Choose timezone
Your profile timezone:
Powered by Indico
v3.3.5-pre
present: Paulo, Franois Vasey, Ken Wyllie, Lauri, Jan Troska, Robert, Torsten, Christian, Alex
*) Q: are all 5 SAMPAs loosing sync at the same time? —> monitoring not precise enough
*) monitoring of frame in GBT Frame was not available and will be available in Q1 23
*) modify loss of lock cnt in GBTx rx PLL => set to low value to see whether rx pll error count goes up —>
Assumption is that rx clock looses lock, but is not seen as the loss of lock state machine requires a programmable minimum of cycles before loss is reported in the counter.
Changes the loss of lock threshold does not solve the problem but would confirm whether rx clock is going down.
*) time behaviour of Sampa 0 to 4 and its frame in short tern scale -> Today TPC monitoring is not precise enough to see whether all 5 SAMPAs stop at the same time or one after the other. This would allow to see whether GBTx0 looses lock before GBTx1, given that GBTx1 receives the clock from GBTx0
(Alex note after meeting: Is this really true? GBTx0 rx clock is forwarded to the GBTx0 tx PLL (within the same chip) and to GBTx1 tx PLL to the 2nd GBTx on the boards. Would there not only be the transmission delay be the difference?)
*) paper exists with SEU cross section —> TPC should lose a link in one hour compared to 1 minute as in ALICE
*) is the regulator rad hard? —> needs to be verified
*) if you pull out the fiber or put in an attenuated will be sampa reset?
*) SEU in receiver diode can be emulated —> contact Jan for details.
*) disable driver on elinks and see what happens? —> contact Paulo for details
*) bGTX 0 in loop bach mode once beam is back —> can be done by TPC
*) for all detector monitoring improvment —> action on TPC and CRU FW team
*) include Sophie & Daniel Montesinos