Latest revision as of 01:06, 1 April 2015

January 16, 2014 FDC meeting

Agenda

1. Installation [1],[2],[3],[4] (all)
2. Engineering (Bill)
   • Cooling system status
   • Gas system status
   • Other: FDC survey, cathode scan
3. Electronics (Chris, Nick)
4. Mini-DAQ results FDC E-log (Beni)
5. Other

Minutes

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

Installation

- The top two quadrants were cabled (see first three pictures attached); Mike Klopf was working till his last minute at JLab, to arrange the cables nicely and uniformly (first picture). Nick is almost done with routing the second quadrant to the electronics. Casey and Tina are working now on the bottom right quadrant and have cabled already two packages (second picture).

- Discussions about the routing of the LV cables. They all have to run to the left side (looking downstream) and it is not clear yet how they will be attached to the face of the magnet in their final position. Therefore, Tom prefers to disconnect them from the LV panel during insertion and then route them again to the panel. Fernando, Bill and other disagree; Beni showed on a drawing how to route the bottom half in order to avoid disconnection from the panel. Lubomir discussed if the extra length of the big bundle at the top right cable tray is enough for the FDC insertion (third picture).

- On Monday we changed the gas to 40/60 Ar/CO2 pre-mixed bottle. The flow was initially increased from 400 to 1000 sccpm (using another mass flow controller with 6000 sccpm range and another rotameter), but at this point the bubblers stopped bubbling: the same old story and we suspect again that somehow gas channels with high impedance are created, making connection from the bubble meniscus to the air. Finally the flow was set to 800 sccpm (200 sccpm per package) so that the bubblers can work.

- Since Wednesday nominal HV, +2100/-500V, has been set on all channels except one (fourth picture showing its current) with +2000/-500V and there were no trips for more than 24 hours. After the meeting that channel was also set to the nominal HV and is stable: all the HV channels are now operational at the nominal HV and gas mixture.

Engineering

- On Wednesday morning we found the chiller off with an active alarm. It turned out that there was disruption in the chilled water supply for several minutes that most likely caused the chiller to fail. The top half of the LV channels were on at that night which increased the Fluorinert temperature to 27degC (as measured with the chiller when it was re-started). Everybody agrees that we need a hardware shutoff of the LV crate in case the chiller fails and this is urgent. Details, how to do this, were discussed at another meeting organized by Hovanes right after the FDC meeting.

- Even if the chiller failed due to the water disruption, Bill wanted to make sure that the chiller itself didn't contribute to that. Dave measured the temperature difference of the water in/out of the chiller and it was 31degF. According to Bull this means we have ~0.5 GPM water flow, while it is recommended to be 1.3 GPM to have smaller temperature difference. Bill is looking for a flow meter to measure the actual water flow.

- Gas system: Bill sent to Brooks one of each: mass flow controller, control unit and cable. Certainly there's a problem, most likely in the control unit. Other than that, most of the gas stuff is ready, still needs to be tested. So when we fix the flow controllers, the system can be used in principle with a manual control and using one bottle (of each, Ar and CO2) at a time. Because of the high flow we use now one bottle per week. According to Dave we expect the next gas delivery for Feb. 5 and we have a little more than one bottle in the moment. Again, we need to find a temporary solution till Feb. 5, either reducing the flow or flushing with Nitrogen and postponing the tests.

- Bill talked to Jim Dahlberg, but there are no results yet from the last survey!?

Electronics

- Chris discussed again with Bill how to connect the ground cable. The problem is the cable is very thick with a big lug at the end that doesn't fit well on the ground plate attached to the packages.

- Fernando proposed to use techs from Fast Electronics, since we lost three people working on the cabling. Eugene agreed.

- Fernando: 13% of the fADC125s have problems in the moment. It means we can fully equip the FDC with modules which will happen soon.

- While testing the signals from the FDC we found several connectors at the detector side with broken locks which Chris fixed immediately, but also two bad connectors at the electronics side. The latter are more difficult to be fixed and Chris wants to do all the bad connectors at the same time when we test all the cables. Three more cables have one or two bad channels due to the conductive tape problems. In some cases one can fix them by re-seating the connector and bending the cooling tube, however sometimes this introduces new bad channels even in the neighboring cards. Therefore for the moment we will not attempt to fix such bad channels if they are only a few.

Mini-DAQ

- Beni presented first results from the mini-DAQ (E-log entry 123). If reading 9 fADC125 per crate the rate is ~6Hz defined by the code running on the ROC. It checks if there's something above a threshold, then all samples are written to a file. Since there's no external trigger, the probability to get a hit in the 800ns window is of the order of 10^-6; practically one can get on average 10 hits per channel for 24 hours, however for some channels (short strips) one gets only a few hits (second plot in the E-log). Beni showed also the hit multiplicity (third plot), indirectly indicating that we are looking at signals, not noise.

- Beni now has two crates with 9 fADCs each, that can be used for testing. The rest of the fADCs don't have firmware; Cody is working on loading the firmware via VME and when this is ready we can use all the fADC125s. Beni also got 6 ROCs so we can have all of them running at the same time. Practically such tests will be used to find bad channels, but testing the bad channels will better be done with a scope. Fernando promised to build a testing board, adapter between the cable connector and the differential probe.