Difference between revisions of "Minutes-3-5-2015"

From GlueXWiki
Jump to: navigation, search
(→‎FDC update)
 
(4 intermediate revisions by the same user not shown)
Line 17: Line 17:
 
# Engineering (Dave)
 
# Engineering (Dave)
 
# Other
 
# Other
−
 
−
<!--
 
  
 
= Minutes =
 
= Minutes =
  
−
Participants: Naomi (CMU) Eugene, Tim, Dave, Fernando, Mike, Chris, Nick, Simon, Beni, and Lubomir (JLab).
+
Participants: Curtis, Naomi (CMU) Eugene, Fernando, Dave, Mike, Chris, Nick, Simon, Beni, and Lubomir (JLab).
  
 
== Gas issues ==
 
== Gas issues ==
  
−
- We had very long discussions what to do with the CDC leakage. Everybody agrees that CDC can't be operated like this and the leak has to be fixed. Different proposals: start removing the HVB cards and check with the sniffer the holes at the cheese plate (Beni, Nick, Fernando), check for leaks at the downstream mylar window with a tube on the sniffer (Dave), take some cosmic data overnight, if the leak is at the cheese plate the straws will still have some gas flow and the data should be OK (Lubomir).
+
- CO2 was monitored for two weeks. Two sets of measurements are shown in the above plots. First plot: starting with certified gas (60% CO2) in EEL126 then moving the sensor (without disconnecting from the laptop) to the Hall and sensing first FDC1 for several days and at the end switching the solenoid valve to the CDC. Shown are the atmospheric pressure, sensor results and the ones corrected for the pressure: ~(dleta_p/p)^n, where the power n was chosen, on the first plot, to make the calibration measurements flat (n=2.4). Second plot is the same except n=1.4 which makes the CDC measurements plat. Third and fourth plots represent the second set of measurements, starting with 5 days sensing CDC and then few days FDC1.  
  
−
- By the time I'm writing the minutes, Nick removed all the cables at the front and a portion of the HVB cards at the bottom. Thus, Mike and Nick were able to sniff the entire cheese plate without finding anything localized. The place with the maximum beeping is at the bottom, between BCAL and CDC, but the gas is heavier than air so it may come from anywhere. On Friday afternoon Tim and Eugene decided to pull the CDC out of the bore. Target people are preparing the target. On Monday morning it will be craned to the floor and the CDC will be pulled out.  
+
- Discussions:generally FDC is off by 4-5% consistent with the time-to-distance results. CDC is off by 8-9-10% while the time-to-distance from the data requires ~6% off (or ~44%). Proposal: to calibrate the sensor with a gas that is closer to the 40% CO2 as measured in the CDC. Need more calibration measurements to be sure the pressure correction is correct. If assume n=1.4 CDC is more stable then the FDC, could be due to the higher flow.  
  
−
- The CO2 sensor came back from repair. We calibrated it with 60/40 CO2/Ar certified mixture in the EEL126, and then started monitoring the CO2 in the first FDC package. Results are linked: some variations for 20 hours, maybe related to the gas temperature, has to be monitored for longer time.
+
- Gas gain calibrations: using Fe source measurements were done in the EEL126 with the spare package at three different HV (first plot). They will be compared to the results from the Hall with the source put at the downstream side of FDC. Second plot shows some change of the gas gain with the atmospheric pressure. Since the gas properties scale with field/pressure, it goes in the right direction.
  
−
== Electronics ==
+
== CDC update ==
  
−
- Fernando: the higher CDC noise (~60mV compared to ~30mV before) related to the loop on HV cables we had before that acted as an inductance. Fernando tested the new HV filter from CAEN on one cable and it worked great: noise <20mV (with or without loop). This was only a prototype box, now we need 13 more to put on all the cables. Fernando agreed that if we remove the CDC, the HV cables can be re-routed with a loop.  
+
- CDC was left without alcohol for a week - very small change (~1%) in the drift-time spectrum.
  
−
-  Cody: testing the f125 with the latest firmware at high rates in block-mode. In mode 8 one can reach 5kHz, in mode 1 up to 20kHz, and in mode 7 up to 50kHz -  these limits correspond to 100% occupancy.
+
== FDC update ==
  
−
== Engineering ==
+
- The Chiller came back from the repair and was connected to the FDC, so far works fine.
  
−
- Dave: CDC flow sensor is ready.
+
- One HV channel tripped, took several days to set it back.
  
−
== CDC update ==
+
== Electronics ==
  
−
- Mike showed a plot linked above, with the resolutions (sigma of the residuals) vs drift time. The residuals were obtained from tracking in magnetic field (somehow the track quality is poorer without field) after Simon's alignment. The resolutions are very good, ~130 microns for >100ns. The poor resolution for <100ns can be explained by poor timing resolution.  
+
- Beni: new f125 firmware on all modules. CDC was read out for 24 hours at 30kHz without crashing.  
  
−
== FDC update ==
+
== Engineering ==
−
 
+
−
- We looked again at the structures seen in the residuals: difference between the cathode reconstructed and the nominal wire position,
+
−
discussing once more (now with Tim) the possibility for wire offsets. The plots above show both the mean and the R.M.S. values of the residuals as function of x,y and their projections on x (last plot).
+
−
 
+
−
- FDC efficiencies for reconstructing 3D hit point in each cell (package 3 shown): it is required to have hits in the two cathodes and the wire plane, the hits to have timing withing 2 samples cathode-to-cathode, and within 48ns wire-to-cathode, as well as the three coordinates to cross at the same point within +/-5mm. The efficiencies are estimated with tracking without magnetic field and shown for 4 different thresholds: 50,100,150,and 200 fADC units. The plateau is at ~95% efficiency. There are places with poorer efficiency due to bad strip/wire channels. In each package, there's also a spot with poorer efficiency which is always at the wire connector side. This maybe due to the timing constraint without taking into account the signal propagation time.
+
  
−
-->
+
- Dave: CDC flow sensor connected and operational. FDC chiller shut off also connected and tested.

Latest revision as of 18:18, 9 March 2015

March 5, 2015 FDC+CDC meeting

Connection

  1. Instructions for Bluejeans meeting connection
  2. FDC meeting ID: 290664653
  3. To join via a Web Browser, go to the page [1] https://bluejeans.com/290664653.

Agenda

  1. Gas issues
  2. CDC update (Mike)
  3. FDC update (Lubomir)
  4. Electronics (Fernando, Cody, Chris, Nick)
  5. Engineering (Dave)
  6. Other

Minutes

Participants: Curtis, Naomi (CMU) Eugene, Fernando, Dave, Mike, Chris, Nick, Simon, Beni, and Lubomir (JLab).

Gas issues

- CO2 was monitored for two weeks. Two sets of measurements are shown in the above plots. First plot: starting with certified gas (60% CO2) in EEL126 then moving the sensor (without disconnecting from the laptop) to the Hall and sensing first FDC1 for several days and at the end switching the solenoid valve to the CDC. Shown are the atmospheric pressure, sensor results and the ones corrected for the pressure: ~(dleta_p/p)^n, where the power n was chosen, on the first plot, to make the calibration measurements flat (n=2.4). Second plot is the same except n=1.4 which makes the CDC measurements plat. Third and fourth plots represent the second set of measurements, starting with 5 days sensing CDC and then few days FDC1.

- Discussions:generally FDC is off by 4-5% consistent with the time-to-distance results. CDC is off by 8-9-10% while the time-to-distance from the data requires ~6% off (or ~44%). Proposal: to calibrate the sensor with a gas that is closer to the 40% CO2 as measured in the CDC. Need more calibration measurements to be sure the pressure correction is correct. If assume n=1.4 CDC is more stable then the FDC, could be due to the higher flow.

- Gas gain calibrations: using Fe source measurements were done in the EEL126 with the spare package at three different HV (first plot). They will be compared to the results from the Hall with the source put at the downstream side of FDC. Second plot shows some change of the gas gain with the atmospheric pressure. Since the gas properties scale with field/pressure, it goes in the right direction.

CDC update

- CDC was left without alcohol for a week - very small change (~1%) in the drift-time spectrum.

FDC update

- The Chiller came back from the repair and was connected to the FDC, so far works fine.

- One HV channel tripped, took several days to set it back.

Electronics

- Beni: new f125 firmware on all modules. CDC was read out for 24 hours at 30kHz without crashing.

Engineering

- Dave: CDC flow sensor connected and operational. FDC chiller shut off also connected and tested.

Retrieved from "https://halldweb.jlab.org/wiki/index.php?title=Minutes-3-5-2015&oldid=64069"