Path: utzoo!attcan!uunet!lll-winken!lll-tis!helios.ee.lbl.gov!nosc!ucsd!rutgers!gatech!mcdchg!ddsw1!karl
From: karl@ddsw1.UUCP (Karl Denninger)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Hard drive / controller problems
Summary: Select lines are on the twisted portion of the cable
Message-ID: <1393@ddsw1.UUCP>
Date: 15 Jul 88 21:25:19 GMT
References: <222@psuhcx.psu.edu> <16800327@clio> <230@psuhcx.psu.edu> <4452@sphinx.uchicago.edu>
Reply-To: karl@ddsw1.UUCP (Karl Denninger)
Organization: Macro Computer Solutions, Inc., Mundelein, IL
Lines: 64

In article <4452@sphinx.uchicago.edu> mobo@sphinx.uchicago.edu.UUCP (Samuel Wilson) writes:
>In article <230@psuhcx.psu.edu> wcf@psuhcx (Bill Fenner) writes:
>>In article <16800327@clio> berger@clio.las.uiuc.edu writes:
>>|
>>|......using a "twisted" 34 pin cable.  
>>|
>>Well, I took the cable apart, removed the twist, set DS2, and bagoomba!
>>20 megs.  Nobody that I talked to, including Seagate Tech support and all
>>the computer stores around town, thought anything about the fact that the
>>cable was twisted.  I know that at least 2 other people are using twisted
>&c...
>
>I put a second ST225 in an ATT6300.  I was using a Western Dig. WX2
>controller.  I was also using a cable with a twist between the 
>connection for drive c and d (or a and b, as the case probably was). 
>the controller could talk to one drive, or the other, but never both
>at once.  
>
>I have some idea how controllers work, and some idea how drives work, 
>but no idea why *that* worked.

It's simple!  Here's the poop:

ATs (and some XTs) have been made with a "twist" in the cable.  These hard
disk control cables (the 34-pin is called the control cable) are *NOT* the
same as a twisted floppy cable, although they perform exactly the same
function with the twist in both cases.

The reason for the twist is to change drive select lines (select, that
is).  This allows people to slam in drives without understanding that they
have to set the little drive-select jumper first -- everyone is on DS 2 and
all works fine (base = DS 1).

However, with the clones being out there, there is no longer a "standard".  
As a result, you will often get a drive with the jumper wherever it was 
during final factory test (usually DS 1).

If you have *no* twist, then drive 1 is set as Select 1 (base = 1 again), and
drive 2 is set Select 2.  IF there *IS* a twist, set BOTH drives to Select
2.  The drive right after the twist in the cable (usually the end connector)
will be drive #1, and the other one will be drive #2.  All this "magic" 
is done by simply swapping the select lines in the wire....

(A note:  It's always a good idea to check the jumpers on a drive before
	  installation.  Sometimes there are a few that aren't supposed to
	  be there present -- I've even seen 'em come out of the factory
	  packaging with TWO OR MORE select jumpers installed!  This will
	  make for one non-functional system, and if it's a secondary you're
	  installing may scramble the primary on boot).

Floppy drives work exactly the same way, as do the twists.  

See, when all the world was IBM all the techs just set the drives to 
Select 2 and off you went.  This has the additional advantage in that the 
drive select is position independant -- you can swap selects just be 
changing the position on the cable.

Note that floppy cables w/twists, while very similar in appearance, twist a
different number of wires.  DO NOT USE a twisted floppy cable for a HD --
all kinds of bad things can happen, although all you usually end up with is
a drive that doesn't work.

--
Karl Denninger (ddsw1!karl) Data: (312) 566-8912, Voice: (312) 566-8910
Macro Computer Solutions, Inc.    "Quality solutions at a fair price"