"V W Wall" said in :
This is not true. Cable select is determined by pin 28 *not* pin 34,

I'm a little more awake than last time. You are correct. Pin 28 is for
CSEL (cable select). Pin 34 is for PDIAG/CBLID. PDIAG was the original
designation and CBLID was added for 80-wire detection. Pin 34 cannot be
shorted to ground in the mobo otherwise every ribbon cable connected to it
would look like an 80-wire ribbon cable. Pin 34 is shorted inside the blue
connector (on the mobo end); see
http://www.pcguide.com/ref/hdd/if/id...Cable80-c.html.

I don't have any 80-wire ribbon cables at the moment where I can see a tiny
(3mm) cutout section of a wire. According to one user in these newsgroups,
"The wire is cut near the motherboard end of the cable so that the drives
will not "see" the ground on that pin." If you have a ribbon cable with the
cutout, for which signal pin is the cutout, or is it in one of the shielding
ground wires?

Although it sounds great to spout specifications, what I've measured doesn't
fully jive with what has often been touted in the newsgroups as the spec for
80-wire layout of signal wires (to pins) and for shielding wires (to reduce
crosstalk). I had been told that the signal wires (going to the pins) were
the odd numbered wires (which meant the outside wire on the striped side was
for pin 1) and the even numbered wires were the shielding ground wires.
When ohming out a couple of 80-wire ribbon cables, that's not what I found.
For my spare 2 ribbon cables (both 80-wire/40-pin), the odd wires were the
shielding grounds and the even ones were the signals (going to the pins).
Then I noticed the above linked article to PC Guide says the added ground
wires can be either the even or odd set of wires. I haven't ohmed out other
ribbon cables to get a consensus. Great, now I don't know if the outermost
striped end of the ribbon cable is for pin 1 or a shield ground. Or looking
at it in reverse, pin 1 could be on wire #1 or #2. I have to test it to
find out.

Pin 34 did show continuity between the black and grey connectors (these go
to the hard drives). So, for the drives, this line is still connected
between them. This is still a requirement in 2-drive configurations as
PCDIAG from the slave uses this to tell the master drive that it is ready.
"PDIAG (passed diagnostics) is a signal used by drive 1 to tell drive 0 when
(and if) it has passed its diagnostics following a power-up or a reset.
Drive 0 uses this information to inform the system of a drive 1 failure"
(according to
http://www.uib.es/c-calculo/scimgs/f...Interface.html).
However, neither of these had their pin 34 connected to pin 34 on the mobo
connector (blue), so there is an open in the wire between the mobo and drive
connectors. I could not see any cutout in the ribbon cable along the wire
for pin 34, so it must be the ribbon cable is cut back on this wire so it
doesn't engage the forked tooth for that pin, the cutout in the ribbon cable
is just above the forked tooth so there is no wire to engage into the tooth,
or the forked tooth is broken from the pin. So you might see the cutout or
you might not.

Also, for pin 34 to be grounded to indicate an 80-wire ribbon cable, the
only way the motherboard could see that is if pin 34 actually went to a
ground pin on the connector. The motherboard can only see what's on the
connector pins. For the blue (mobo) connector, pin 34 showed continuity to
pins 2, 22, 24, and 40 (there are more ground pins but that was enough; see
http://www.interfacebus.com/Design_Connector_IDE.html for a pinout of the
connector). So pin 34 (but only in the blue mobo connector) is shorted to
ground.

So, to summarize how pin 34 is used:

- You don't know if the signal pins use the odd- or even-numbered wires.
You'll have to check by ohming out the pin to see if it is connected to wire
#1 or #2. For my 2 samples, pin 1 went to wire #2 (so signals were on the
even-numbered wires).

- For the 80-wire/40-pin cable, pin 34 on the mobo connector (blue) is
disconnected from the ribbon cable. Pin 34 on the drive connectors (grey
and black) are connected. You may see a cutout in the ribbon cable, you
might not. I've never seen a cutout in any of my 80-wire/40-pin ribbon
cables.

- Pin 34 on the pin side (only to the mobo) of the blue connector is shorted
to ground.

Me: I don't use cable-select ribbon cables.
V W Wall: There is no such thing--all cables made to the ATA spec can be
used.

"To use cable select, both devices on the channel are set to the "cable
select" (CS) setting, usually by a special jumper. Then, a special cable is
used." "On a cable select cable, one of the connectors (the "master
connector") has pin #28 connected through to the cable, but the other (the
"slave connector") has an open circuit on that pin (no connection)." This
according to http://www.pcguide.com/ref/hdd/if/ide/conf_CS.htm. You don't
use "special" cables for standard installations. I had heard of cable
select cables but never bothered to use them or even investigate into them
because I always want to determine which drive is slave and master instead
of worrying about which is connected to which connector (although the
preferred hookup is to have the master at the end of the ribbon cable). So
then I ohmed out my 2 sample 80-wire/40-pin ribbon cables for pin 28.

You're right ... mostly! Although they are 80-wire/40-pin ribbon cables and
NOT used at anytime for a cable select configuration, they appear to be
setup to be usable for cable select (if you configure your drives that way).
Pin 28 on the mobo connector (blue) shows continuity to pin 28 in the far
end drive connector (black) but pin 28 in the middle connector (grey) is not
connected to either of the other connectors. In fact, by the feel of the
needle I was inserting into the grey connector, it feels like there isn't
even a female pin in that hole. So at one time it might've been that pin 28
was connected across all the connectors for non-cable select setups and a
special cable-select ribbon cable was needed for cable select setups, but at
some point all ribbon cables became cable-select cables whether you used
them that way or not.

From what I see for the ATA specs at www.t13.org, CSEL (being open on pin
28) was *optional* in ATA-1 (http://www.t13.org/project/d0791r4c.pdf) and
ATA-2 (http://www.t13.org/project/d0948r4c.pdf), plus they mention "Special
cabling can be used by the system manufacturer to selectively ground CSEL".
Again, you don't use "special" cabling for *standard* setups. Sure sounds
like there used to be normal (non-cable select) cables and cable-select
cables. I didn't find a link for ATA-3, and for ATA-4 and up they want me
to buy the spec from ANSI (no thanks). So at what point the "special"
cabling that would provide CSEL became standard is unknown to me. I do
remember that at one time you had to be careful to NOT get stuck using a
cable-select version of the ribbon cable if you were configuring the drives
as master and slave, or, the other way around, you had to be sure you used a
cable-select version of the ribbon cable if you wanted to use cable select
on the hard drives.

--
__________________________________________________ __________
*** Post replies to newsgroup. E-mail is not accepted. ***
__________________________________________________ __________