I tried to connect Muse 2 with brainflow a couple days earlier (as suggested also by wjcroft) but did not manage to do it..
Anyways thank you so much for your reply!
Didn't know it had a good community so maybe I should try it again and ask for help.
re: typical EEG amplitudes, if your electrodes, amp, and filtering (removal of < .5 Hz 'DC offset'), are correct: typical EEG is below 80 or 100 microvolts. If you are seeing more than that, it is likely EMG muscle artifact. On headband systems this can happen easily with movement of jaw, eyebrows, etc.
I already did and they helped already! I was using the old CSR Dongle instead of BLED112.
I will purchase the correct one and find out if it works in Brainflow. Thanks a lot!
Hello again!
I'm re-evaluating the use of Ganglion for my experiment and I'm trying to figure out if my EEG values are considered reasonable.
I attached gold cup electrodes to F3, F4, P3 and P4 positions referenced to A1 and grounded to A2. I do not own Ultracortex, so Ganglion is not attached to my head, it lies at my desk next to me.
Impedance was up to ~42kΩ for the whole session.
So when opening the (raw) recording "OpenBCI-RAW-2022-01-25_19-06-45" with notepad++, I see that electrodes 2, 3, and 4 do not seem to alternate their value. It's always negative or positive!
And more specifically, electrode 3 has an average of raw -900μV while the others have raw values of 300-400μV... Does that mean the electrode 3 is broken?
Also, the strange fact is that if I open this Recording file from the OpenBCI GUI, due to the filters applied (Notch 50 and 1-50Hz bandpass), nothing seems strange. The GUI indicates that the filtered EEG value is between -20 and 20 μV. Is this possible?
Maybe electrodes 2 and 4 are also broken? But if I open another recording file with notepad++ ("OpenBCI-RAW-2022-01-25_18-37-43") from the same session, electrode 2 and 4 seem to alternate their value this time.
What's happening here? Are the values reasonable?
Is there any general rule to infer the signal / electrode quality of raw data so that I can know with a quick glimpse if my data seems reasonable or not? I could also check the GUI for filtered accepted values but that is not possible every time that' why I insist on raw values.
I've also noticed that retiutut was asking about the accelerometer widget. What can we infer from that?
Do you have any 'new' gold cups you can try, to compare? The raw text CSV files do look like they have a consistent DC offset, which usually does not happen with Ganglion. May be an indication of a chipped or damaged gold plate coating. When this happens a 'galvanic' battery effect is produced between the gold plate and the underlying metal electrode (usually tin or silver). Previous comments on this thread show ways to clean gold cups which do not involve any friction or loss of the gold plate.
No, I do not have any new ones.. But I think the package had some more electrodes so I will check and post back.
Is the constant positive / negative value an indication of DC noise or the large raw values (300, 900μV)?
What if the reference or ground electrode is damaged? How can I know that?
Thanks.
Now that you are talking about DC noise, I must also say that I had to place Ganglion about 1 meter away from my laptop because the noise was clear in the GUI when not 1 meter away.
Also, I noticed that if my feet were not on the ground, DC noise was also clear in the GUI! (Shouldn't the ground electrode take care of this?). These two were taken care of before starting the session.
Hi celticharp,
The constant positive/negative value you mention is the result of the galvanic effect that William mentioned. I also had this issue (discussed earlier in this thread, neg 3000-4000!), and was able to mostly rectify it with new electrodes.
I think there is mention of a way to test the ground also in this thread....
Yes, William mentioned a technique for troubleshooting that may help:
"With clean, 'new' electrodes, you can use a glob of paste to connect a test channel, reference, and ground (bias) together. This should always produce a channel signal of very close to zero microvolts. If your Ganglion cannot do that with non-chipped electrodes, then something is wrong with the Ganglion board."
@celticharp said:
Now that you are talking about DC noise, I must also say that I had to place Ganglion about 1 meter away from my laptop because the noise was clear in the GUI when not 1 meter away.
Also, I noticed that if my feet were not on the ground, DC noise was also clear in the GUI! (Shouldn't the ground electrode take care of this?). These two were taken care of before starting the session.
This is interesting. Let us know how you go with fixing this...
Just disconnect your ground lead while you're streaming data, see what happens to the output.
Hi
I've used Cyton with dry electrodes for my experiment. I'm getting amplitude between +/- 100microVolt for most of the trials. But sometimes the values gets over 2500microVolt. Is this because of some problem with my hardware or is it normal? I'm getting these kind of values at random channels.
I'm attaching some plots from experiment.
Momentary (very) high values at channels are usually due to motion artifacts, EMG (involuntary scalp muscles, eye movement), etc. Most legitimate EEG is below 100 uV, even below 80 uV.
Comments
I tried to connect Muse 2 with brainflow a couple days earlier (as suggested also by wjcroft) but did not manage to do it..
Anyways thank you so much for your reply!
Didn't know it had a good community so maybe I should try it again and ask for help.
The Brainflow community support platform is on Slack, sign up through,
https://brainflow.org/
re: typical EEG amplitudes, if your electrodes, amp, and filtering (removal of < .5 Hz 'DC offset'), are correct: typical EEG is below 80 or 100 microvolts. If you are seeing more than that, it is likely EMG muscle artifact. On headband systems this can happen easily with movement of jaw, eyebrows, etc.
William
I already did and they helped already! I was using the old CSR Dongle instead of BLED112.
I will purchase the correct one and find out if it works in Brainflow. Thanks a lot!
Hello again!
I'm re-evaluating the use of Ganglion for my experiment and I'm trying to figure out if my EEG values are considered reasonable.
I attached gold cup electrodes to F3, F4, P3 and P4 positions referenced to A1 and grounded to A2. I do not own Ultracortex, so Ganglion is not attached to my head, it lies at my desk next to me.
Impedance was up to ~42kΩ for the whole session.
So when opening the (raw) recording "OpenBCI-RAW-2022-01-25_19-06-45" with notepad++, I see that electrodes 2, 3, and 4 do not seem to alternate their value. It's always negative or positive!
And more specifically, electrode 3 has an average of raw -900μV while the others have raw values of 300-400μV... Does that mean the electrode 3 is broken?
Also, the strange fact is that if I open this Recording file from the OpenBCI GUI, due to the filters applied (Notch 50 and 1-50Hz bandpass), nothing seems strange. The GUI indicates that the filtered EEG value is between -20 and 20 μV. Is this possible?
Maybe electrodes 2 and 4 are also broken? But if I open another recording file with notepad++ ("OpenBCI-RAW-2022-01-25_18-37-43") from the same session, electrode 2 and 4 seem to alternate their value this time.
What's happening here? Are the values reasonable?
Is there any general rule to infer the signal / electrode quality of raw data so that I can know with a quick glimpse if my data seems reasonable or not? I could also check the GUI for filtered accepted values but that is not possible every time that' why I insist on raw values.
I've also noticed that retiutut was asking about the accelerometer widget. What can we infer from that?
Thank you!
P.S. The Recording files: https://we.tl/t-suYL4sj4U3
Do you have any 'new' gold cups you can try, to compare? The raw text CSV files do look like they have a consistent DC offset, which usually does not happen with Ganglion. May be an indication of a chipped or damaged gold plate coating. When this happens a 'galvanic' battery effect is produced between the gold plate and the underlying metal electrode (usually tin or silver). Previous comments on this thread show ways to clean gold cups which do not involve any friction or loss of the gold plate.
William
No, I do not have any new ones.. But I think the package had some more electrodes so I will check and post back.
Is the constant positive / negative value an indication of DC noise or the large raw values (300, 900μV)?
What if the reference or ground electrode is damaged? How can I know that?
Thanks.
Now that you are talking about DC noise, I must also say that I had to place Ganglion about 1 meter away from my laptop because the noise was clear in the GUI when not 1 meter away.
Also, I noticed that if my feet were not on the ground, DC noise was also clear in the GUI! (Shouldn't the ground electrode take care of this?). These two were taken care of before starting the session.
Hi celticharp,
The constant positive/negative value you mention is the result of the galvanic effect that William mentioned. I also had this issue (discussed earlier in this thread, neg 3000-4000!), and was able to mostly rectify it with new electrodes.
I think there is mention of a way to test the ground also in this thread....
Yes, William mentioned a technique for troubleshooting that may help:
"With clean, 'new' electrodes, you can use a glob of paste to connect a test channel, reference, and ground (bias) together. This should always produce a channel signal of very close to zero microvolts. If your Ganglion cannot do that with non-chipped electrodes, then something is wrong with the Ganglion board."
This is interesting. Let us know how you go with fixing this...
Just disconnect your ground lead while you're streaming data, see what happens to the output.
Hi
I've used Cyton with dry electrodes for my experiment. I'm getting amplitude between +/- 100microVolt for most of the trials. But sometimes the values gets over 2500microVolt. Is this because of some problem with my hardware or is it normal? I'm getting these kind of values at random channels.
I'm attaching some plots from experiment.
Razen, hi.
Momentary (very) high values at channels are usually due to motion artifacts, EMG (involuntary scalp muscles, eye movement), etc. Most legitimate EEG is below 100 uV, even below 80 uV.
William