July 10, 2019 online meeting minutes
present: Sergey Boyarinov, William Gu, Nathan, Stepan, vpk
1. HPS status
- FADC fix: bug was found at high rate in HPS reading hodoscope FADC; confirmed in test setup with event rate >30kHz with high data rate- one FADC with zero threshold, close to 1Gbit/sec; fixed by Ed with help from Hai and Ben; new test setup was established with VME crate in CODA group lab and server in hall counting room; after bug was fixed, Ben recompiled hall b FADC firmware (control chip version b3 now); Sergey tested new firmware in hall b test setup and confirmed that problem fixed; new firmware installed in HPS, will be installed in CLAS as well
Stepan tensioned that some experiments may need to run FADCs with zero threshold, it may be needed to investigate if we can read all 256 channels and what event rate can be achieved (hall C experiments in 2-3 years)
- SVT readout: TIPCI board was modified to route 'good' clock to the output connector, new wire installed; firmware was modified accordingly, output clock set to 125MHz for HPS svt usage; cables between TIPce and ATCA blades were modified as well; Ben and William checked clock signals by scope, 'bad' one had significant jitter (hundreds ps) and new one decreased it at least factor 10; preliminary testing shows significant readout stability improvement
Bug in rol1 was found by Omar and fixed by Sergey: svt builder header event counter was overwriting tag 0x16 when event number exceeds 27 bits (>134 MEvents); tested, problem solved
Performance tests shown svt readout time per event about 20us on clonfarm2 and 40+ us on clonfarm3, and occasionally those numbers swapped during the run; svt team is working to understand the performance limitations
- PRs are in process for new comps, TVs, fileserver etc