HIFI Cryomag: K2700 IOC Improvements and fixes

LiamPanchaud commented 5 years ago

In the day of testing I have performed with the 2700 on HIFI, I've been observing the temperatures as they warm up. On 2 separate occassions, a contiguous block of sensors (from Shim Z Outer A to 2nd Stage B) has ceased updating after 40-50 minutes of operation. This needs to be investigated and fixed.

When a Keithley 2700 has been running for a long time its timestamps become a mess. It starts reporting that the timestamp as some number to the power 37 (which is very wrong).

The VI still reports temperature correctly but its timestamp and therefore drift columns go haywire.

The IOC, on the otherhand, just shows nothing. I believe this is due to overflows in the record fields. i.e. in the IOC log messages like No Channel Found for value: 210.000000 (TE:NDW1801:KHLY27c00_01:BUFF:READ: 11 bythes surplus input "+a9n.nel 90)E37,101" (note that this is also interleaved output which makes it even more confusing to read)

2 issues are covered by this ticket:

[ ] Investigate and fix the issue where a block of sensors stop updating
[ ] Investigate potential solutions to the timestamp overlfow and get input from the instrument scientists.

LiamPanchaud commented 5 years ago

I have so far been unable to reproduce the error in which channels stop updating

LilithCole commented 5 years ago

The equation given here as used in the original VI looks like it will slowly build up to the correct value over a large number of measurements, possibly asymptotically. It may be better to gradually smooth the drift, or not show it until a fully smoothed drift has been calculated.

ISISComputingGroup / IBEX

HIFI Cryomag: K2700 IOC Improvements and fixes #4219