impedances, less with pressure; what is normal range? [resolved]
I have several problems to measure EEG using OpenBCI_GUI.
The problems may consist of several causes.
The impedance of the target electrode sometimes goes very high ( > 5000 kOhm, or Railed ).
When I push down the headset, the impedance decreases to several hundred kOhm, but if I release the hand, the impedance goes back to the high value.
The electrode may not contact the skin, but it seems to contact it as far as I can see.
Can you suggest what is wrong?
It is difficult to set the headset correctly when the subjects have much volume. Can you have any good technique to avoid hair and contact the electrodes and skin?What is the appropriate impedance value?
The smaller the value, the better the impedance I think.
How much should I decrease the value at least for the proper signal quality?
100 kOhm, 300 kOhm, 500 kOhm?The values of EEG signals sometimes go to near 0 uV.
I cannot identify the reason.
After I switch the board off and on, it works correctly.
I think something wrong with the board.
Comments
Can you please provide a screenshot?
Hiroki, what electrodes are you using, dry, paste, gel?
Please see this thread with suggestions on RAILED condition.
https://openbci.com/forum/index.php?p=/discussion/1481/how-to-become-de-railed
Regards, William
Hi everyone,
I would like make sure a minimum quality of my signals before I start the data acquisition step of my project.
So, I think needing to have a good impedance on electrodes, but I'm having some trouble.
I did three tests, the first using my Ultracortex Mark III printed by myself, below is a screenshot where I'm measuring the impedance to the electrodes 7 and 8:
The second test was using the headband kit, below is a screenshot where I'm measuring the impedance to the electrodes 1, 2, 4 and 5:
And the last test was using the gold cup electrodes with Ten20 Conductive Paste, below is a screenshot where I'm measuring the impedance to the electrodes 1, and 2:
Looking the readme of the OpenBCI_NodeJS_Cyton github repository, I know that a Good value is < 5k Ohms, which is not the case for any of my tests.
If anyone can help me, thank you very much!
Mauricio, hi.
5K ohm impedance is only possible with wet electrodes (gel or paste). And generally some type of skin prep is usually required to get as low as possible. Such as removing skin oils or skin abrasion with NuPrep.
The 5K impedance figure was a relic of the older amplifiers which had low input impedance (in the megohms range). Cyton (ADS1299) has very HIGH impedance inputs (1 gigaohm at the front end), which means higher skin-electrode impedances are acceptable for both dry and electrolyte contacts.
https://openbci.com/forum/index.php?p=/discussion/2517/expected-impedance-range-with-ultracortex-cyton-resolved
https://openbci.com/forum/index.php?p=/discussion/2206/what-is-the-relationship-between-high-impedance-on-eeg-channels-and-50-60hz-noise
Regards, William
There will be "noise" when impedance check is active. The only thing that is meaningful is the XXX kOhm reading on the left of each ChannelBar in TimeSeries.
Richard, thanks. I was spacing out the 31 Hz generated frequency used to find the impedance. Yes, this is actually normal. Though I seem to recall really older GUI's did not show it so prominently. Maybe because of the way the scale factors were set.
So, the impedances in the test with my Ultracortex Mark III (which is what I want to use) are acceptable.
Thanks for the quick answers!
It is difficult to find a way to reduce the impedance.
What is the best and easy way to make the electrodes contact the skin at low impedance with no (less) pain?
Please read this new thread regarding expected impedance ranges with the dry electrodes and Ultracortex,
https://openbci.com/forum/index.php?p=/discussion/2517/expected-impedance-range-with-ultracortex-cyton-resolved
William
Hi William,
Thank you for the thread about the impedance ranges.
I understand that the range around < 100 - 200 kOhm is good.
And > 2000 - 3000 kOhm is not appropriate because the OpenBCI GUI often displays RAILED when the impedance is high, that means the electrodes don't contact skin.
Now I want to know 'how not to make subjects feel pain when I adjust electrodes?'
Did you read the previous thread link?
https://openbci.com/forum/index.php?p=/discussion/comment/13328/#Comment_13328
1000Kohm (1 megohm) impedance is only < 0.1% error. 200Kohm is only possible in forehead area with flatter electrodes. Richard was seeing 900Kohm in his hair areas (best case) with the short combs. Even 2 megohms will only result in 0.2% error. Richard was not seeing any RAILS at either 2.8 megohms, 4.8 megohms, etc.
https://openbci.com/forum/index.php?p=/discussion/comment/13315/#Comment_13315
Generally RAILED condition can be other factors. You don't need to induce pain to get lowest possible impedance. Please re-examine Richard's @retiutut 's numbers.
William
Oh I'm sorry.
I'll check that.