EEG amplitude is way too high.
Hi
So my Ganglion says that my eeg amplitude is up to 3,299 uv
with an average of -977. And then ran again, and maximum
of 3151 and average of -345. Obviously this is wrong.
For now I am just using one channel, so one electrode, which is this type:
https://shop.openbci.com/products/openbci-eeg-headband-kit.
How can I fix this?
Thanks
Katie888
Comments
Where are your electrodes placed? What are you using for reference and location? Ground?
If this Ganglion has worked correctly for you in the past, then suggest issues with the electrodes, skin contact, cables, etc.
A simple 'sanity' test you can always do is to connect the three electrodes together (channel, reference, ground). This should result in close to zero microvolts.
The dry electrodes you mention are HIGHLY dependent on the pressure applied to the skin and skin impedance. Have you checked impedance with the GUI widget?
https://docs.openbci.com/Software/OpenBCISoftware/GUIWidgets/#ganglion-signal-check
William
Hi, thanks for getting back to me so quickly!
So per your advice I put the channel electrode, reference, ground together and checked impedance and it is very high.
So to be on the safe side I powered up the battery, but the results (see below) were basically the same.
Hi, thanks for getting back to me so quickly!
So per your advice I put the channel 0 electrode, reference, ground together and checked impedance and it is very high.
So to be on the safe side I powered up the battery, but the results (see below) were basically the same.
Channel[0] 386.5
Channel[1] 479.5
Channel[2] 479.5
Channel[3] 386.5
Sorry about the multiple posts, I was trying to add 2 photos but finally see that not allowed for this area.
I guess I am not understanding what is going on. With the channel, reference and ground shorted, the time series should show close to zero microvolts and the impedance should also be very low.
I suggest you email Customer Service at (contact at openbci.com) to further investigate. Your wires or electrodes may be damaged.
I asked previously if this Ganglion has ever worked correctly?
Hi wjcroft
Not sure if they were shorted. I attached the ground and reference to the OpenBCI EEG Headband right next to the one channel 0 electrode. Is that the correct way to short?
And it turns out that the Ganglion has had this problem all along (I have files from when I received the Ganglion), but I didn't realize that there was a problem. Once I got the Ganglion I looked for inexpensive software that could import a bdf or txt file to graph and analyze the results but did not have any luck. So I took a class in Python and wrote a program to graph the results. And the bar graph of bands showed either mostly Delta or mostly Gamma, both of which makes no sense, and realized it was due to the numbers being out of whack.
Anyway I really appreciate your help in getting to the bottom of this and I will email support.
By 'shorting' or 'connect together', it is meant to physically contact all three electrodes. If you have some kind of clamp or alligator-clip this can help to that. If you have Ten20 paste, you could stick all three into a blob of that.
re: importing csv or bdf. I might have mentioned in the past that EDFBrowser can do this. And other signal processing graphing apps can work as well, such as EEGLAB.
Hi wjcroft
You are right. I exchanged the dry electrodes for some wet electrodes (Channel 0, Reference and Ground),
spooned and taped them together and then ran the impedance with this result:
Channel[0] 000.5
Reference Impedance 0.05
And then I took the same electrodes (Channel 0, Reference and Ground)
and taped them to my forehead with Ten20 with pretty good impedance:
Channel[0] 061.5
Reference Impedence 95.5
So that is encouraging.
But the exported txt file is off. The maximum votage was 623 and
minimum was -1885 with an average of -539.
Any suggestions?
Katie888
P.S. and regarding EDFBrowser and EEGLAB, I did check them out, but
they don't have the graphs that I am looking for, so glad to use python instead.
Plus they really are designed for those with lots of expertise (definitely not me).
Note that transient artifacts will distort your maximum and minimums. When you use the GUI Time Series, what do you see in the text box 'rms Uv', which is the average microvolts. Should be below 100 uV most of the time and even lower in quiet periods.
William
Hi wjcroft
So I just ran the Ganglion, this time with more Ten20 paste, and the impedance was only:
Channel[0] 15.00 and Reference Impedence 18.00 and the 'rms Uv' was only 47.9
But out of 5 seconds the voltage peaked at over 200 uv each time.
Anyway I emailed a screenshot of the GUI to your support email,
can you please have a look and let me know what you think?
Thanks!
Katie888
I am only the Forum admin here. Your Customer Service query will be answered by another staff member.
Have you tried the basic Alpha test?
https://docs.openbci.com/GettingStarted/Biosensing-Setups/EEGSetup/#4-alpha-brain-waves-eeg
To conclude this thread, Katie found the issue (1 Hz noise bursts) was caused by electrical noise from a nearby doorbell transformer:
Hi wcraft, exactly, thank you very much for your support and patience!