The WXT surface met sensor stopped reporting for an as yet undetermined reason.  Attempted to replace with another WXT on the same DSM port but that gave the same sensor not found error, suggesting a DSM problem.  To replace it we installed a Lufft ws700 adjacent to the WXT and that appears to be recognised by the DSM and it's data is being ingested (from around 11UTC, 17 Dec).

 

1 Comment

  1. Gary Granger

    I'm guessing the "sensor not found" problem was a message you saw when running "rs 3", something like below:

    daq@dsm-sodar:~ $ rs 3
    connecting to inet:localhost:30002
    connected to inet:localhost:30002
    sent:"/dev/ttyS3
    "
    line="ERROR: sensor /dev/ttyS3 not found."
    Exception: IOException: socket: read acknowledgement: ERROR: sensor /dev/ttyS3 not found.

    That is actually a bug in the rs script.  Port 3 should map to device /dev/ttyUSB3, but rs tries /dev/ttyS3 instead.  The workaround is to give the full device path: rs /dev/ttyUSB3.

    So it's possible the replacement WXT would have worked.  Nothing is being recorded on /dev/ttyUSB3 at the moment, so I presume the replacement WXT was removed.

    Perhaps you knew this, or perhaps it was just luck, but port 1 was already configured for RS485 half-duplex, and it was already configured in the NIDAS XML file, otherwise you would not have been able to see any data from it.

    The rs script would be fixed by an update on the DSM, but I am avoiding that until someone can be around to fix it if it breaks.  As long as the surface messages are being recorded, an update is not necessary.