Date   : Mon, 11 Mar 1991 09:39:24 PST
From   : Sprague.Wbst311@xerox.com
Subject: Obvious Solution the Motor On

Forget the AND gate to generate Motor On, forget connecting Motor On to Ground,
just forget all that sort of stuff.  The solution was so obvious that I am
surprised that I and others didn't think of it first.

I would have written this sooner, but due to the massive ice storm we had a
week ago, I only made it into work two days last week, and those were very busy
days.

Back Track:

This has to do with replacing 8 inch drives with 5.25 or 3.5 inch high density
drives.  The problem was that an 8" interface does not supply the Motor On
signal, and that many of the new high density drives do not provide a jumper to
tie the Drive select signal to Motor On.

I had suggested using an AND gate to provide the signal.  Zoltan Fekete had
suggested that the signal just be grounded.  Grounding the signal is not good,
since as long as a floppy has been inserted, the heads will be loaded and the
drive will spin.  As I prepared to make a new interface board (this one with
room to add an AND gate chip), I realized that even with this addition, both
drives will spin if each has a floppy, and either one is selected.

Then the solution this me.  Use an IBM type twisted cable!!!  This way, the
Drive Select signal could be tied directly to that drives Motor On, and nothing
else was needed (and you can buy them for about the same cost as it would take
to make one).  Since I had a spare twisted cable lying around, I modified my
old interface board, and it worked fine.  There is still a problem with drive
ready, but I'll get to that later.

In the IBM twisted cable, all wires go directly to the first drive connector
(drive B) and then wires 10 through 15 are twisted and installed backwards for
the second drive connector (drive A).  The ground wires all come out correct,
so this results in swapping positions for DS1/Motor ON and DS2/DS3.

On your interface board, connect the controllers Drive Select 1 signal to the
Floppy Interfaces Drive Select 1 (Motor On for Drive A) and Drive Select 3
(Drive Select 2 for Drive A).  The controllers Drive Select 2 still connects to
the floppy Interfaces Drive Select 2, and also to Motor On.

Both A and B drives should now be configured as Drive B, because of the twist.
Drive A goes on the end of the cable, after the twist.  The termination
resistor should now be in drive A, rather than Drive B, because A is at the end
of the cable.

Have an controller that supports 4 drives?  No problem.  Add a second 34 pin
header to your interface board, and connect as above, substituting the
controllers Drive Select 3 and Drive Select 4 signals for Selects 1 and 2.
Again, Drives C and D must be configured as Drive B, and a termination resistor
should be in drive C (which goes at the end of the cable).

Hmmm, for those who might have been lost by my "Interface board", you need a
way to connect the signals coming from the 8" drive controller to the 5.25 (or
3.5) inch drives.  I took the controllers signals directly using a 34 wire
ribbon cable.  That connects to a 34 pin wire wrap header on my Interface
board.  There is a second 34 pin wire wrap header on the interface board, and
that is for connecting to the floppies themselves.  Using a wire wrap tool, I
then connected all the interface signals to their proper pins on both headers.

Last, I mentioned a Drive Ready problem.  On my Chinon 1.2Mb drives, there was
a jumper to select between the Disk Change Function and Drive Ready on pin 34
of the floppy.  I set it for Drive Ready .... and it almost worked.  The signal
*IS* compatible with the 8" drive ready signal.  The problem is that my
computer times out before the drive is ready and gives me an error.  Since the
drive is still spinning (the drive spins of about 10 Seconds after it is
de-selected, in case you want to use it again), if I try the command again,
everything works fine.

Still it's a pain to boot twice, or type DIR, wait then tell it to retry.

The obvious solution would be to tie the controllers drive ready signal to
ground.  As Zoltan Fekete pointed out, this will probably work fine for reading
the drive, but could cause problems when trying to write to it.  I have had no
problems writing to the drive .... yet.  But I have not used the system enough
to know if I can get away with this or not.

The other solution of course, would be to modify the BIOS (I think) of the
computer so that it takes longer for the error time out when spinning the drive
up.  Sigh!

I assume that most of you who have done this conversion have just tied Drive
Ready to ground?  If so, how well has it worked?

                               ~ Mike  (Sprague.Wbst311@Xerox.Com)