Need Shugart 851 manual

From: Richard Erlacher <>
Date: Tue Mar 5 11:45:36 2002

What??? You mean that the controller doesn't have a normal FDD cable
connector? Is it possible that there's an intermediate signal distribution
board, a concept not unheard-of among Intel products, that you don't have?
I'll check my old doc's just to see what I can find, but I don't remember that
there was such a thing. Isn't there someone somewhere who's got one of these
boards working? Perhaps they can shed light on what works and what doesn't.

As Tony pointed out, 8" floppy drives expect to see a 50-conductor cable, of
which the odd pins are grounded. There's a very much standardized pinout,
though some makers didn't stick with it, e.g. PerSci, but, at least in that
case, the differences were only a few (4) displaced signals. After all, there
are only so many signals a drive and controller can reasonably use to
communicate with one another. Unfortunately, this particular board doesn't
seem to have adhered to that standard. Most of their later boards did use the
standard pinout, so I conclude that there may have been an intermediate board,
of which there would have been several versions for several different drive
subsystems. It's possible that there was an external clock extraction circuit
(data separator) as well, as was the case with at least one of their
Multibus-1 HDC's.

I'm unable to locate my Intel doc's from that era, so I can't verify what the
interconnection scheme they promoted for that system was, but if it's not the
"standard" FD cable, I'd say you have a bit of research to do, since this
apparent lack of a drive interface specification suggests, to me, at least,
that there's a physical piece of the hardware that's missing, along with the
associated documentation, and that's where you're more likely to find the
drive-subsystem-related information that's been so conspicuous in its absence.
This also fits together with Tony's observation that most such products simply
included the drive manufacturer's doc's. Since this product may not have been
intended for use with a specific configuration, it wouldn't have documented
that configuration, or any other artifact of the drive subsystem, since the
had a product that handled that.

The absence of a drive-compatible connector of any sort on the controller
board explains the absence of any drive information in the document you have
for the controller. I have a, much later of course, hard disk controller that
uses an off-board data interface to the several sorts of drives it supports,
and, not surprisingly, while the controller manual tells lots of things about
the controller, it says absolutely nothing about the drive interface. That's
clearly left up to the documents relating to the varous drive data interface
boards they require.

I'd give that controller manual a good read, just to see what, if anything, it
says about the floppy drive interface board to which it is probably supposed
to be wedded. There are probably several candidates. Intel was infamous for
taking what would easily have worked on a single board and making two or, if
they could, three, out of that function, since they could charge more for
that. Intel boards often cost $25 less than a competitor's product, only to
turn out to cost $2500 more once you'd bought the necessary companion products
that made them useable. They liked, particularly, to do that with their own
products, since you couldn't go anywhere else to buy the functions you wanted,
and, you wouldn't readily give up the Intel products they supported, having
invested a huge amount in the development system for Intel's. Motorola was
the same way, of course, but it wasn't quite so blatant.

----- Original Message -----
From: "Joe" <>
To: <>
Sent: Tuesday, March 05, 2002 9:25 AM
Subject: Re: Need Shugart 851 manual

> At 07:46 AM 3/5/02 -0700, you wrote:
> >see below, plz.
> >
> >Dick
> >
> >----- Original Message -----
> >From: "Joe" <>
> >To: <>
> >Sent: Tuesday, March 05, 2002 6:50 AM
> >Subject: Re: Need Shugart 851 manual
> >
> >
> >> At 04:31 PM 3/4/02 -0700, Richard wrote:
> >> >I'd advise you to pursue this from the controller end, initially.
> >>
> >> Good idea. Unfortunately the card connector is numbered 1-100 with odd
> >> numbers on one side and even one the other (exactly same as S-100). But
> >> it's connector is numbered A1 through A50 on one side and B1 through B50
> >> the other. It makes things confusing! It's not something that you should
> >> do when you're tired!
> >>
> >
> >Agreed ... I'm having a little trouble with that one myself ... there are
> only
> >25 pairs on the cable.
> There are (or can be) 100 wires per cable.
> What do they do with the numbers greater than 25 on
> >one side and 50 on the other?
> There are 50 contacts per side. All the odd numbered wires are on the A
> side and all the even numbered ones are on the B side. But here's one place
> that it gets very weird, A1 is pin 99 and A50 is pin 1 !!! The same thing
> happens on the B side, B1 is pin 100 and B50 is pin 2!
> I came up with a couple of formulas to help figure out the wire numbers.
> Wire # for contact Ax is (51-x)*2 and the wire # for contact Bx is
> ((51-x)*2)-1. I *think* that's right! it was late when I was working on it.
> >
> >Perhaps you'll want to give up on using th '85x's on the Intel box and use
> >80x's instead. There will be a cleaner path for you to follow, though,
> >clearly, it won't be totally obvious.
> I would use 800s if I had any but it seems like all I can find are 85xs
> :-/ I've got two extra 800s but they've been out in the weather and they
> look nasty!
> Joe
> >
> ><snip>
> >
> >
Received on Tue Mar 05 2002 - 11:45:36 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:35:09 BST