Eugene,
Ok, I think I misunderstood your meaning. You are saying that the
full energy bite of the microscope corresponds to +/- 2 sigma of
the typical final state energy resolution in GlueX. That tells me that
the tracking+calo resolution is just good enough to say whether or not
the beam photon was somewhere in the microscope, but not good enough to
narrow it down much further within that range. That appears to be
true, although there will probably be some accidentals reduction coming
from a chi-square cut on energy conservation. I guess I was focused on
trying to help clear up the questions raised by David about how the
different numbers for the tagger energy resolution fit together into a
coherent story. I hope that story is clearer for him at this point
- GlueX physics requires < 60 MeV rms resolution in the tagger
microscope for reconstructing exclusive final states.
- Segmentation for rate purposes requires individual counters
subtend < 10 MeV full width in the microscope.
- Intrinsic tagger resolution limits from electron beam energy
spread and magnetic optics are around 3 MeV rms in the region of the
microscope.
- Choice of 2mm fibers leads to 8 MeV full-width energy channels,
consistent with the rate requirements and exploiting the full
resolution of the magnet + beam for a final instrumental RMS energy
resolution around 5 MeV rms.
Richard Jones
Eugene Chudakov wrote:
Pine.LNX.4.64.0905202242321.5025@jlabl1.jlab.org">
Richard,
correct me if I am wrong. I calculated the full event energy resolution (RMS) of 200MeV. I used PYTHIA events at E>8 GeV with all the tracks reconstructable, assumed that nothing
is missing, and applied the parametric MC resolutions. Now, one should look for the matching signal in the tagger in an energy window at least 4 times wider, or 800 MeV.
Eugene
On Wed, 20 May 2009, Richard Jones wrote:
Eugene Chudakov wrote:
One can look from another angle. The spectrometer can measure the full energy of the event, assuming nothing is missing. I estimated the RMS of about 0.2GeV for some class of events (any comments?). Taking +/-2 sigmas gives 0.8GeV of the energy interval. This is about the "microscope" width.
There seems to be an order of magnitude left out here? For a reconstructed event resolution of 0.2 GeV (200 MeV) and a reduction factor of 2.5 sigmas, I get 80 MeV -- assuming that your 0.8GeV was a typo. Your estimate comes close to the 60 MeV that we say is required for optimal reconstruction of exclusive final states with GlueX. That is not close to the microscope width, however, which is down another order of magnitude at 8 MeV.
At the full beam the microscope counts about 300MHz (3 times more than the photon beam in this range, because of the collimator). Assuming the ideal 2ns timing resolution, one gets 60% of the accidentals in the microscope. As far as I understand, this is considered acceptable (treated as a background?). Above 9 GeV the tagger rate per GeV is a factor of 2 lower than at 9 GeV. One may expect a 30% rate of accidentals in the right energy range. As far as I understand, the problem with the fixed array is not the accidentals, but the rate of a single counter, which may be too high for a PMT at the gain required.
Agreed.
-Richard Jones
|