Notes on tests of Module 28
1. Warm tests (hybrid at 25C).
Observations (and possible solutions to problems, courtesy of pwp):
- Strobe delays were found to be ~38-40 instead of ~12-14 as found by RAL.
> CLOAC was probably using the wrong time bucket, ie we needed trig delay of 129 instead of 130?
- The characterisation sequence terminted 'normally' after the trim-range scan, ie the noise occupancy did not start.
> Our PC is a little slow. We need to increase the sleep in ST.cpp (routine st_start_stopper) from 1sec to 2sec.
- Strange result from the trim range scan - looked fine for ranges 0 and 1, but there was a significant shift
in the trim settings, and trimmed thresholds, for ranges 2 and 3
> Not understood. But the optimised trims used range 2, so all subsequent scans (ie Response and Noise occupancy)
showed a very significant shift in threholds!!
When tackling these problems, we started many scans in various random orders, so the data is not easily available
in an ordered way.
2. Cold tests (hybrid at 0-3C)
Generally much better behaved, ie
the strobe delays were now fine (after the second strobe delay scan!!), and the trim range scan looked fine (ie no shift as
observed at room temperature). Why the difference? Apart from the temperature, the freezer tests use much longer cables (~6m compared
to ~2m).
We ran a normal characterisation sequence at this temperature, which worked
fine (apart from stopping after trim-range scan, as we have not recompiled after changing from sleep(1000) to
sleep(2000)) - this is the data available on the summary page.
Long term test - was running well, but hung after 22 hours due to LV trip. The sctdaq s/w did not
try to restart the LV, and pwp will look into this.