OASIS98: ADAMS Messages: 2 Entries..

Jul 29

Marigold down: 7/28 22:00 - ~10:30 CDT
Ragwort down briefly:  7/29 around 9:40, 10:40 and 13:00 CDT

Upon arriving at base station, discovered Marigold was down.
To isolate problem we used components from Ragwort, eventually finding
some small bites in the fiber optic cable.  Replaced cable with 
spare armored fiber cable spool between ragwort and marigold.

Datel S/N 440690 seems to be marginal, and causes backplane
problems.

When ragwort (S/N 9006) is the Matrix CPU, it does not receive A/D
data (I assume it is not receiving interrupts from the Datel).  The
bootup goes well, no errors are seen. The green LED is pulsing at a 1
second interval (the A/D interrupt period) which I don't think is
normal.

A week ago ragwort was reporting "HSTSTA ACK failed" messages on boot
indicating it couldn't talk to the Datel at all (though I'm not sure
that it was with #440690).  At that time I blamed ragwort.  It would
also intermittently have problems talking to the ironics, reporting
"serialSock: read 1024 bytes, result -1: S_errno_ECONNRESET", which
causes the whole data acquisition process on the Matrix to crash.

When cosmos (Matrix S/N 940211101) is swapped in (configured as
ragwort), it seems to work with Datel 440690 (the A/D data comes
through), but gets an ECONNRESET error talking to the Ironics.

When the 440690 is replaced with Datel 320689 all is OK, with either
Matrix card, so that is why I'm blaming the Datel, and I'm hoping
that both of these problems (no A/D data and ECONNRESET with the
Ironics) can be traced to the Datel.  It's either that or El Nino.

Further tests:  Tried Datel 440690 with marigold.  It came up ok and
actually worked for a few minutes, then died.  It then reported
HSTSTA ACK errors on bootup.

Currently we're running 440690 in cosmos, with only channel 107
and no serial channels active.  So far so good.  It could act
as a spare datel if used with cosmos in this way.