Primary Collimator Live Page
Tony's Status Report
Klaus's Status Report
Michael's Web Page
Qweak Simulation Forum
Keep Out Zones a la Allena
Previous Reports
Virginia Tech Update for Tracking Group,
Simulations are still under way for beamline/shielding/lumi/plug optimization. The major issue to be overcome is still the problem with the rates for the far forward scattering angles.
The scattered ep elastic rate (only primary electrons) over all origin theta is greater than the incident rate (Figure 1, black curve). We understand the basics of why this is so, but we have not yet developed a prescription of how to properly weight small angle events. We are working on it! For now the best we can do is give what we believe are upper and lower estimates on the ep elastic rates (or any backgrounds caused by them).
|
|
Figure 1 - A plot of theta at the scattering vertex for all elastic ep particles (black) and for those "hitting the plug" (green). |
There wouldn't be a problem if none of those low angle events made it to any of our detectors of interest. The small angle electrons that make it in are those that emit very energetic bremsstrahlung photons and are left with small (~ 100 - 300 MeV) energies. These electrons have a signficantly larger mean multiple scattering angle, so many of them get scattered out to large angles.The green curve shows the events that make it into the angular range subtended by the first plug.
There is no obvious solution to this, because GEANT will treat events at an angle of .5° pretty much the same way it treats events at .01° - so the suggestion of trying to divide by the number of scatters per event is probably not going to work. To try to clarify the issue, it is not scattering in GEANT which is the problem - it is that the cross section is so high the the probability of interacting in the target is greater than 1.
Probability of interaction = &int d&sigma/d&Omega d&Omega t NA/ A
where d&sigma/d&Omega is the differential cross section ( &mu b sr-1)
and t is the target thickness (g &mu b-1)
and NA/A is Avogadro's number over the atomic mass (#particles/g)
When we integrate this from .027° to 180°, we get 1 for our target.
To reiterate, the same problem comes into the point target estimates that are independent of the simulation. The rate calculation only gives you the number of particles/second if each particle only interacts once in the target.
We built an aluminum frame to support and ground the amplifier/discriminator cards. Figures 2 and 3 show this frame on the half-chamber. Figure 4 (Green trace) shows the oscillation on the fast OR output of a Nanometrics N-277-C amplifier card. By raising the externally applied threshold on the amplifier card to 2.2V we can get rid of the oscillation and see drift chamber signals. Figure 5 shows:
Yellow trace - Scintillator trigger signal
Green trace - Drift Chamber Fast-OR signal
Blue trace - Discriminated Drift Chamber signal.
When we raise the field wire voltage above about 1200V we get a trip (spark) in the drift chamber. We took it back to the clean-room to open it up and investigate.
|
|
|
Figure 2 - New aluminum supports for the cards.
|
Figure 3 - More stable grounds for the cards?
|
|
|
|
Figure 4 - The oscillations are still present (yellow - scintillator, green - chamber, blue - discriminator).
|
Figure 5 -
|
