Present
Joel, Tivadar, Tuan, Pippo, Olivier
Many detectors are willing to start tests at the same time
- TPC :
- several issues were found, but not reproduced in the lab, after reloading firmware it was OK
- DCS : Test with differents boards
- MCH : Will have a PRR review next month, so they need to collect data
- MID : They don't have a CRU, they work with GRORC, mainly on DCS.
- ITS :
10G PON
(Mazsi)
- On leave
- Merge to be done
GBT
(Erno)
- Standalone design was rebuild with the new GIT repository
- Busy with debugging Calcutta board
- Try to fix the external reference PLL clock for V1. On V2 it will be a different PLL anyway, but we need to understand the hardware issues (for general understanding)
- We don't aim to use external jitter cleaning on the V1
- Tivadar : The use of the external clock system is not yet decided for V2. If internal PLL is fine, we won't use it. But we want to check what is the best quality we can achieve. For that the PLL should work with a 240 MHz recovered clock instead of using the 40 MHz.
- Please remove GBT core library
DCS
(Pippo)
- Mainkly working with sub-detectors, no news.
PCIe
(Pippo)
- QSYS could not be automatically generated by the make command.
- A temporary fix was to commit the generated files
- Now, it seems that using qsys-generate manually before typing make works and solves the issue. Makefile will be updated to add a make ip
- A problem was discovered with a CRU (low performances)
- That means that working CRU exists, but that they are not performant enough! The cause of the problem is not clear hardware or firmware. Seems to be correlated with power module.
- We have to check if there are many retransmitted packet
- The TPC card has to be borrowed to run the same test and confirm that is correlated to old power modules. If it is they will be changed.
Datapath wrapper
(Olivier)
- Some documentation will be commited
Integration
(Olivier, all)
- Preint was integrated with the other => now a single repo for the whole project
- PCIe modification to use two reference clocks -> to be done latter on
- pcie40-v2 firmware requires two issues to be solved:
- TTC-PON modification (to drive 2 transceivers) → Mazsi? At least after merging last modifications
- pcie40-v2 compiles up to synthesis, issues arise in fitter stage (see text file attached)
Simulation
(Olivier)
Production
(Tivadar)
- First cards from India were tested
- Card #2 has a single issue left (PLL work well), GBT #13 does not work. The likely defective is a borken capacitor. If correct, it will be sent to CERN.
- One more new power module is available.
- Another card form india is in shipping, should arrive soon.
Travels/vacations
Feb 22-23 : Olivier on leave
There are minutes attached to this event.
Show them.