October 8th 2002 CFT Trigger Commissioning MeetingIndexHardwareSequencers etc.DFE_ware, link testsL1CTTL2CTTReadout, ExaminesDiscriminator StudiesDocumentationAOBhardware (Jamieson)One DFEA power supply is back from the manufacturer, and will be installed during the next access. The 5 V supply, run at 3.3 V, will provide increased headroom for remote sense operation. The DFE power supplies will be added to Julie's monitoring GUI and to the SMT power supple control GUI. back to indexSequencers etc.The Sequencer Controller mods have been implemented on the platform. Some AFEs had their VSVXs reprogrammed, but none so far for the stereo layers. back to indexDFEWare, link testsMakoto added a new Mixer masking library, allowing masking at the four-sector level. For completely independent (sector-level) masking, Stefano needs to write new firmware.The fake/real data switching bug in the AFE LVDS control was fixed. The STOV/STSX link problem is not due to the cables; the next step is to swap transition boards. The CTOC/CTTT tests worked better after the power supply voltage adjustment, but some links still seem bad. The chain from Sector 22 through the CTOC in slot 4 to the CTTT is ok, though. L1CTTOnce our ethernet hub on the west platform is powered up again (next access), we should be able to start L1CTT chain tests on the platform. Crate x13 has to be taken out of the run before turning off/blanking CTOCs. L1MUON will not be affected by DFEA output mode switching; the output modes affect the link to the CTOC (and STOV/STSX) only.L2Steve is working through the STT 'wish list' agreed upon in a meeting last week between Uli, Georg, Steve and Stefan.STOVs 3 and 6 have LVDS problems (see dfe-ware above). L2 preprocessor contacts:
Jamieson is almost done with the DFEA L2 code. Once this works, we can proceed to test Thei's STOV/STSX firmware. At that point it will make sense to start discussing Thei's next trip to Fermilab. Readout, ExaminesCTOC readout: Kyle sees sometimes 5, sometimes all 8 CTOCs. The data format for the DFEA info seems to have some extra scrambling, compared to the CTOC-built header. All in all the test vector readout results are very encouraging! Real data have no tracks at all, due to the VSVX pipeline being set to the previous 132 ns 'crossing', which is guaranteed to be empty in 396 ns operation.Discriminator Studies (Yurii & Yuri)Yurii has been checking the data for illegal SVX codes; this work will be integrated into the header and data integrity check done by Lucas's examine.DocumentationEverybody was reminded that we need documentation for every piece of firmware. Currently this area looks pretty bad.AOBGeorge reemphasized the importance of peaceful coexistence with D0 data taking activities. In particular, crate x13 should be taken out of the global run BEFORE it goes terminally front-end busy due to blanked or otherwise removed inputs. Whenever you change the configuration for AFEs, Mixer, DFEAs and/or CTOCs, be mindful of the possible side effects for the CTT readout chain!Stefano described recent changes to the Mixer system, mostly having to do with improved diagnostics and monitoring. He has also provided firmware files for Yurii's remote download archive. We do not plan to activate this feature at this point in time, though.
Last modified October 17th, 2002, by SG. |