Cyton channels switched off during recordings - occurs randomly

tzvettzvet Hefei, China
edited November 2018 in Cyton
Hello! We have a problem with channel(s) that suddenly switch off during recording sessions.

We have a Cyton+Daisy 16 channels system + MarkIV and we record 16 channels in the Board Mode 4 (in the GUI - Marker Mode widget) that allows to send Time stamps with commands (like: echo -n '`2' > /dev/ttyUSB0) and we write the data to the SD card for offline analysis. Now, the problem is that the system works generally fine, but in some sessions one channel suddenly is switched off (becomes zero!).

This is a rather awkward problem because:
1. it seems to appear randomly - sometimes it happens, sometimes it does not...
2. the electrode number is not related systematically to one of the Time stamp commands we send during the session (we got electrode nbs: 2,5,7,9,15 - a lot of 5 - that went off in different sessions while the triggers we send are 2,5,9 - but to switch off electrodes 9 and 15 one needs to send characters q and u - see OpenBCI Cyton SDK)

We do not have any idea why this suddenly appeared. I worked with the system during spring 2018 and did not see such behaviour.
We tried different "software communications" OpenBCI GUI v3.2.0 (+ its rc2) and v3.4.0 - all gave it (although it seems the 3.4.0 has less of that problem).

any idea why ?

from a simpler point of view: Could anyone point to the place in the code where I can stop the option "Turn channels off" during streamData mode. This would be far simpler than digging to find why the problem appears! And does this mean that I have to reupload the code to the Cyton/Daisy (i guess this will end up there...)?

Comments

  • wjcroftwjcroft Mount Shasta, CA
    Tzvet, hi.

    When did you purchase this Cyton? Do you know what revision of firmware it has? 

    It sounds like you are using Linux or Mac? Are you using the standalone GUI or running from Processing? If standalone, have you tried the Processing IDE method?

    Regards,

    William

  • tzvettzvet Hefei, China
    William, hi!

    We bought the system autumn 2017 (Sept-Oct), Cyton+Daisy+electrodes, but somewhere in Jan-March 2018 i re-uploaded a new firmware on the Cyton following a suggestion for solving other issues. It worked. The current Firmware is 3.1.0.

    We are under Linux (Lubuntu 16.04) with standalone GUI. Not running from Processing. We have not tried the Processing IDE method.

    Regards,
    Tzvetomir

  • wjcroftwjcroft Mount Shasta, CA
    You mentioned when recording, that you are sending Linux shell echo (Cyton SDK) commands directly to the Cyton serial port. Is it possible that the process(es) doing this could get confused and send the wrong SDK commands. Or for example some glitch occurs where multiple processes are writing to the port at the same time or with different baud rates, etc. Resulting with garbled SDK commands being sent?

    What if you comment out the echo commands and do an hour long test run. Do you lose any channels with that?

    Best,

    William

  • tzvettzvet Hefei, China
    Good point. But for your proposed test, as i said, the problem is not always appearing. We would need longer or multiple sessions to check it. Have to see.

    thanks for the help.
    Best
    Tzvetomir
  • wjcroftwjcroft Mount Shasta, CA
    You hinted that you might want to mod the firmware to disable channel deactivation. To do that edit this file,


    Search for the references to: "deactivateChannel"

    For example you could return immediately with no actions from that subroutine.
  • tzvettzvet Hefei, China
    Great! thx!
Sign In or Register to comment.

© Vanilla Keystone Theme 2026