noise in signal from my DIY ADS1299 board (not Cyton)
Rand
England
Hello everyone. it is my first question in the forum. I have a little problem.
When i measure EEG signal in a Y- axis i have not stable signal (signal is good, but Y-axis). It is looks like in attached photo. In ads1299 i use bias output signal.
Comments
Rand, hi.
What program are you using to read from the Cyton? OpenBCI_GUI? Or Brainflow, etc.? Does the signal look ok in GUI?
Discontinuities and jumps are seen in this signal (straight lines), which could indicate some kind of signal loss or miscalculation. Is this Cyton (ADS1299) from OpenBCI, or something you built yourself?
Also be aware of DC offset, which is usually removed with pre-filtering.
https://openbci.com/forum/index.php?p=/discussion/201/large-millivolt-data-values-fbeeg-full-band-eeg
William
Dear William, thank you for your quick reply. It is my private board with ADS1299. This signal i receive without any filter (raw data). Thank for your link. It is mean that high pass filter will be enough for decide this problem?
Your straight lines in the graph look problematic, evidence of dropped samples or something weird.
Typically EEG is filtered first with a bandpass filter say from .5 Hz to 45 Hz, then with a notch filter at the mains freq (50 or 60 Hz). Although you might get by with just the highpass at .5 and the notch. EEG at gamma range (40 Hz and up) typically has some EMG (scalp muscle) contamination.
William
PS I moved your category to Built-It-Yourself, from Electrodes.
Hello, everyone, can not understand many days which type of noise in the signal, how do you think which reason can be for this anomaly signal? what signal so quickly jump sometimes. It is real-time
. In the pic length about - 3 sec. The signal from dry electrodes in Fz position
@Rand, hi. I merged your new thread into this existing thread. Which was concerning noise and anomalies in the signal from your DIY board that you built yourself. I assume you are not using the OpenBCI Cyton board layouts.
This looks like another potential glitch with your circuit. What improvements have you made since the original thread?
William
William, thank you for your reply. Now I use pass filter 1-40 Hz. I can detect blinking and chewing. I have a problem only with this -high-frequency noise. In the pic. - it is real-time. In the pic length about - 3 sec
But your original graph showed actual GAPS in the signal. How did you fix that?
Designing mixed signal analog / digital boards, is considered somewhat an art form. And many considerations must be met to achieve a good mixed signal design.
https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html
original graph - it is not the correct image, it was a problem with code for visualization. The last image is correct.
Do you use "star-grounding" in your mixed signal design? See previous link.
You originally posted the new thread under 'Electrodes' category. I moved it here to Build-It-Yourself. What made you think the issue is with electrodes? A good way to test an electrode issue, is to use (for example), gold cups and paste for all 3 connections: channel, reference, and ground. That would eliminate any possible issues with the dry electrode you are using. You did not mention the electrodes used for reference and ground.
I didn't use "star-grounding" I will read about it more carefully. I suspect that problem with the connection electrode to the scalp, because when I shorted the measurement electrode and reference electrode I receive a stable result - 0.5 (+-0.2 mkV) mkV.
Today I will try to make a test with golden electrodes.
I used dry electrodes from fri-fl.com. For bias and reference electrodes used for reference and ground, I bought in aliexpress these electrodes -" EEG Ear-clip Electrodes, Gold-plated Ear-clip Electrodes".
Quote, thank you for your advice. This result I received with gold electrodes, it looks better. Looks like the problem was with the dry electrodes. But how I can decide this problem, need more pressure, or replace reference electrodes?
> @wjcroft said:
The FRI dry electrodes are silver chloride. It sounds like you are mixing gold (reference / ground) and silver chloride. It's best to always use the same metals throughout. I think I mentioned this to you previously. Otherwise with mixed metals you can get a battery ('galvanic') effect.
Can you clarify?? Gold for all 3 connections? Paste?
In addition to the dry comb and cone electrodes, FRI also sells silver chloride ear clip electrodes. These would be recommended for use with an FRI dry electrode setup. For your reference and ground connections.
https://www.fri-fl-shop.com/product-category/electrodes/eeg-electrodes/
Quote, yes it was Gold for all 3 connections.
I will try to use silver chloride ear clips and compare them with gold chloride ear. Today I will make an order.
Thank you so much.