From: bryce@hoser.berkeley.edu (Bryce Nesbitt)
Subject: Re: noKLICKstart
Date: 11 Sep 87 10:17:42 GMT
Organization: Tubular Transport Devices, Ltd.
Summary: the final word on detecing disk insertion without a 'click'
Keywords: Matsuhita, Mitsumi, Sony, NEC, Chinon, Panasonic, Fujutsu


[ Cross-posted to sci.electronics because it's about electronics,
  though the problem showed up with the Amiga Computer. Please edit the
  group line if your response is not approprate to both groups. ]
[ You need to know that the 90mm disk drives of the Amiga Personal Computer
  make a clicking noise when empty.  You also need to know that 90mm ~=
  3.5 inches ]

Recently I posted a Kickstart patch to turn 1.2 Kickstart into 1.2
noKLICKstart.  I know it was popular because hundreds of people
responeded (Email and otherwise).  Sadly, many of those
complained that it did not stop the click of *their* drives.  What
follows is a technical desciption of what goes on and how to fix
it:

Seems that Amy has been "sleeping around" with a lot of different drive
manufacturers.  No offical word, but my spies report:

Mitsumi
Matshshita
Sony
NEC
Chinon          (Amiga 500)
Panasonic

Which reads as a "whos who" of 90mm drive manufactures.

The reason for the clicking in the first place is to detect disk insertion.
The drives all have a /CHANGE signal, which is latched LOW when a disk is
removed.  This must be a latch, or you would run the possibility of missing
a disk change.  To see if a new disk is in, you must RESET the latch.  This,
for reasons byond my comprehension, is done with the /STEP signal. (BTW:
I know the reasons, I just don't agree.)

So now you know why the drives click.  To prevent it I patched Kickstart
to continutally step OUTWARD instead of forth and back.  When the head
reaches the track zero sensor the drive says "hey", "you can't go that
way".  It resets the /CHANGE latch, but does not step or click.  Saves
wear and tear on the drives, is silent, and is generally a good idea.

"But no" said some manufacturers.  The drives that don't like this are a bit
dim... you try to step the head past zero and they simply let you bash
the head into the stop.  No noise or wear and tear savings here!  These
drives all have ample circutry to do it the first way... they just don't.


Drives that are know to work include----
Sony
NEC
Chinon          (Used in the Amiga 500, single +5 supply, very low power)
Panasonic

Drives that don't----
Matshushita
Mitsumi

Drives that I have received a conflicting report on----
Panasonic (one person says it's the same as Matsushita... unconfirmed)

Drives that are not used in the Amiga, but won't work anyway----
Fujitsu


SOME drives provide an alternate /CHANGE reset signal, often on Pin 1.  But
all the drives that do, *already work fine*.  If you want to get a non
working drive to work you would need to either:

A> Cut /STEP on the drive PCB so it gets to the latch, but not the stepper
motor.  Now add a gate that will disallow /STEP if the track zero sensor
is active and the direction is out.

B> Tap the change sensor on the board and drag it external.  Use /STEP to
reset your external latch (ignore the internal one).  Now add the circuit
to disable /STEP when track zero is active and the direction is out.

Both are simple, but painful to do.  I don't expect that anyone will try
them.


There seem to be no *real* standards in 90mm drives.  Every manufacturer
is a little different.  The /CHANGE signal is the most volitaile of all.
The definitions of Pins 1,2 and 34 are wildly different from manufacturer
to manufacturer.  Such is life...


|\ /|  . Ack! (NAK, EOT, SOH)
{O o} .
 (")    bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce
  U