Difference between revisions of "Minutes-10-23-2014"

From GlueXWiki
Jump to: navigation, search
Line 17: Line 17:
 
# Commissioning [https://hdops.jlab.org/wiki/index.php/Hall_D_Commissioning Wiki Page] (all)
 
# Commissioning [https://hdops.jlab.org/wiki/index.php/Hall_D_Commissioning Wiki Page] (all)
 
# Other
 
# Other
 
<!--
 
  
 
= Minutes =
 
= Minutes =
  
Participants: Curtis, Naomi (CMU), Beni, Dave, Mike, Eugene, Chris, Nick, Simon, Fernando, Vlad, Cody, Sergey, and Lubomir (JLab).
+
Participants: Curtis, Naomi (CMU) Mike, Fernando, Chris, Eugene, Tim, Dave, Beni, Nick, Simon, Vlad, Cody, Sergey, and Lubomir (JLab).
 
.  
 
.  
 
== DAQ and electronics ==
 
== DAQ and electronics ==
  
- Info from Beni and Cody: JTAG cable was connected (as proposed by Sergey) to the last module of crate FDC9 and chipscope installed on one of the FPGAs (the processor one). Thus, they were able to capture the problem and analyze the signals when the module didn't report data. So far, one thing is sure, the module does receive a trigger but at some point one FPGA doesn't transmit data to the next one. The plan is now to install chipscope on all FPGAs. After the problem appears even the reset doesn't help indicating is a hardware problem.  
+
- Cody generated a new version of the firmware that is much more stable, as Beni tested it over the last night. With the new firmware Sasha was also able to have the CDC crates running at >1kHz without crushes for some time (before, they crushed at about 300Hz). However, with the new firmware, some new modules started showing problems, as tested by Beni and Sergey. A chip-scope was installed on the last module of ROCFDC5, but turns out other modules were crushing the DAQ now.  
  
- After the meeting Sergey and Cody informed me that the problem is that the 80MHz clock sometimes doesn't come to one of the FPGAs, indicating most likely bad soldering that changes with the temperature. The plan now is to reduce the board frequency to 40MHz.    
+
- Fernando explained the changes in the firmware: in the previous version, Cody let the CAD software do the planing of the changes (in particular going from 16 to 72 channels) to the f250 firmware. As a result, some of the clocks were not routed in an uniform way. Now Cody manually required a separate (clean) area to be used for the clock distribution.  
  
- Nick: One CAEN HV module in the CDC has been replaced. In the moment we have 3 spares. The hardware limits for the modules should be written somewhere on wiki, to be used when replacing modules (actually they are recorded in the archiver).  
+
- Cody: the new bad modules are most likely related to a different problem, could be a hardware problem. We are not confident in the quality of the boards (Fernando). The first article f125 from the new vendor will arrive next week.
  
- Dave is implementing reading temperature and humidity in the Hall, also transmitting the FDC pressure differences as a new type epics variable.  
+
- Eugene: For the first days of running we should simply exclude the bad modules from the DAQ till we find (if any) a stable configuration that runs for at least an hour without crushes. Then, during the several days of down time we may tray to fix/replace the bad modules. At least for the first days, the main priority will be to have the DAQ working in any configuration (even without the chambers!). Then we can think of how to test the detectors, which may be postponed till December.
  
- The gas panel in the Hall was connected to UPS.
+
- Bryan Moffit is working on a software reset of the module if it crushes, so the you re-sync the events and can use the information from the module further without stopping the DAQ.
  
 
== CDC update ==
 
== CDC update ==
 +
 +
<!--
  
 
- Mike is working on the efficiency plots, also charge vs track length, to be used for gain calibration.  
 
- Mike is working on the efficiency plots, also charge vs track length, to be used for gain calibration.  

Revision as of 15:23, 23 October 2014

October 23, 2014 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. DAQ (Beni)
    • f125 status
  2. Electronics (Fernando)
  3. CDC update (Mike)
  4. FDC update (Vlad, Lubomir)
  5. Commissioning Wiki Page (all)
  6. Other

Minutes

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

DAQ and electronics

- Cody generated a new version of the firmware that is much more stable, as Beni tested it over the last night. With the new firmware Sasha was also able to have the CDC crates running at >1kHz without crushes for some time (before, they crushed at about 300Hz). However, with the new firmware, some new modules started showing problems, as tested by Beni and Sergey. A chip-scope was installed on the last module of ROCFDC5, but turns out other modules were crushing the DAQ now.

- Fernando explained the changes in the firmware: in the previous version, Cody let the CAD software do the planing of the changes (in particular going from 16 to 72 channels) to the f250 firmware. As a result, some of the clocks were not routed in an uniform way. Now Cody manually required a separate (clean) area to be used for the clock distribution.

- Cody: the new bad modules are most likely related to a different problem, could be a hardware problem. We are not confident in the quality of the boards (Fernando). The first article f125 from the new vendor will arrive next week.

- Eugene: For the first days of running we should simply exclude the bad modules from the DAQ till we find (if any) a stable configuration that runs for at least an hour without crushes. Then, during the several days of down time we may tray to fix/replace the bad modules. At least for the first days, the main priority will be to have the DAQ working in any configuration (even without the chambers!). Then we can think of how to test the detectors, which may be postponed till December.

- Bryan Moffit is working on a software reset of the module if it crushes, so the you re-sync the events and can use the information from the module further without stopping the DAQ.

CDC update