Minutes-5-23-2013

From GlueXWiki
Jump to: navigation, search

May 23, 2013 FDC meeting

Agenda

  1. Production Construction Tracking
    • Spare package status
  2. Update on the corrosion test at MePhi results (Lubomir)
  3. Engineering (Bill)
    • Cooling system status and tests
  4. Electronics (Chris, Nick)
  5. Full electronics tests FDC E-log (Beni)
  6. Other

Minutes

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

Production

- Spare package: On Friday all the channels of cell #2 were tested and worked fine. Over the weekend the cell tripped from time to time. After we identified one HV sector (28 wires on the side of big wire numbers) that trips, we opened the package for a third time. In addition we noticed that the dark current goes down when removing one of the pre-amps in the same area at the signal side of the wires. There we found one suspicious soldering that was fixed. At the other side of the wires we found a small epoxy ball that was removed. After some additional cleaning of the wires in this area we closed the package again. It is being flushed with gas since Tuesday and will be checked on Friday.

Update on the corrosion test at MePhi

- Vlad finished the test with 6 EPDM samples at 100degC at 100% humidity. The results are similar to what he got before with two samples: all the samples died between 6 and 24 hours. Now with higher statistics he estimates the activation energy to be 0.35+/-0.07 eV (consistent with the literature). Assuming the same chemical reaction for the Viton corrosion (just that Viton contains less amount of Sulfur) we estimate 1055 +/- 211 days live time at room temperature, but at 100% humidity. We don't know how to extrapolate the humidity down to 50%. It would be good to place O-ring samples outside of the packages to be tested whenever the FDC is taken out from the magnet for some maintenance.

Engineering

- Bill, Casey and Beni performed the test of one loop of the cooling system. 22 pre-amps were attached to the cooling tube and powered with LV. The flow (at the right pressure) appeared to be in a good agreement with the ANSI model, however the temperature drop between the end and the beginning of the loop was 1.2deg instead 2deg as expected from the model. We discussed possible reasons for that: the exact power on the cards (we measure the current only, but the voltage on the card was not measured), the hot elements on the card, the regulator and the ASICs, are not in equilibrium for the short time of the experiment (about 10-15 min), bad thermal contact at the pre-amp clamp, or at the place where the thermo-couple is attached. The most important however is what is the temperature at the voltage regulator and ASICs when the cooling is on. After the meeting Beni and Bill repeated the test and the max temperature on one of the card at the voltage regulator was measured to be 42degC. Note that this card was connected with an individual LV cable. Also the ends of the thermo-couples were clamped together to check their calibration: no more 0.2deg difference was found. In any case, the test demonstrates that we can cool the pre-amps enough to keep the temperature of the hot elements well below the their limit.

- The current cooling set-up (SMC chiller + 2kW water chiller) can ran for no longer than 15 min. If you cool down the water in advance (say to 3degC) it can ran probably for not more than an hour. Thus the only way to do the test with the full electronics will be to use the water chiller directly to cool the pre-amps.

- After the meeting Bill got response from the SMC experts: the chiller produces 2kW from the pump + 2kW from the compressor + max 2kW cooling power. So it requires 6kW water cooling. The water chiller we used in 126 is 2kW that's why it didn't work, but we will have no problems at the Hall.

Electronics

- Nick and Beni once more identified the problem with the CAEN HV module; it is there even without a load. Nick will contact CAEN.

Full electronics tests

- Beni for the firmware/software for fADC125: working progress. If it's not ready by the end of next week we will read out only the old modules.

- The truck was not available this week, so we will move the package next week. As discussed above, we will use water for cooling.

Other

- Vlad is arriving on June 2 for six months.