG not detecting

Postby MrFloppy » Thu Jan 05, 2017 6:16 pm

Hey there,

I am building a device to detect EEG signals using the Bitalino EEG 151015. I use a STMicroelectronics Nucleo L432KC as a microcontroller. Now I added the EEG Sensor and I get really strange values (see Graph attached to this topic).

When I provide a constant voltage of 3.3V to the same Input I get the Graph that I expected (continous signal at the maximal sample value) so I can exclude the ADC as a source of error. I am using reusable Electrodes (made by Bitalino) - new ones that were never used before.
I even exchanged the electrode lead cable to exclude that one.

The sensor is powered with 3.3V (from the nucleo) and ground is connected to the nucleo ground as well.
I provided VCC/2 with a voltage divider (by using 2 100kOhm resistors between ground-vcc/2 and 3.3V-vcc/2).

Can you please help me on this?
Thank you very much!
Attachments
EEGNotWorkingGraph.png
Strange Values
EEGNotWorkingGraph.png (12.06 KiB) Viewed 1749 times

BITalino
Site Admin
Posts: 567
Joined: Tue Aug 27, 2013 3:47 pm

Re: EEG not detecting

Postby BITalino » Thu Jan 05, 2017 6:40 pm

Hi there,

How to you have the electrode leads connected? It would be really helpful if you could share a few pictures of your setup.

Perhaps there are noise sources interfering with the signals... please refer to the following thread on our forum for additional information:
http://forum.bitalino.com/viewtopic.php?t=236#p614

Connection-wise things seem good, with 0V on GND pin, 3.3V on VCC and VCC/2 on REF (Ax being the sensor output).

Best regards,
The BITalino Team

MrFloppy
Posts: 10
Joined: Tue Oct 11, 2016 12:59 pm

Re: EEG not detecting

Postby MrFloppy » Thu Jan 05, 2017 8:02 pm

Thank you very much for replying.

Taking pictures is difficult as I am trying around with different positions. For this purpose I am using reusable dry electrodes and just place them to the spots (reference to the bone behind the ear, + and - at the forehead 1 inch above and right (from my point of view) to the nasion) and push them down with my fingers.

I am wondering why the signal is so shaky - it gives me 0 or almost 100% jumping between those two all the time.

I know about that noise problem and I already set up a device with EDA/ECG using Bitalino sensors with very clear signals using the same setup. The problem should be more about the EEG sensor/electrodes/positions/etc..

BITalino
Site Admin
Posts: 567
Joined: Tue Aug 27, 2013 3:47 pm

Re: EEG not detecting

Postby BITalino » Thu Jan 05, 2017 8:13 pm

Are you using gel on the electrodes? Perhaps that's the culprit... the typical signals should be like this:
https://youtu.be/LOFUTNEgrv4?t=226

From your screenshot the signal trace looks like there is an electrode not connected or something alike.

If you're interested, we can also schedule a Skype / TeamViewer session to further look into the issue.

To move forward with the later, please contact us by e-mail to bitalino@plux.info

Best regards,
The BITalino Team

MrFloppy
Posts: 10
Joined: Tue Oct 11, 2016 12:59 pm

Re: EEG not detecting

Postby MrFloppy » Sat Mar 11, 2017 12:49 pm

Dear Bitalino team,

I spent a while testing different setups and electrodes.
The following is related to tests that I performed with pre gelled sticky electrodes (TIGA-MED Gold).
I attached the Electrodes as supposed in the video you have posted.
I did many attempts and gathered data over a period of 2-10 minutes.

In the attachment you can find some datasets that show how my data looks like.
It seems like I am not detecting EEG.
Can you please help me on this?

If I perform an FFT Analysis it looks like the brainwaves from 0-4 Hz are the most active ones in every situation which should not be the case.
I calculated the single sided FFT according to this Matlab page: https://de.mathworks.com/help/matlab/ref/fft.html
I attached the FFT comparisons between the three posted signals.

What am I doing wrong?

Edit: The data was detected with 200 Hz sample rate.
Attachments
eegFftRestComparedToCreativeTask.png
eegFftRestComparedToCreativeTask.png (17.59 KiB) Viewed 1346 times
eegFftRestComparedToWorking.png
eegFftRestComparedToWorking.png (17.61 KiB) Viewed 1346 times
eegFftWorkingComparedToCreativeTask.png
eegFftWorkingComparedToCreativeTask.png (18.15 KiB) Viewed 1346 times
eegWorkingOnALaptop.png
eegWorkingOnALaptop.png (8.68 KiB) Viewed 1346 times
eegCreativeTask.png
eegCreativeTask.png (12.87 KiB) Viewed 1346 times
EEGRest.png
EEGRest.png (15.47 KiB) Viewed 1346 times

BITalino
Site Admin
Posts: 567
Joined: Tue Aug 27, 2013 3:47 pm

Re: EEG not detecting

Postby BITalino » Sat Mar 11, 2017 1:23 pm

Hi there,

From the time series plots you've shared, it seems like the sensor is still saturating. It would be really helpful if you could share a photo of your setup or a schematics showing how you have things wired up.

Do you have the chance to power the sensor from VCC and REF voltage regulators in an independent way of the power lines that are connected to the MCU / transmitter?

We can eventually ship you a replacement sensor, but all parts have automated tests prior to shipping. The normal signal trace that you should be obtaining can be further confirmed on the data sheet available here:
http://bitalino.com/datasheets/EEG_Sensor_Datasheet.pdf

Best regards,
The BITalino Team

MrFloppy
Posts: 10
Joined: Tue Oct 11, 2016 12:59 pm

Re: EEG not detecting

Postby MrFloppy » Sat Mar 11, 2017 1:53 pm

Thanks for you answer so far.

Please see the wiring in the attachment. You can see a battery pack which is not connected, because I did not get a voltage regulator yet.
The MCU is powered via USB by a Macbook which is not plugged to any power source (it runs from the battery). The sensor is powered by the 3.3V output of the MCU.
Attachments
IMG_3677.JPG
IMG_3677.JPG (178.01 KiB) Viewed 1340 times

BITalino
Site Admin
Posts: 567
Joined: Tue Aug 27, 2013 3:47 pm

Re: EEG not detecting

Postby BITalino » Sat Mar 11, 2017 2:15 pm

Thanks for sharing the photo... there are a few things that may be able to help you achieve better results. The electrode cable assembly for example appears to be a bit flimsy.

Perhaps we can send you a replacement sensor (just to account for possible damages during shipping). It should also be possible to send you a more flexible cable to interface with the breadboard as well:
https://store.plux.info/handy-tools/234-professional-grade-wire-for-the-bitalino-freestyle-100-cm.html

or:
https://store.plux.info/handy-tools/237-miniaturized-high-performance-bitalino-revolution-cable-100-cm-810122919.html

You could also consider the purchase of a SnapBIT-DUO + ground lead:
https://store.plux.info/breakout-boards/219-snapbit-duo.html
https://store.plux.info/cables/228-1-lead-electrode-cable.html#/57-cable_length-100_cm/25-connectors-none

The SnapBIT-DUO generally helps; eventually a battery-powered supply block could be used to power both the sensor and the ST MCU... our PWR block has 3 voltage regulators for that (1x AVCC for sensor; 1x DVCC for MCU + BT; 1x REF for midpoint voltage):
https://store.plux.info/bitalino-components/21-power-pwr-block-810121702.html
https://store.plux.info/29-batteries

See attached photo of an alternative assembly (on the sensor side at least). If you're interested in learning more about how this could be put forth, please e-mail us at bitalino@plux.info

Best regards,
The BITalino Team

IMG_9765.JPG
EEG + SnapBIT-DUO assembly
IMG_9765.JPG (221.19 KiB) Viewed 1337 times


Return to “Sensors”




cron