From GlueXWiki
Jump to: navigation, search

April 5, 2012 FDC meeting


  1. Production Construction Tracking (Dave)
    • New procedures for the O-ring replacements
    • Production status
  2. Results from the oxygen studies (Lubomir)
  3. Engineering update (Bill)
  4. Electronics update (Chris)
  5. Possibilities for cluster counting (Lubomir)
  6. Other


Participants: Bill, Dave, Chris, Nick, Mark, Simon, Beni, Eugene, and Lubomir.


- Dave: Chris put components on two wire planes for package #3, they remain to be deadened. Anatoly is gluing daughter cards on cathodes for package #4. One end window for package #4 that was recently made has a wavy mylar surface. We suspect the surface of the transfer ring was slippery, since before it was used to tension clear mylar for the gusset ring. The rule will be always to sand the transfer ring before using for the end windows. We will have to redo the end window: first sand the old mylar and tension it again. Tomorrow, after the first cell of package #3 is assembled, we will resume stringing wire boards; stringing was suspended for several weeks due to the oxygen problem.

- On Monday we disassembled the third package. The old O-rings of the two end windows and the wire frame for the first cell were removed using aceton to clean the super glue. Casey invented a cover for the wires to prevent aceton spills. Then the grooves were double coated with Hysol, Viton O-rings were installed today and we assembled the first cell right after the meeting. Dave wrote the procedures for the replacement of the O-rings, waiting now for signatures. The depth of the groove after coating: according to Bill the depth should not be reduced by more than 10 mils; Dave measured it: 1-5 mils.

- Two or three type-1 frames have to be "converted" into type-2. Bill explained how to do this: put Scotchweld in the groove, use razor blade to level it and after curing possibly sand it.

Results from the oxygen studies

- Lubomir started working on a document describing the oxygen tests. The most important table with the measurements using the testing chamber is linked above. Important step from configuration #7 to #8: only the G10 groove was coated resulting in oxygen change from 1400 ppm to 280 ppm. To estimate the change factor we have to subtract the oxygen in the supply line. Beni measured it before using T-connector right at the chamber inlet: 200ppm. Lubomir measured it again by directly plugging the supply line into the sensor using the same tubing: 50ppm. The difference might be due to contamination from the chamber itself when using T-connector, or due to the higher pressure that is created when you plug directly the supply.

- Eugene: what's the acceptable oxygen level; must be <0.05%, but in the prototype we had ~0.1% and was still operational. Assuming we will have ~300ppm per cell at 200cc/min then we will have 1800ppm for the whole package, but if operating at 400cc/min for the whole package it will be 900ppm or 0.09%. However, for this estimation we have to subtract the contribution from the oxygen in the supply gas. More realistic estimation can be done at Blue Crab with the real package.

- Results of the oxygen studies: contamination is an effect of a combination of the O-ring type and the surfaces around the O-ring. When g10 groove coated there was no difference which O-ring used, but with the plastic spacer EPDM showed higher oxygen. Viton and Buna-N showed no difference; Viton was chosen for the production due to reports of outgassing for Buna-N. We want to continue the studies to understand the mechanism of the oxygen permeation, but later when we have time.

- Oxygen sensor system was moved to Blue Crab and the four cells of the package #3 were tested at about the same flow of 200cc/min (measured with rotarmeter): ~7000ppm, which is in agreement with ~10000ppm for the whole package #1. Bill proposed and Beni agreed to move the new gas system (Brooks) to Blue Crab to better control the flow, while using the old MKS at JLab.


- Bill is finalizing the fiducial scheme: will have 4 fiducial holders on each package visible at all times. Difficulties finding non-magnetic low-mass materials. Eugene and Beni had different suggestions especially concerning the z-position determination. Bill will coordinate his plan with the survey group. Bill is working also on the spacer between the packages, decided to have 3 supporting rods instead of 4, related to the fiducial scheme and other issues.


- Chris: there's a plan how to do the grounding of the packages. It's time now to work on this: Chris and Fernando will start testing the grounding on Monday. The first package which will be available for testing in 126 till the third package is moved there, in about a month and a half.

Possibilities for cluster counting

- There will be a workshop on the GlueX detector upgrade, especially on PID detectors. Lubomir presented an idea to use cluster counting technique for PID with an FDC-like detector filled with He-based gas mixture (slides are attached above). He gas is needed for low cluster density and low drift velocity to match the fADC125 and fADC250 bandwidths. On the other hand, to have enough statistics a wide (4cm) cell with ~3cm homogeneous filed is proposed. Full simulations of the signals taking into account also the electronics response functions, were performed using Garfield code (v.9). Assuming two packages, each with 8 cells (placed in between FDC1, FDC2, and FDC3 packages), one can separate pi from K in the relativistic rise with ~2 sigmas if using fADC125 and ~2.5 sigmas with fADC250. Advantages discussed: higher angular acceptance due to proximity to the target, lower mass (?) and additional tracking may improve momentum resolution, can be built using FDC design and materials with some modifications. Disadvantages: lower separating power than Cherenkov or TOF (can be used complementary to them), requires single electron sensitivity, He leakage (Eugene: can be blown away from the TOF and FCAL), additional cables (Bill: will be a big problem). Beni: huge drift times up to 6-7 microsec; Lubomir: that's why we have strips that will see the same specific cluster pattern and will disentangle events overlapping on the same wire. Eugene pointed to several experiments that have used cluster counting technique.