IBMs and Quantums are very reliable, and Maxtor, JTS, and Fujitsu tend to
work without problem. The problems seem to come with Seagate and certain
chipsets (VIA MVP3, Intel TX, Intel 810) and Western Digital more generally
(WD actually blows off the CRC). Some of the Seagate+chipset problems
occur only at certain bus speeds, like 75 and 100 MHz FSBs and disappear at
95, 83 and 66. And, yes, cabling is also important. I have seen some
Mobos freeze solid when presented with one brand of UDMA 80-pin and work
fine with another. More like plumbing than science :-<
Civileme
Ron Stodden wrote:
> Civileme wrote:
> >
> > Having gone through the mill more than a few times, I have to ask this
> > right away.
> >
> > What disk drives are you using? There are some which do have signal
> > reflection errors on Pentium/Pentium II code and ATA/66.
>
> Pentium III, 550 MHz. Both drives and the CDROM are ATA-33. The new
> motherboard is ATA-66. New motherboards come with only one ATA-66
> ribbon (80 conductor) cable. I have tried the following cable
> combinations: primary 66, secondary 33; primary 66, secondary 66;
> primary 33, secondary 33 - with no effect on the crash incidence.
>
> hda is IBM-DHEA-36480 6197MB w.476kB cache, CHS=790/255/63 (LBA)
> hdc is IBM-DJNA-3515209 14664MB, w/430kB cache, CHS=29795/16/63 (BIOS
> says LBA???)
>
> Substituting another ATA66 motherboard on the retailer's premesis led
> to all problems disappearing.
>
> Are these recent IBM drives reflection free? I tend to blame the
> motherboard and am getting it replaced. I do know that the proper
> mix of 33/66 m/b, 40/80 conductor cables, and 33/66 hd devices is
> critical - get it wrong --> instability.
>
> --
>
> Regards,
>
> Ron. [AU] - sent by Linux.
--
BETA-testing Netscape 6
and its mailer
