Moller: analyzer or hardware?

Log entry time 10:36:55 on February 11,2002

Entry number 47367

This entry is a followup to: 47362

keyword=Moller: analyzer or hardware?

The problem with the Moller analyzer is actually a problem, I believe, with
the data.

Analyzing run 43600, which ends with a floating point exception, I note that
scaler group 3, which is supposed to count during EITHER helicity state,
appears quite sick:
RAW SCALERS
scaler1 scaler2 scaler3 3-(2+1)
521562656 521535968 -551752640 -1594851328 01: Microsecond Clk
254433184 254414320 -540612032 -1049459584 02: Unser Monitor
218658672 218699552 1332684672 895326464 03: BCM1
217754368 217791120 1300100352 864554880 04: BCM2

and so on. The third scaler (3rd column) should be very close to the sum of
the first two.

Running ~cdaq/gen/coda2/crl12.2/RPC68k/ratemon (which is quite similar
to, but not as pretty as, realmoller), I get

------------------- Moller Detector Live Rates -------------
Live Time interval: 0.000 s
ID Raw_Change Rate (Hz) Calculated

Clock 0 0.0 1.000 MHz (by definition)
Left Singles 0 0.0 -0.000 Hz/uA
Right Singles 0 0.0 -0.000 Hz/uA
Conicidences 0 0.0 -0.000 Hz/uA
Accidentals 0 0.0 -0.000 Hz/uA
PHT Errors 0 0.0 0.000
BCM1 0 0.0 -0.156 uA
BCM2 0 0.0 -0.155 uA

which tells me that NOTHING is counting. Hmmm.

Logged in to ROC8 and did a reboot. Reboot is OK, but the scalers still don't
report correctly, and some of the processes soon die.

I next power-cycled the crate. After it booted, the scalers report correctly.
I believe things are OK now, but I don't know what started this problem
or why a power-cycle was required.

As for the problem runs of this morning, the results quoted from 'realmoller'
are probably meaningful -- but Marko should confirm this to be true.

The tcl script 'realmoller' might need another box to monitor the status
of the third scaler.

User name hcf

Log entry time 10:36:55 on February 11,2002

Entry number 47367

This entry is a followup to: 47362