|NOTE: Old contents of BU CMS wiki. These (green) pages cannot be edited|
See www.hcal.info for more links to active TWikis
Wu found problem with upper bits. Try new temporary version of oslbve.mcs.
We have a working test setup now for 5 oSLB at once, describe in the test procedure document.
It requires Wu's old htr_oslb_28.mcs firmware for the oSLB boards, and the special [http://cmsdoc.cern.ch/cms/HCAL/document/CountingHouse/HTR/firmware/HTR_oSLB_tester/HTR_oSLB_tester.mcs HTR_oSLB_tester.mcs] (a variation of v57) on the receiving HTR.
Today I am testing Tullio's new [http://cmsdoc.cern.ch/cms/HCAL/document/CountingHouse/HTR/firmware/HTR_59/MCS/HTR_00059.mcs HTR_00059.mcs] which should include the oSLB pattern tests in the standard firmware. New features:
Above seems to work OK, but with 2c0 = 0 (and oSLB map=0xfd) I would expect to see the BCN counting. No luck. However, with oSLB mode=2 I do see BCn6:0 in the upper bits as expected. Hmm...
Tullio made special HTR firmware to support oSLB testing. It performs automatic checking of pseudo-random data generated by oSLB when bit 4 at 0x4c is set in oSLB.
Test run overnight (ending 1/15 AM) using HTRoSLBTest.cc in ~hazen/src/oSLB on cms2:
---Fiber 5--- ErrAcc: 0x 0000 0000 0000 0000 NonDVAcc: 0x 0000 0000 0000 0000 RDErrAcc: 0x 0000 0000 0000 0000 0 errors; BER: < 5.8514e-15 ---Fiber 6--- ErrAcc: 0x 0000 0000 0000 0000 NonDVAcc: 0x 0000 0000 0000 0000 RDErrAcc: 0x 0000 0000 0000 0000 0 errors; BER: < 5.8514e-15 ---Fiber 7--- ErrAcc: 0x 0000 0000 0000 0000 NonDVAcc: 0x 0000 0000 0000 0000 RDErrAcc: 0x 0000 0000 0000 0000 0 errors; BER: < 5.8514e-15
Received 69 stuffed oSLB. Wu is programming the CPLD on the first one. Firmware rev C (as documented here ) is the latest. There is no real software support for this version yet.
Simple check-out of first board OK. Fixed test pattern and (new in Ver 0xd) LFSR random data received ok (low statistics).
Histogram mode seems to work. Used some updated scripts in ~hazen/src/htr/oSLB/2008 and the 'TTC/VI' command in htr.c to send pulses at fixed BX, and the correct buckets in the histo are filled.
Wu has programmed CPLDs on all new boards.
''Test Plan for Andrew''
Could test multiple boards at once. We have 7 HTRs... if we had enough fibers we could install 5 oSLB and cable the 15 links to one HTR.