From GlueXWiki
Jump to: navigation, search

July 11, 2013 FDC meeting


  1. Production
  2. Engineering (Bill)
  3. Electronics (Chris, Nick)
  4. Full electronics test FDC E-log (Beni)
    • Pre-amp contact issues (Vlad)
  5. Thick cell results [8],[9],[10], [11](Lubomir, Vlad)
  6. Other


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


- Final package installation (see the pictures linked above): First, Casey found a tiny hole in a cooling tube and he successfully fixed it with solder. We tried to check the cooling system for leakages using CO2 at 100psi, but the leak detector was not sensitive at all to CO2 even when you have obvious leakage. The first two packages were put on the installation cart and the spacer between them installed. The spacer fits tightly on the gusset ring and it was especially difficult to insert it in the second package. For the next spacer Bill proposed to make several holes, so that one can pull and make the circular shape at the end of the spacer in order to insert it into the gusset. We also had to do several cuts at the end to make space for the fiducial holders. Finally, it turned out there was a gap at the top between the spacer and the gusset ring, meaning the two packages were not exactly parallel. Bill tried to use the adjustment on the holding brackets to fix the problem, but it didn't work. Therefore, we inserted an aluminum threaded rod (in addition to the three fiberglass rods) on the top (see last picture), to tighten the two packages together and the gap was fixed. It is possible by this we may have introduced some frame deformation, therefore Bill suggested to survey the packages after the final assembly at Blue Crab to make sure the packages are parallel and there's no deformation.


- We discussed what else is needed to finish the package installation. We will install plastic manifolds for the gas (instead of the aluminum ones that we have now), and will make also the connections between the cooling manifolds. Important question about the grounding connections. Important result of the test with the full electronics is that we don't need additional ground connections. According some experts, the connection of the detector to the Hall ground is needed only for safety (in 126 there's no such connection and the noise is low), therefore, we can have one ground cable for all the packages and it doesn't have to be thick.

- We discussed if we can use 1/4" tubes for the gas inside the solenoid. These tubes will fit in the gas manifold. Outside tubes are 1/4".


- Nick will check how many spare pre-amps we have to be used with the new F1TDCs in the spare package.

Full electronics tests

- Beni explained his latest results (FDC Elog 115): after fixing the trigger cable to the first crate (fADCs cells 1 and 2) and changing the algorithm for applying thresholds, now the strip efficiencies are ~98% (note also we are at 2200V). Strip problems: one bad connector/daughter card/conductive tape in the middle of cell 5 with about 10 channels not working, several channels with problems outside of the detector, some of them in the fADC, and also few channels showing problems from time to time. Wires: one translator board not working (half of a cell), identified by swapping the cables. Beni discussed also the slopes of the wires as reconstructed by the strips. On some of the planes there are systematical deviations up to 1mm over 1m (Bill: difficult to attribute this to some mechanical problems), but one can correct for these deviations offline.

- Vlad also did a lot of work fixing the strip channels: some were fixed just by touching the cards, some by re-seating the cards, and some by using canned air and contact cleaner. It is not clear where exactly the problems were; it could be that by using canned air and cleaner, we just cool the connector and due to the thermal conductivity also the conductive tape.

- We discussed different options what to the with the bad place in cell 5: open the package and fix the connector (Eugene didn't support this option), try to fix it without opening by inserting thin plate in between the conductive tape and the wire frame which is against the cathode frame, or do nothing (very small percentage of all the channels). After the meeting Lubomir and Bill inserted a thin g10 plate 2" wide to cover the whole conductive tape and we immediately saw an improvement, only 2 problematic channels. On the next day, by playing carefully with the plate we fixed all the channels of this card. Casey cut the plate so that it doesn't stick out of the cathode and we will leave it there permanently. Thus, the plan will be to take data at three different HVs over the weekend and move the third package back to Blue Crab on Monday.

Thick cell results

- Lubomir and Vlad collected a lot of data files (about 6k) at the end of last week with waveforms from the scope at different HV: 2450,2500,2525,2550, and 2575V (the maximum at which we could operate the two cells). For the new measurements we doubled the flow and the oxygen contamination went down from 100 to 60ppm (further reduction is possible if we filter the gas from the bottle). Thus, the reduction of the number of electrons with the time was significantly smaller: see first picture attached, black at 100ppm all the rest at 60-65ppm for different HV values. Interestingly, at electron recombination depends also on the HV.

- The response of the chamber on a single electron arriving at the wire is shown in the second plot. Basically, it represents the distribution of the gas gain (as seen by the pre-amp in Volts) for different HVs. In some of the measurements the noise was higher which explains the double-peak structure, but the fact that we see the extremum of the signals that must be present at high gain means we can study the so called Fano factor. These distributions are important for gas detectors that want to detect individual electrons (like Cherenkov detectors) and we think the results have to be published.

- The last two plots show the cluster counting: one cell vs the other and their difference. Each cell counts about 50 clusters per event and the resolution as derived from the difference is ~7 which corresponds to the above number. Thus, just with the two cells that we already have, we can reach 10% resolution to be compared to ~15% difference in the cluster number for K and pi in the relativistic rise.