Main INDEX
Monthly INDEX
PREV
NEXT
User name gilman
Log entry time 14:43:34 on January 9,2003
Entry number 91216
Followups:
keyword=helicity setup
After Bob's confirmation with the VTOF that we really are delayed 9 cycles, instead
of 8, I have tweaked the timing of the delays we put on the helicity and QRT signals
that we send to the spectrometers.
We use the transitions of the helicity signal to generate 200 us delayed
start and stops which are then used to generate the delayed helicity signal.
We delay it 200 us so that during the 500 us MPS=false time, we can readout
the scalers with T9 event and get the helicity of the last pulse (that is, the helicity
of the pulse presumably 8 pulses earlier when in the usual delayed reporting
mode) that was integrated by the scalers. I have changed the delay from
200 us to 350 us. It should be stable to a couple us.
(Aside: MPS is the signal that tells us when the helicity is stable. It is
roughly true for 33 ms, false for 0.5 ms, and iterate...)
The QRT signal tells us when it is the first pulse of a group of 4.
The pattern is always +--+ or -++-, chosen pseudorandomly,
so one can always uniqely find the first one, in the middle of 1001-0110
or 0110-1001 patterns, for example. But QRT makes recognizing the first
one trivial.
The QRT pulse is generated by a 200 us delay, and has a fixed "33" ms
width. Over time, the module width had narrowed a few tenths of a percent,
so that the T-->F transition was 100 us delayed, whereas the F-->T transition
was delayed 200 us. Since Seonho's analysis saw only QRT=False for the T9,
this leads to the suggestion that for unknown reasons we typically read out the
scalers 150 +/- 50 us after MPS goes false, rather than nearly instantly as
we expect. I have readjusted the width of the QRT pulse, and increased the
delay to 350 us.
We have just started runs 20244/1243 (no, I don't know why we are out of sync.)
with this adjusted timing. It would be good to see if we see QRT=True in 1/4 T9 events.