The channel of C3 electrode railed if connect the BIAS channel
Shawn
Chicago
in OpenBCI_GUI
Hi there,
We are trying to use Mark IV to collect EEG data, 8 electrodes.
We found the channel of the C3 electrode is always railed if we connect both SRB and BIAS. Once we disconnect BIAS or disable it in the setting, seems like everything works well.
Our question is: can we feel confident collecting EEG data without the BIAS channel? How would the result be impacted? Or How should we fix the problem (railed channel with bias channel connected)?
Any help will be greatly appreciated!
Comments
We are using Cyton + Daisy board.
Hi Shawn,
Is it possible you have a worn / bad electrode at C3?
Here are other tips on resolving the RAILED condition:
https://www.google.com/search?as_q=railed&as_sitesearch=openbci.com
https://openbci.com/forum/index.php?p=/discussion/1481/how-to-become-de-railed
William
Hi William,
Thank you for your comment.
We tested this by changing it with some works-fine electrodes but still doesn't work. And this also couldn't explain why it works without the BIAS channel connected.
Thanks anyway for your kindly advice.
Shawn
None of the tips worked?
https://openbci.com/forum/index.php?p=/discussion/1481/how-to-become-de-railed
'BIAS' has a dual function, it acts like 'Ground' on most other EEG amplifiers. And it also provides a 'counter-mains' injection to reduce the mains noise. It is ok to use the GUI settings panel to turn off Bias on your weird C3 channel, if that helps with the RAILED. However you should always have both Ground and Reference (SRB2) earclips connected to the Cyton.
See "Hardware Settings" section, on how to switch off the Bias injection for a given channel number.
https://docs.openbci.com/GettingStarted/Biosensing-Setups/ExGSetup/#visualizing-the-data-in-openbci-gui
Thanks for your replying.
We tried impedance testing for all channels and found the C3 channel's N status is ridiculously higher than others (4500 to 20 k΄Ω). Does that mean something?
If we turn off BIAS for C3 channel, is there any way to denoise and fix the data manually or by coding?
Thanks in advance!
If one channel's impedance is hugely different than the others, it means you have some issue with either the electrode, or the skin contact at that location. Or possibly with the spring in the electrode holder. You say you already tried swapping with known good electrodes. So investigate other causes.
I explained how Bias and Ground are combined functions on the ADS1299 / Cyton. Turning off the Bias injection sensing for one channel, will not require any adjustments to your signal processing. Because Bias functions are still happening on the other electrodes.
But, the C3 channel data will be impacted and noisier than others, right?
No, that is not the case. You can read more about how the Bias sensing and injection works in the ADS1299 data sheet.
https://www.ti.com/product/ADS1299
The C3 channel data will be fine. In fact your impedance is not that far out of line with the values mentioned in this link about expected Ultracortex impedances.
https://openbci.com/forum/index.php?p=/discussion/comment/13328/#Comment_13328
thanks for all these info, I just found that it still railed if I just turn off BIAS for the C3 channel. It can only work when I take off the BIAS earclip or turn BIAS off for all channels. And I found one more channel railed this afternoon.
Any tips on this situation?
I still believe you have issues with your electrodes. Assuming you read the tips here,
https://www.google.com/search?as_q=railed&as_sitesearch=openbci.com
https://openbci.com/forum/index.php?p=/discussion/1481/how-to-become-de-railed
re: working with Bias turned turned off for all channels in Hardware Settings.
That can work. Bias still acts as Ground. The only thing you would miss in that case is a bit of mains noise rejection. If you have your mains notch filter turned on, and you are not seeing significant mains in the FFT, then you are fine.
Do you have more in detail for this tip:- _Check board function: Connect SRB, BIAS, and railed Channel using jumpers / y-cables / ... This should give you roughly 0.00 uV. See picture (1)._ Since the picture link does not exist anymore.
>
Is there any code or other way to reduce the main noise in the post data analysis step?
Thanks in advance!
A common test of EEG channels, is to simply using paste or alligator clips, to connect all three wires / pins together: channel, SRB2 and Bias. This will result in 0uV if the channel is working correctly.
I already said that you just need to check your FFT widget and see that the notch filter at your mains frequency is being effective. That is all that is needed, a notch filter at the mains freq.
Is this what you mean in the picture? So the channel is all good right?
https://drive.google.com/file/d/1acpRCX5dipXrbK3qV4s4R5NER4SHfzft/view?usp=share_link
I did not post that 'picture'. No picture is really needed, I explained the steps. Connecting the three wires verifies that your channel properly zeros itself.
"all good" is assuming there are not other issues with the channel. This test just confirms that the differential amplifier correctly sees 0uV when channel and reference are connected.
Sorry, I mean in the picture I posted in my last comment.
Anyway, I figured them out, just because of the electrodes! The unbelievable part here is, we have more than 1/3 of spare electrodes malfunctioned. And the most coincident part is there is an electrode that worked for another channel but somehow didn't work after I moved it to channel C3! That's why I thought it could NOT be because of the electrodes.
Anyway, Thank you so much, William, thanks for all these days' help and all the info you shared with me!
Many thanks!
However, still wonder why the railed channel could work without the BIAS connected if the electrode is worn.