Guys, I came across something rather enlightening, and
troubling for those of use that use external HD enclosures.  In
a particular PC build for a customer, the HD's are very close
the IDE connectors on the mobo.  Sometimes it's like this,
other times the IDE connectors at way at the top of the mobo.
Since I'm very particular about the aesthetics of PC's we
build, and air flow, I wanted to use 6" IDE cables for a
cleaner look, and less obstructive to air flow (yes I also
sleeve my cables as well).  I happened to have one, that
happened to come with a pack of FireWire (FW) cables.  It was
an 80-pin cable, which is supposed to be for ATA/UDMA66+ HD's.

Come to find out, even though they are indeed 80-pin IDE cables
which are supposed to be UDMA/ATA66+, they are not.  I hooked
up a HD with one and the BIOS kept complaining it was an ATA100
or higher HD, but the cable was only "40-pin" and of course
would only run at UDMA 33.  Apparently, (as the included FW
cables indicate), even though they are 80-pin, these IDE cables
are **ONLY** for USB or FW external enclosures.  I thought they
were simply just a shortened ATA66+ cable, apparently not.

Now this brings up an interesting area of the data transfer
speeds.  UDMA/ATA33 is 33MB/sec max theoretical transfer speed,
USB 2.0 is 60MB/480mb/sec, and FW400 is 50MB/400mb/sec.  Both
of course more than 33MB/sec.  So any external enclosures using
these shortened pseudo 80-pin cables are not getting the
maximum transfer of which they are capable.....right?  I doubt
anyone has ever removed the 4-6" 80-pin IDE cable in an
enclosure and tried a typical ~18" 80-pin IDE cable.  It would
be interesting to do some benchmarks before and after.  I may
when I have the time.  Perhaps there is "something" in the USB
or FW interface that does not require real 80-pin cables.??
Even so, you're still only dealing with a cable that can only
transfer 33MB/sec so it seems we're getting short-changed here
with the drives in these enclosures not performing as well as
they could.

However, it could be due to the fact the mobo is where the
"80-pin cable detector" so to speak, is located.  This is why
an 80-pin cable has a different color, usually blue, that
connects to the mobo, and the gray or black connects to the
drives (master or slave depending on jumper position).  If you
reverse the cable, the BIOS will think it's an old 40-pin UDMA
33 cable.  Maybe these shortened cables ARE the same cables,
just missing this 80-pin cable detector connector pin since
they are not required because they are in an external
enclosure.  But, it seems like it would be cheaper to just make
shortened 18" cables, instead of a special 6" cable void of the
80-pin cable detector pin.

Any info or thoughts welcomed.

(BTW in case anyone is wondering about USB 2.0 Vs FW400,
FW is actually a little bit FASTER sustained speeds in actual
use since USB 2.0 has more "overhead".  Small files, it's not
going to matter.  But huge files like digital video, or
copying/backing up your entire main HD, FW will be a bit
faster.  FW800 is out now which as the name implies is
supposed to be twice as fast as the old FW, and there's
probably a new USB protocol out now that's faster than 2.0.
If not, there no doubt soon will be).
-Clint

God Bless
Clint Hamilton, Owner
http://OrpheusComputing.com
http://ComputersCustomBuilt.com
============= PCWorks Mailing List =================
Don't see your post? Check our posting guidelines &
make sure you've followed proper posting procedures,
http://pcworkers.com/rules.htm
Contact list owner <[EMAIL PROTECTED]>
Unsubscribing and other changes: http://pcworkers.com
=====================================================

Reply via email to