High Band Power + FFT Plots
Hello!
I just got my Cyton board + gold cup electrodes a few days ago, and have been experimenting with collecting some of my EEG and ECG data. I'm very new to this, and am wondering if anyone else has faced the following problems and / or has any idea on how to debug or a crack at a solution.
First, on my Time Series plot, it looks like I'm seeing a bit higher uVrms than what I saw in the tutorial; sometimes they go all the way up to a couple thousands, and am wondering if this is just interference, or if there is anything I can do to get a cleaner / lower / more accurate number.
Secondly, I think these inflated numbers may have an impact on my FFT plot, as it nears 100 max uV (and sometimes surpasses 100), so I have to adjust the scale to 1000 in order for these values to look more normal. It looks like the Band Power graph is also affected, as the bars look static (capped at 100 for all the waves, all the time).
Any ideas on how to debug, or what I could do to fix the situation?
Comments
Michelle, hi.
Generally when most / all channels are showing similar patterns, suspect your reference connection. Since reference is shared in common with all channels.
Where is your reference and ground (Bias), located? Are you sure that you have good connections there and on the scalp? You can press the 'ohm symbol' (Ω) to the right of the channel number, to check the impedance for each channel. With gold cups and paste, you should be getting fairly low readings. Say below 20Kohms.
Also be aware that certain hair care products such as conditioners, can interfere with the skin contact. Below is a tutorial on cup electrode placement.
https://openbci.com/forum/index.php?p=/discussion/176/placing-electrode-sensors-on-the-scalp
Regards, William
Hi everyone, thank you so much for your quick responses!
@wjcroft, for the first question my reference and ground are connected on A1 and A2. I tried reattaching those on my earlobes and ones on my scalp, but am still getting high uVrms (including one at Fp1 that is not obstructed by hair).
As a side note, I did change the gain to 8x for channel 4 (ECG) based on this tutorial. Should I reconfigure it back to 24x?
Is there anything else to do now when it comes to debugging other than experimenting with the physical hardware? Should I install and start prototyping on the new GUI v5.0.0-beta.0? Or hold off until the date it is released to the public / when should I look forward to this release?
Again, really appreciate all the help from everyone so far!
I would expect 5.0.2 in the next week or so. It has been in testing for some time, and it seems to be stable.
Michelle, try checking your impedances. They should be relatively low, below 50K or 20K ohms. Back in the 'old days' even 5K was a target using older amplifiers. Be sure you do not have any hair conditioners or skin lotions that may interfere with the contact.
Hi @Prageethda, thanks for following up.
I've been spending some time trying to debug, but yesterday I downloaded the 5.0.2 update that was released this week, and it's been stable ever since.
https://github.com/OpenBCI/OpenBCI_GUI/releases/tag/v5.0.2
Hope it fixes the problem for you too!