In message , Paul
I like to draw pictures for people.
For a single IDE drive, it *always* goes on the end, like this.
Cable_Select (if 80 wire, CS is allowed)
IF I understood the Wikipedia article, some 40 wire cables _did_ allow
CS, by omitting line 28 between the remote connectors, and master having
to be the middle one, with the unterminated bit of cable providing all
the reflections you'd expect )-:. Probably why CS wasn't used much in
the early days. I _think_ I _do_ remember seeing 40 wire IDE cables
where the cable between the remote connectors was two ribbons, i. e. had
a gap in it - not a twist like a floppy connector, just a gap, which had
I looked closer I'd have seen was the omitted line 28. (Presumably line
28 was just connected to 0 or 5V on the mobo.)
When you add a second drive, it can be like this. Or,
you can run CS on both drives, if you are using an
80 wire cable (with that twist in it). I didn't want to junk
(Not a twist, just an omission. Apparently often done by just omitting
the insert on the middle connector's line 28.)
up the diagram, by adding CS to the table for both drives.
Master with Slave (some brands have a
distinction on the jumpers)
(I'd forgotten that! Wonder why.)
Do *not* do this, as the end of the cable constitutes a
stub and causes excess reflections and corrupted data.
Even in the best of circumstances, the signals on that
cable look horrible. The signals look more horrible
in that last case.
Must have been even worse with the 40 line cables!
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)[email protected]
A biochemist walks into a student bar and says to the barman: "I'd like a pint
of adenosine triphosphate, please." "Certainly," says the barman, "that'll be
ATP." (Quoted in) The Independent, 2013-7-13