Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!samsung!munnari.oz.au!brolga!bunyip.cc.uq.oz.au!qut.edu.au!lemon From: lemon@qut.edu.au Newsgroups: comp.periphs.scsi Subject: Re: Maxtor, Wangtek & WD FASST 7000 problem Message-ID: <1991Apr19.175336.27422@qut.edu.au> Date: 19 Apr 91 22:53:36 GMT References: <1991Apr12.150258.26918@qut.edu.au> <6639@gssc.UUCP> <1991Apr18.020832.19028@bohra.cpg.oz.au> Organization: Queensland University of Technology Lines: 52 In article <1991Apr18.020832.19028@bohra.cpg.oz.au>, als@bohra.cpg.oz.au (Anthony Shipman) writes: > In article <6639@gssc.UUCP>, bradd@gssc.UUCP (Brad Davis) writes: >> In article <1991Apr12.150258.26918@qut.edu.au> lemon@qut.edu.au writes: >> >The problem is that when a tape is placed in the tape unit the hard drive >> >makes chlunking sounds(well the same sounds it makes at power up, a reset?) >> >and also after reading a number of blocks from the tape unit the hard drive >> >does the same thing, which results in the software hanging. It also hangs >> >when trying to write to the tape unit, the hard drive make the same sounds. >> >> >Has anyone any idea as to what may be wrong ? >> >> >PS : the SST software reports the correct ID's for the drive's. setup is on >> >a 386 20Mhz, AMI bios. CHIPS chipset. >> >> This sounds something like a problem I'm having with an Adaptec 1540 >> and SCSI tape drives. After a little experimentation, I found that it >> worked fine on Compaqs and failed in the same way (hanging on reads) on >> three out of three Chips & Technologies motherboards. > > My motherboard is 20MHz 386, (old)AMI BIOS, C&T chipset. I use an AHA1542B and > Archive 2150S (+Maxtor 4380S) and it all works very well. How big is your power supply, as the problem turned out to be the drain on the power supply when the tape unit was started up which caused a power drop and the maxtor seeing a low power level, did a reset. >> >> I believe the problem in my case, and possibly yours, is a bug in the >> C&T chipset's DMA circuitry that shorts a couple of address lines together >> at a crucial time. C&T supposedly issued an errata sheet with work arounds, >> but apparently all the motherboard OEMs ignored it. > > My motherboard is Taiwanese, some people called "Sure". Perhaps they got it > right? Slowing down the DMA did help a little, it was able to write a few more blocks before hanging. The power supply was the problem, It nows works fine with the two power supplies, at lest untill I get a 300W unit. PS Thanks to Al Amet who hit the nail right on the head, but your message got to me as I was writing this reply. -- -------------------------------------------------------------------------------- John Lemon _--_|\ INTERNET: Faculty of Information Technology / \ lemon@qut.edu.au Queensland University of Technology \_.--._/ lemon@snow.fit.qut.edu.au Queensland, Australia v John.Lemon@f555.n640.z3.fidonet.org BBS: Ph +61-7-372-8321 fidonet 3:640/555 --------------------------------------------------------------------------------