Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site abnji.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!mhuxv!abnji!jeff
From: jeff@abnji.UUCP (jeff)
Newsgroups: net.micro
Subject: confused about disk drive specifications
Message-ID: <735@abnji.UUCP>
Date: Wed, 10-Jul-85 15:43:28 EDT
Article-I.D.: abnji.735
Posted: Wed Jul 10 15:43:28 1985
Date-Received: Fri, 12-Jul-85 01:44:53 EDT
Lines: 140

[whurr! clank! clikety-click!]

	I am confused about floppy disk drive specifications.
I see that the major choices are:
	.  single sided single/double density 48 TPI
	.  double sided single/double density 48 TPI
	.  double sided single/double density 96 TPI
full or half height.

	John J. Meshna Jr. (a surplus catalog) has the Mitsubishi
no. 4853 for $175.  It is double sided, double density, 80 tracks
per side.  Is this the same as a DSDD 96 TPI drive?

	The way I understand the specs:
single sided - read/write heads on one side only
	You cannot flip over the disk to use the other side unless
	the sector hole is in the center, or the sector hole is not
	used (as in Apple computers).
double sided - read/write heads on both sides
density - depends on the external controller, thus any drive
	that can handle double density can be used single density.
	Does quad-density mean another doubling of the density per track
	by the controller, as long as the drive can handle it?
	I think density refers to the number of bits per track.

	How do they specify how many tracks the head moves (and is this
necessarily all the tracks on the diskette?)?  I think the TPI rating
and the number of tracks per side are related as follows:
	2.25" outer track radius
      - 1.4373" inner track radius
	------
	.8125" head movement edge to edge
so 48TPI means (48 tracks/inch) (.8125 inches) + 1 = 40 tracks
(a track is a discrete head position), thus the 80 tracks
means 96TPI.

	If a drive claims 160 tracks - is that 80 tracks by 2 sides
at 96 TPI, or 160 tracks per side at 192 TPI?

	Is this a reasonable approximation of reality?

	Let us try to read the following specs:
   mfg.		model	tracks		no.heads
   ====		=====	======		========
a)Shugart	400	35		1
b)Shugart	400	35		1
c)BASF		6108	80 (both sides)	2
d)Pertec	FD 250	70 (80 optional)2
e)MPI		B92	160 (both sides)2
f)CDC		9409T	160 (both sides)2
g)Mitsubishi	4853	160 (both sides)2

Drives a and b could be 48 TPI with the head not traveling all the
way to the hub, or (35 tracks)/(.8125 inches) = 39.4 TPI.
Which is it?

Drive c is 80 tracks total, 40 per side thus 48TPI.

Drive d - Does the stepper mechanism have two modes,
	stepping .8125/(70-1) inches or .8125/(80-1) inches per track?
	With 70 tracks, the TPI is 1/(space between tracks) = 85 TPI.
	With 80 tracks, it is 1/(.8125/(80-1)) = 97 TPI.
	What did I calculate wrong - the head traveling distance?
	How - by changing the lead screw?

Drive e, f, and g are 160 tracks total, 80 per side thus 96 TPI.

In conclusion:
I think it is too easy to confuse tracks total with tracks per side
(total tracks = tracks per side times number of sides).

The TPI rating says the distance between tracks, but not how many the
heads will reach.  It should not be confused with density although
a 96 TPI disk will hold twice the data as a 48 TPI disk.  I wonder
if some of the 'quad density' drives are really double density
at 96TPI as compared to single density at 48 TPI.

The number of tracks the head can position itself is almost standard
(80 for 96 TPI, 40 for 48 TPI) but other possibilities exist.
Does this mean that the tracks for a 35 track and 40 track drive
won't align due to unequal spacing (thus different TPI ratings)?

Density is the number of bits per track, determined by the
controller.  The drive dictates how fast the data can be
read/written thus limiting the density which is otherwise
determined by an external controller.
One unit of measure is the number of bits per inch along
the track.  Most controllers write the same number of bits on the
outer track as the inner track, thus the number of bits per track
is constant.
Since the circumference of the outer tracks is larger than the inner tracks
and assuming the density in bits per inch along the track being
constant, some controllers
therefore put more bits on the outer tracks so the density of
bits per track can vary from the outside to the inner track.

Height has no effect on the drive's operation except that
some full height drives can lift the heads when the disk is
not in use and even shut off the motor.  The half height drives
apparently don't have the room to do this and it makes for a
less expensive head mechanism at the expense of additional
disk wear.

Speed of rotation is constant, isn't it?  That would affect
the density.  Are the speeds (in RPM) standard?
I suppose a disk written at one speed can be read at another speed if
the format is self-clocking and the controller is adaptive enough.
Is wow and flutter a consideration, or does the controller compensate
for speed variations.  Even so, does this affect the reliability of
the drive (a more stable motor meaning less read errors?)?

Density of the medium - the finer the magnetic particles in the coating
the more magnetic fluxes per inch can be stored.  Is the density
uniform in all directions or only in a particular orientation
(such as along the tape's or disk's tracks)?  Tapes are rated
in FCI (flux changes per inch, which determines BPI and TPI).
Why not disks?

There are several mechanisms for moving the heads.  One I have seen
moves the heads via a metal band directly to the motor pulley.
It is quite fast and quiet.  Others use a lead screw, which
needs occasional cleaning and lubrication.  Do any use linear
motors, or is that an unwarranted expense for a floppy drive?
This might appear as a spec on head seek time.

My Teac drives have a 'button type' head.  I have seen DSDD
drives that have small heads on a leaf spring.  Are there others?
Unless you get the manufacturer's specs, this is not usually
part of the drive specification.

My fingers grow weary, so I conclude with a request for someone
knowledgable to offer a dissertation on disk drive specifications.
Thank you for reading all this, and in advance (for whatever).


			+----------------------------------------+
			|  Jeff 'I am sooooooooo confused' Skot  |
			|  at the bewildering ATT IS Somerset    |
			|  {ihnp4 | mcnc | cbosgb} abnji ! jeff  |
			+----------------------------------------+