From: Jonathan Fulcher [jonathan.fulcher@cern.ch]
Sent: 21 May 2004 14:23
To: Coughlan, JA (John)
Cc: Geoff Hall (E-mail)
Subject: CERN Fed Progress
Hi,

Here is a short progress report. Most time seems to be being spent in the LAB, hands on with the system debugging the DAQ readout in preparation for the beamtest. For this reason I have fallen a little behind on certain software tasks.

1. a number of improvements have been made to the readout code including speeding up of the event readout, and introduction of polling between fragments. This has been tested and works well.
   

    2. I have implemented an Abstract Base Class for the FED, the Fed9UVmeDevice and the Fed9UDescription now derive from this class, so we are now able to call the matching interface from a pointer to the base class, this enabled me to have an abstracted gui widget to set values in the fed either on the description or the fed real, seamlessly. It was one weeks work, but was deemed as a big enough advantage for future development.

    3. the Gui now enables the user to build descriptions and/or directly interact with the fed registers in real time. I am just in the process of moving the debug readout code from FedDebugSuite.exe into the gui. At this stage we will then have both a complete debug gui and a command line interface. These tools have proved invaluable during testing so far.

4. I have also migrated one of Matt’s classes, Fed9UCrateStatus, from the test bench to the Fed Software which probes the crate and produces an array of FedEntries which also contain pointers to the constructed Fed Objects. This has been integrated in the gui, making it a real simple matter to detect your feds in the crate in preparation for any debugging tasks you wish to perform

5. System semaphore has been added to the Fed9USupervisor to ensure atomicity of the readout. This has also been tested with the DAQ, and appears to work.

6. A lot of time has been spent last week trying to find the cause of a strange readout error that causes a random percentage of events to be corrupted. This slowed down progress and preparation for the beamtest, but some progress has been made today, and we believe that we may be approaching an explanation.

7. I took the liberty of setting up a savannah account and we have started to submit bugs and tasks to this, I would like to continue to use this as much as possible, so far it has proved to be a real help to me in managing the status of the project. If you have an internet connection there you may want to have a look at it. Perhaps greg may feel it a good idea to set up projects for the FedTester and future projects.

8. Costas has agreed to set out the S-Link cards soon, so we plan to test first in the Lab at CERN, before going to the test beam area.

9. The tests on the FEDs have revealed that 003 is still operation, despite it’s small issues with booting. And 014 has proved to be working fine. 013 is installed in Lyon and has been working fine reading out data in their lab.

10. Pisa reported that since I upgraded their software, they have been taking data with no errors for two weeks.

11. Installation scripts still haven’t been completed, this is due to lack of time with all the work in the lab. However, I plan to have a new stable release prepared by the middle of next week. This will be posted on the website and the new installation scripts will be posted with it. This should fix a number of the issues posted on savannah. Thanks to Ian and Rob for testing. Hopefully the next version release should contain all the temperature and voltage monitoring code and the Eprom stuff.

Plan is to move the setup to the beam area on Monday. Most of this can be done by Laurent and Patrice leaving me free to prepare the software release. As for software, I have a current stable release on afs, we will keep this, but initially move to the new version. If any problems are encountered we can revert with no delay. This would only mean a minimal loss of functionality.

I am in contact with the Crack and Lyon about their timing for FED requirements. Will organize this at the start of next week.