Hello,
with help from tavi i recollected these recent/known C-related facts and propose a plan on how to get out
of the impasse. I realize ricardo is not hooked up to the blast_tof mailing list but I hope you have been
checking the flow of emails on blast talk (of course ricardo you can easily subscribe yourself)
* C detectors were gain matched (flasher) and a high efficiency for cosmics was measured in the DTF
* Now, for the same voltages, single C pmt's do not appear gain matched (when triggering on the flasher
pin diode). Furthermore now we can't use cosmics for the efficiency. For the same flasher
input, and the standard operating voltages, I measure 1.0 and 1.5 V signals for different tubes in
the same box. The sum from the An.Adder of all pmts in a box is a funny number (see below)
* Recent very clean ep runs with field on H2 (3 weeks ago) show a number between 60 -70 % for the
efficiency (nikos). Nikos is further studying this, for selected boxes, also as a function of cut
the goal is to reach a significant number using ideal elastic events, although this may not be the
true "average C efficiency"
* The analog adders in the D-tunnel were found to be largely non-linear. The sum out of many tubes is not
very different from a single tube signal. Note we expect about 1 photoelectron per tube.
In addition, the large tube "noise" events (several volts) pass the analog adders but fall back in the
range of good events (ca 0.5 V)
* With this adder performance and this set up (one box, one signal) we have no way to trigger on events for which more than one tube fired. Therefore the scalers readout are completely dependent on tube
noise. This has been seen on the scope to be up to a few KHz (note cosmics and flasher are few Hz)
* However, the total "cosmics" (beam off) rate can vary greatly (factors of 10-100) from one box to
another. How come tube noise is so different ? It may instead relate to operating voltages. We also
need to check for light leaks (this hasn't been done in the hall)
* tried several things but I still can't attenuate the flasher signal. Since the optical connection may
now be different (and provides a very fast slope) we can't relate the pulse height of a cosmic event
in the dtf to the pulse height of the flasher now.
So here is a plan. It requires hall open which should be the case this week.
_ Check for lights leaks. At any voltage. Look at scalers and scope with lights on/off. Note, lights
are always on in the hall.
_ Using the small paddles and a TOF counter, trigger on cosmic in the hall. Tune the pmt HV to have
300-400 mV. Requires some configuration changes (and ladders)
_ To save time, once a tube has been set for cosmics, we can look at its response for flasher and set
the other tubes in a box to that flasher pulse height. We don't really need analog attenuators since we
are interested in HV and not pulse height (this was only a concern for the caen analog adders)
_ Results found in this way be extrapolated in the next steps without relying heavily on assumptions about
PMT linearity since we are already in the good HV range. Of course different tubes set for cosmics (few
photoelectrons) may respond differently to the flasher signal (many more photons ?) but there is no
way around that.
_ Replace the analog adders with (truly passive) fan outs being built at bates. We should have the first
units ready by tuesday at the latest. This will allow splitting the TOP and BOTTOM signals. I suggest
moving the whole thing over to the (almost empty) racks with the FB crates
_ Using this new set up we can now trigger off meaningful events. In this way we can related the scaler
rate across different boxes
_ One tube at a time, the gain matching can be checked against adc spectra, trigger on the flasher (we may
need more attenuators if the flasher falls out of the ADC range).
_ Once we are "gain matched" we can go back to eff studies
_ At any convenient time for the Wch people, checking C timing to the trigger (build C trigger). Note that
Wch people only need "some" C efficiency.
________________________________________________________________________________
Tancredi Botto, phone: +1-617-253-9204 mobile: +1-978-490-4124
research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST