[comp.os.cpm] Using 3.5" floppies instead of 8" floppies

ralph@uhheph.phys.hawaii.edu (Ralph Becker-Szendy) (02/09/91)

Since someone asked what the method is to hook up a 3.5" floppy disk drive
instead of an 8" drive, here is a listing of pin-numbers on the cables. A 5.25"
HD or 3.5" HD floppy is electrically nearly equivalent to a 8" DD floppy. The
only major difference is that some 5.25" HD and all 3.5" HD floppy drives spin
at 300RPM (instead of 360RPM, what it used to be for 8" floppies), so you can
pack more data onto each track (but you don't have to). Some 5.25" HD drives
switch to 360 RPM when switching to HD mode though (usually jumper selectable).

All odd pins on both cables are ground. 8" floppies use a 50-pin cable, 3.5"
and 5/25" floppies use a 34-pin cable:

Pin   Pin   Signal
50p   34p
==================
12     4    Disk change (may be on other pins too, not very standarized)
14    32    Side select (some 8" drives use no side select, instead two
            different drive select lines)
18     2    Head load (all 5.25 and 3.5" drives ignore this, sometimes use
            the pin for something else)
20     8    Index pulse
22    34    Ready (some 5.25 and 3.5" drives have no ready)
26    10    Drive select 0
28    12    Drive select 1
30    14    Drive select 2 (not all 5.25 and 3.5" drives have that)
32     6    Drive select 3 (not all 5.25 and 3.5" drives have that)
34    18    Step direction
36    20    Step pulse
38    22    Write data
40    24    Write gate
42    26    Track zero
44    28    Write protected
46    30    Read data
      16    Motor on (only some 5.25 and 3.5" drives)

Going from this list to a special-wired cable should be easy. Supposedly there
are some 3.5" floppies which also get their +5V supply fed through the data
cable. Never seen one yet (at least never taken one apart).

To emulate a 8" drive you have to set up the 5.25" or 3.5" drive to work in HD
mode. Some 3.5" drives can sense the density-indicator hole in the floppy,
others need a DENSITY SELECT signal from the controller, sometimes on pin 2,
sometimes on pin 34. If the controller doesn't supply that, one can emergencies
just use ground or Vcc instead and leave them permanently in HD more. Sometimes
one can find the multiplexer on the controller which switches between 5.25"
SD/DD mode (250kHz data rate) and 8"/HD more (500kHz data rate), and use the
signal which controls that multiplexer to switch the drives between normal and
high density.

If the 5.25" or 3.5" drive really needs the MOTOR ON signal (most newer ones
can be jumpered to do the motor control themselves), just strap it to ground,
that keeps the engines under full steam.

If your controller needs a READY signal from the drive and the drive does not
generate one, take the SELECT signal you use to select the drive, delay it by a
few hundred ms with a 74121 and use that for READY. Some controllers even work
with a connection between READY and SELECT (or even more cruel, just grounding
READY so the drive is always ready); you might get read errors when selecting
the drive for the first time (while the floppy spins up to speed), but one or
two retries will cure that; most controllers / BIOSes will retry automatically.
Problem: The controller (and thereby your BIOS) can not determine whether there
is a floppy in the drive. On the real 8" drives and on 5.25" drives I get a
"NOT READY" error immediately when I try to use a drive with no diskette in it;
on the 3.5" drive it takes about 30 seconds, while the controller re-tries like
mad, and finally I get a "RECORD NOT FOUND" error from my BIOS. Annoying but
survivable.

For the real meaning of the DISK CHANGE, READY, MOTOR ON and DENSITY SELECT
lines you have to consult the manual for the drive. I have access to manuals
for a few drives (Toshiba, Sony, Shugart, Panasonic), mail me if you need help
with those. Most drives can have all these critical pins changed via jumpers.
On some 5.25" HD drives there are about 30 or 40 jumpers, so setting up the
drive can turn into a real science, and can probably not be accomplished by
trial and error any more.

Some 8" disk drives can have very elaborate selection schemes (binary select,
where SELECT 0 is a global select and SELECT 1-3 are a 3-bit drive address), or
they use weird combinations of select lines instead of SIDE SELECT to switch
between the two sides. If one really >has< to emulate such madness, a few TTL
gates might be required.

One more weirdness: 3.5" drives usually have non-removable terminators in the
kOhm range, instead of the good old ~200 Ohm removable terminators. I have
never tried using just a 3.5" drive at the end of a long flat cable with no
other termination, and it probably is a real bad idea since reflections must be
fierce.

I never understood why people would want to replace 8" drives. They are
friendly, solid, reliable, and teach respect for computers. I have lost a 3.5"
disk drive in the pile of stuffs on my desk; that could never happen to an 8"
drive. Installing a 2-drive set with power supply is very good exercise and
builds character and muscles. They heat the office, and provide a pleasant
acoustic background (I call mine the "afterburner"). You will never forget a 8"
floppy in your shirt pocket and run it through the washer by mistake. In
emergencies, you can remove the floppy from the jacket, and for a limited time
read and write on a naked floppy. If you screw up aligning the heads really
bad, a little pile of brown dust will form under the drive and you hear a
screaching noise; on 3.5" disk drives you can't even align the heads yourself. 
On the other hand, they don't have that cute sliding shutter which one can play
with. Well, guess I can't have everything.
-- 
Ralph Becker-Szendy                          UHHEPG=24742::RALPH (HEPNet,SPAN)
University of Hawaii                              RALPH@UHHEPG.PHYS.HAWAII.EDU
High Energy Physics Group                                  RALPH@UHHEPG.BITNET
Watanabe Hall #203, 2505 Correa Road, Honolulu, HI 96822         (808)956-2931

slsw2@cc.usu.edu (02/11/91)

In article <11375@uhccux.uhcc.Hawaii.Edu>, ralph@uhheph.phys.hawaii.edu (Ralph Becker-Szendy) writes:
> Since someone asked what the method is to hook up a 3.5" floppy disk drive
> instead of an 8" drive, here is a listing of pin-numbers on the cables. A 5.25"
> HD or 3.5" HD floppy is electrically nearly equivalent to a 8" DD floppy. The
> only major difference is that some 5.25" HD and all 3.5" HD floppy drives spin
> at 300RPM (instead of 360RPM, what it used to be for 8" floppies), so you can
> pack more data onto each track (but you don't have to). Some 5.25" HD drives
> switch to 360 RPM when switching to HD mode though (usually jumper selectable).

Actually, there is a major difference between 3.5" and 5.25" that I came 
across in a floppy controller data sheet that probably explains why my attempt
to replace the 8" drive on my system failed miserably; I've not tried again
using the new information, so I can't say if that's the problem.

3.5" drives use CMOS drivers for the signals that they drive, and therefore
cannot drive the terminators generally used for 5.25" and 8" drives. The
data sheets for the Intel 82077A therefore recommend that if you are going to
mix 3.5" drives with 5.25" drives, you include a buffer between the 3.5" and
5.25" drives so that the drive doesn't have to try to drive the 5.25"
terminator.
-- 
===============================================================================
Roger Ivie

35 S 300 W
Logan, Ut.  84321
(801) 752-8633
===============================================================================

slsw2@cc.usu.edu (02/11/91)

In article <11375@uhccux.uhcc.Hawaii.Edu>, ralph@uhheph.phys.hawaii.edu (Ralph Becker-Szendy) writes:
> Pin   Pin   Signal
> 50p   34p
> ==================
> 12     4    Disk change (may be on other pins too, not very standarized)
> 22    34    Ready (some 5.25 and 3.5" drives have no ready)
>       16    Motor on (only some 5.25 and 3.5" drives)

Oh yeah. IBM put Diskette Changed on pin 34, so any 5.25" or 3.5" HD drive
that you buy these days does that. I've also never seen a 5.25" drive that
ignored Motor On, even way back in the old days.

I've also noticed that PC drive vendors tend to be soldering in the drive-
select jumpers these days. I was looking for a drive for a CP/M machine, and
had to look for a while before finding one that wasn't hardwired to be drive
1; that's because IBM put this funky twist in their drive cable that swaps
the drive selects around and gives then individual control over the drive
motors. Of course, the BIOS just goes ahead and turns on both drives at the
same time...
-- 
===============================================================================
Roger Ivie

35 S 300 W
Logan, Ut.  84321
(801) 752-8633
===============================================================================

gary@cdthq.UUCP (Gary Heston) (02/13/91)

slsw2@cc.usu.edu writes:
> 3.5" drives use CMOS drivers for the signals that they drive, and therefore
> cannot drive the terminators generally used for 5.25" and 8" drives. The
> data sheets for the Intel 82077A therefore recommend that if you are going to
> mix 3.5" drives with 5.25" drives, you include a buffer between the 3.5" and
> 5.25" drives so that the drive doesn't have to try to drive the 5.25"
> terminator.

Shucks, pull out the terminator. If they're driving the circuit with
CMOS, you won't need it--they were designed originally for open-
collector drivers, which CMOS isn't. If all else fails, build a 
replacement terminator out of higher value resistors. 

It's been my experience that drives and controllers work fine even
when double terminated, both hard and floppy. We've never had a
problem at work (we build ISA systems) with weak signals from 3.5"
drives, and we intermix them constantly.

Gary Heston, at home.....
gary@cdthq.uucp

Sprague.Wbst311@XEROX.COM (02/13/91)

>>       16    Motor on (only some 5.25 and 3.5" drives)

> I've also never seen a 5.25" drive that
> ignored Motor On, even way back in the old days.

While I agree with the satement, note that some drives have a jumper which
allowes you to use the Drive Select signal for Motor On.  This allows the drive
to ignore Motor On.

> I've also noticed that PC drive vendors tend to be soldering in the drive-
> select jumpers these days.

ARRGH!

				~ Mike  (Sprague.Wbst311@Xerox.Com)

ralph@uhheph.phys.hawaii.edu (Ralph Becker-Szendy) (02/19/91)

In article <1991Feb11.092925.46847@cc.usu.edu> slsw2@cc.usu.edu writes:
... stuff deleted ...
>3.5" drives use CMOS drivers for the signals that they drive, and therefore
>cannot drive the terminators generally used for 5.25" and 8" drives. The
>data sheets for the Intel 82077A therefore recommend that if you are going to
>mix 3.5" drives with 5.25" drives, you include a buffer between the 3.5" and
>5.25" drives so that the drive doesn't have to try to drive the 5.25"
>terminator.
... stuff deleted ...

Maybe sometimes true. But not always. I was worrying about the termination 
problems when I first added my new 3.5" drive to the existing system. But it 
turns out to work just fine with a Sony 3.5" drive. Currently my setup is: 
Controller in the middle of two cables. On one side about 10' twisted-pair 
cable to an 8" drive, properly terminated at the drive. On the other side 
maybe 2' of regular flat cable to a 3.5" drive and two 5.25" drives, again 
properly terminated at the last 5.25" drive. I know it shouldn't work, it is 
doubly terminated, and NONE of the drive outputs should handle the load of
two terminators, least of which the 3.5" drive. On the other hand, CMOS output
drivers (see 74HCxxx) are very good at driving resistive loads, sometimes I get
the feeling the are more reliable at that than LS drivers. So trial and error
is in order.

-- 
Ralph Becker-Szendy                          UHHEPG=24742::RALPH (HEPNet,SPAN)
University of Hawaii                              RALPH@UHHEPG.PHYS.HAWAII.EDU
High Energy Physics Group                                  RALPH@UHHEPG.BITNET
Watanabe Hall #203, 2505 Correa Road, Honolulu, HI 96822         (808)956-2931

del@fnx.UUCP (Dag Erik Lindberg) (02/22/91)

In article <11375@uhccux.uhcc.Hawaii.Edu> ralph@uhheph.phys.hawaii.edu (Ralph Becker-Szendy) writes:
-Since someone asked what the method is to hook up a 3.5" floppy disk drive
-instead of an 8" drive, here is a listing of pin-numbers on the cables. A 5.25"
-HD or 3.5" HD floppy is electrically nearly equivalent to a 8" DD floppy. The
		.......
-
-If the 5.25" or 3.5" drive really needs the MOTOR ON signal (most newer ones
-can be jumpered to do the motor control themselves), just strap it to ground,
-that keeps the engines under full steam.

**** DON'T DO THIS !!!! ****
Most newer 5/25" or 3/5" drives do not onload the head when the drive is not
being access.  If you leave the drive running continuously, you will wear
out the diskette at an alarming rate.  I have several souvenier diskettes
where nearly all the magnetic oxide has been worn off the diskette from
being left in drives that were running continuously.  You could practically
read a newspaper through one of them.

If you can't do it properly with a real motor control, then either a) tie
motor-on to the drive-select and always delay 1/2 second after selecting
before accessing drive, or b) build a small timer that will pull down the
motor on line for 20-60 seconds after a select.

-Some 8" disk drives can have very elaborate selection schemes (binary select,
-where SELECT 0 is a global select and SELECT 1-3 are a 3-bit drive address), or
-they use weird combinations of select lines instead of SIDE SELECT to switch
-between the two sides. If one really >has< to emulate such madness, a few TTL
-gates might be required.

Or a single PAL device.

-One more weirdness: 3.5" drives usually have non-removable terminators in the
-kOhm range, instead of the good old ~200 Ohm removable terminators. I have
-never tried using just a 3.5" drive at the end of a long flat cable with no
-other termination, and it probably is a real bad idea since reflections must be
-fierce.

I've had no trouble with 6 foot cables.

-I never understood why people would want to replace 8" drives. They are
-friendly, solid, reliable, and teach respect for computers. I have lost a 3.5"
-disk drive in the pile of stuffs on my desk; that could never happen to an 8"
-drive. Installing a 2-drive set with power supply is very good exercise and
-builds character and muscles. They heat the office, and provide a pleasant
-acoustic background (I call mine the "afterburner"). You will never forget a 8"

Computers and human beings should not share the same room when either is
in operation. :-)

-- 
del AKA Erik Lindberg                             uunet!pilchuck!fnx!del
                          Who is John Galt?