H. Crannell, A. Longhi, D. Sober CUA 4-Feb-98
Initial timing:
Set discriminator widths to 30 ns, pending later measurment
of relative timing.
Plateau:
Set PS B field to nominal value.
Round 1 (PS singles):
Use individual PS scalers (EPICS or DAQ).
Hold L1-4 HV fixed, raise R1-4 HV in steps.
Hold R1-4 HV fixed, raise L1-4 HV in steps.
Set HV at approximate plateau value.
Round 2 (gated T-counter scalers):
Set PS logic to L.OR.R
Set T-counter scaler gates to PS and PS-accidental.
Turn off all but one PS PMT.
Vary HV, and record Gated/Ungated ratio (less accidentals)
for all T-counters.
(If no gated counts in any T channel, change B field.)
Setting magnetic field:
Set PS logic to L.AND.R
Set T-counter scaler gates to PS and PS-accidental.
Record Gated/Ungated ratio (less accidentals) for all T-counters.
Vary B field and repeat.
The optimum B field gives:
a) Gated/Ungated ratio non-zero for any T counter
b) Gated/Ungated ratio higher at the T1 (high-k) end
than at the T61 (low-k) end, to compensate for
the bremsstrahlung spectrum in statistics.
Set PS logic to L.OR.R, and repeat.
Note: If there are no holes in the PS efficiency with PS=L.AND.R,
we should probably run that way, unless the L.OR.R
efficiency is much larger (e.g. factor > 4)
PS counter timing differences:
DAQ trigger = PS Left, PS Right (separately)
Check TDC values of R1-4 vs. Left trigger, and of
L1-4 vs. Right trigger.
(This assumes that the Discrim.-Logic cable lengths are
matched and Discrim.-TDC cable lengths are also matched.)
When these differences have been determined and delays added
if necessary, then reduce discriminator output widths.
PS calibration vs TASC:
DAQ trigger = Master OR
Rates = 10kHz, 30kHz, 100kHz (TASC in beam)
1 MHz, 10 MHz (no TASC in beam)
Identify T counter from TDC's, find tagging fractions
from TASC and PS TDC's.
Compare with Gated/Ungated scaler ratios.
Systematic checks on PS:
Using data from above runs (or short runs with trigger=PS),
identify T-counter and PS-counter from TDC's.
Histogram T-counter distribution for each PS combination
L1*R1, L1*R2, ... L4*R4, and compare with Monte Carlo.