Syncwind, hi. Use the search box on the lower right. There are lots of threads discussing impedance issues. In a nutshell, the impedance of the ADS1299 is very high, 1 Gigaohm. Therefore electrode / skin impedance is not such an issue.
The paper below demonstrates that even skin impedances in the range of ~50K ohms or higher had no impact on the measured EEG. Compared to older EEG amps that required ~5K ohm impedance.
uVRMS is displayed for each channel during offline data playback, if I understand correctly this value relates to electrode impedance? (I usually used to see this values in kOhms).
I'm a bit confused though of how is it possible to compute this in playback mode(not live) when no such data is saved in the actual log(I was under the impression that lead off detection should be used for that).
> uV RMS is displayed for each channel during offline data playback, if I understand correctly this value relates to electrode impedance?
Mike, as you say, the impedance measurement only makes sense with live data streaming from the board. See Joel's previous post. When the buttons are pressed (under P and N columns), the ohms value will appear.
Well, as far as I could tell the impedance values during playback were pretty much the same as the impedance values that were computed during live session. That's why I'm surprised.
If impedance values are being shown during playback, then yes that could likely confuse users and should probably be disabled in the OpenBCI_GUI software. So just avoid that anomalous readout in playback mode.
To check if the impedance values look valid or reasonable, in live mode, press or lift on the electrode in question. You should see ohms values change.
The impedance check is actually done by injecting a small sine wave into the signal path. See Chip's posts,
So indeed, this same calculation could be done on any playback signal; however we would then have no idea whether the ADS1299 channel in question had actually been configured for that injection or not. So any ohms values shown in playback should be ignored. (Just uncheck the buttons.)
The ADS1299 amp input impedance is actually degraded somewhat in this injection mode. So best to leave it off when not using it. Nominal input impedance is 1 Gigaohm. When injecting it drops to something like 500 Megohm.
Correct, it's not using the hardware "lead off" detection. Instead it's the signal injection I mentioned previously and explained further on Chip's blog posts. So that same "information" is present both live and from the recording. However from the recording we DO NOT know the status of the ADS1299 injection "switches" (controlled by the P and N column checkboxes) -- hence impedance values shown during playback may or may not be accurate.
@Mike asked a question on another thread that I moved over here:
One more thing, I would like to make sure that the Bias and reference electrodes are also properly connected, is it possible to measure the signal quality of these two signals?
See the two blog links by Chip @chipaudette on the April 7 post above. He mentioned that Reference impedance could be measured with some caveats. But not Bias.
And also see the pdf on the very first Comment of this thread. Since the ADS1299 amp input impedance is so high (1 Gigaohm), actual skin-electrode impedances are not that critical. And in real life saline electrode systems, such as the EGI Geodesic Sensor Net mentioned in that paper -- no degradation was found.
Comments
When the small square in the 'N' column is checked, screen will show the measured impedance value for that channel.
channel during offline data playback, if I understand correctly this
value relates to electrode impedance?
Mike, as you say, the impedance measurement only makes sense with live data streaming from the board. See Joel's previous post. When the buttons are pressed (under P and N columns), the ohms value will appear.
To check if the impedance values look valid or reasonable, in live mode, press or lift on the electrode in question. You should see ohms values change.
The impedance check is actually done by injecting a small sine wave into the signal path. See Chip's posts,
http://eeghacker.blogspot.com/2014/04/openbci-measuring-electrode-impedance.html
http://eeghacker.blogspot.com/2014/04/impedance-of-electrodes-on-my-head.html
So indeed, this same calculation could be done on any playback signal; however we would then have no idea whether the ADS1299 channel in question had actually been configured for that injection or not. So any ohms values shown in playback should be ignored. (Just uncheck the buttons.)
The ADS1299 amp input impedance is actually degraded somewhat in this injection mode. So best to leave it off when not using it. Nominal input impedance is 1 Gigaohm. When injecting it drops to something like 500 Megohm.
William
See this post for more background.
http://openbci.com/forum/index.php?p=/discussion/84/ads1299-electrode-impedance-measurement-algorithm