High impedance
I have been working with the CYTON board for a little while, but do not get impedances lower than 500 kOhm. Initially I thought there was a very large resistor in series for each channel, but I found out that the serial resistor should be 2.2 kOhm. I then tried shorting the bottom pins (IN1N through IN8N) to SRB2 and this gave results from 540 kOhm to 630 kOhm. I found the following appendix about measuring impedance (https://openbci-stream.readthedocs.io/en/latest/notebooks/A2-electrodes_impedance.html) and it looks like I should be getting 0 or 2.2 kOhm from those measurements, depending on whether the 2.2 kOhm serial resistor is withdrawn from the result before showing it in the OpenBCI GUI. From a post I found (https://openbci.com/forum/index.php?p=/discussion/3773/question-on-using-both-top-and-bottom-pins-for-eeg) I could see that the top pins (IN1P through IN8P) as a default should be bussed together as the SRB2 pin, so I tried shorting the bottom pins to the corresponding top pins (e.g. IN1N to IN1P) instead of shorting to SRB2. This consistently gave 1 kOhm while shorting to SRB2 gave 540-630 kOhm, so it looks like there is something wrong with the bussing together of the top pins to the SRB2 pin.
It looks to me like there is something wrong with the board, or the bussing together of the top pins could be disabled. Has anybody experienced this or have other ideas as to what the issue could be? And if the bussing together of the top pins is disabled, how do I enable it again?
Comments
Hi Torsten,
Are you familiar with the Hardware Settings for Cyton?
https://docs.openbci.com/Software/OpenBCISoftware/GUIWidgets/#hardware-settings
Check what you are using, this is the example shown on that page,
Also note the Load and Save buttons on that page. If you have changed any of the defaults and 'Save'd them, then these settings persist and override the defaults. The first line shows what the user has adjusted himself. The other lines show the default settings. You want your settings to be the defaults.
If indeed you do have the default settings, then let us know, there may be something odd with your board.
William
Hi William
I am familar with the hardware settings panel. I have been using it to change the gain and turn bias on and off. However, I have not changed anything else and I have only used the "send" function. Never the save or load functions.
Also, I think that it does not matter for the impedance measurement what the setting shown in the picture are, because they are back to the default after the impedance measurement. I just tested (pins shorted to SRB2) with different gains and bias on/off, and everything resulted in impedances og 530-600 kOhm
Sorry, I just understood better what you wrote. Yes the settings are the default of x24 gain, normal input, bias included, SRB2 on, and SRB1 off, for all channels.
Please send an email to customer support (contact at openbci.com) and reference this Forum thread.
I wrote to support and they asked me to make measurements with conductive gel/paste and send pictures of time series measurements.
I started doing that, and decided to go back to doing exactly the same as explained in the EEG headband kit guide, in stead of the setup I was using. This gave lower impedances. After a few tests I found out that there is low resistance to the BIAS channel. Removing BIAS electrode gave high resistance, while removing SRB2 electrode did not change resistance. I tried shorting the lower pins to SRB2 and BIAS, and it still gives around 500 kOhm to SRB2, while the impedance was 0 Ohm to BIAS.
I have written to customer service with these results.
Can I just use the BIAS channel in stead of the SRB channel and then everything is fine? I also thought about making an addon that shorts the top pins to SRB2?
I don't understand the current situation with the Hardware Settings that the Impedance Widget is using. I was under the impression that everything worked fine with the older impedance 'buttons' (on Time Series graph) that preceded the implementation of the Impedance Widget.
No, you cannot eliminate the SRB2 electrode from the head. This is what each channel is using as a reference electrode. In the default Hardware Settings configuration, all the IN1P-IN8P channel pins are bussed together and appear as SRB2 pin. These act as reference for all the IN1N-IN8N channel pins.
William
I've added this thread as supplemental info for https://github.com/OpenBCI/OpenBCI_GUI/issues/1204.
So as I understand it, right now we do not know whether it is the hardware or the software that is not working as intended? Do we have anyone that can confirm correct measurements from the cyton signal widget?
If IN1P-IN8P should be bussed together and appear as SRB2, does that mean that they should be considered as short circuited? If so, then measurements from the lower pins (e.g. IN1N) should be more or less identical to all those pins. In my case those measurement are far from identical (see attached table of measurements). Again, if my understanding of "bussed together" is correct, then the following options exist as I see it:
1: My hardware is not working as intended
2: The software makes the hardware have settings so that IN1P-IN8P are not bussed together to SRB2 when performing impedance measurements
3: The software does not perform impedance measurements between SRB2 and the respective lower pins
Some of my measurements are similar to what is reported in the other thread, though some are also very different. As example, I have not yet seen a 0 kOhm measurement when connecting lower pins to SRB2, without having other connections present also.
To be consistent with the ADS1299 data sheet,
https://www.ti.com/lit/gpn/ads1299
I've edited the pin names to their proper format. I.e. changed INP1 to IN1P. Sorry for the typos. But no need to change your table, the meaning is clear.
The OpenBCI staff engineer / manager who has more expertise in this area, was out of the office yesterday. He may be back today or later in the week. @philippitts.
I'm not sure you can assume that the INxP bussing that SRB2 performs, is exactly the same as a mechanical relay / switch would function, i.e. as shorting those pins electrically. The action happens inside the input multiplexer signal block of the ADS1299 as shown in various data sheet diagrams if you search for SRB2.
No, the question is, does the current Impedance Widget in the GUI operate exactly the same as the (previous GUI versions) 'ohm' buttons that were present in the left part of each channel in the Time Series window. It is possible that the Hardware Settings that the old impedance system used differ than what the widget is doing.
As you can see from the Github issue, the impedance issue has been previously reported with other customers, so it is unlikely that your hardware (ADS1299 chip or Cyton board) is at fault. Therefore a software fix seems more probable.
Torsten, hi again.
Philip was taking a closer look today. I believe a resolution of this is possible on the software side (GUI). Did you examine the 'openbci-stream' page below? (Mentioned on the Github issue thread.) He is only connecting between IN1N and SRB2 with a 10K ohm potentiometer, and it works just as expected.
https://openbci-stream.readthedocs.io/en/latest/notebooks/A2-electrodes_impedance.html
That was based on a university developed interface package from 2019 to 2022. However the Brainflow library is what the GUI uses and is what we recommend for all users wanting to interface Cyton or Ganglion directly to their own code. I'm sure that the engineers will be comparing how the impedance implementation differs between these implementations.
Again, apologies for the time delays in sorting this out. Almost all of the engineering staff time is being consumed by the Galea development and testing.
Regards,
Torsten, great news. Philip just found the version of the GUI that introduced the impedance issue. See the Github issue link above. A fix to latest GUI should be forthcoming. There is also the option of running the old GUI v5.0.8, if you need immediate functionality.
Philip has found the fix! A newly built GUI version should be available shortly. Will update with the number / download here.
That sounds very good. Thanks for the looking into it :-)
Until the new GUI app download binaries are regenerated, you can apply the fix yourself using the GUI source code.
https://github.com/OpenBCI/OpenBCI_GUI/pull/1211
https://github.com/OpenBCI/OpenBCI_GUI/pull/1211/commits/b3918c8814e1d7869e0c9430812724671bfda3ec
https://github.com/OpenBCI/OpenBCI_GUI/issues/1204#issuecomment-1970392296
Currently most staff are busy on the Galea project, several important demos at conferences are happening this week. Once things settle back to normal, new download binaries should be available.
https://docs.openbci.com/Software/OpenBCISoftware/GUIDocs/#running-the-openbci-gui-from-the-processing-ide
Thank you. However, I have delivered my headbands + cyton board to my project partners who are going to use it, so I will not be able to do anything in the near future.
Suggest passing on the link to this thread, so they have a workaround until the new GUI version binaries are posted.