Specifications
• I
-
-.
J
~
••
J
.
'.
("
...
-'
-:
.'
..
...
6.1
TRACK
fORMAT
SECTION
VI
RECORDING FORMAT
The
pupose
of a format is
to
organize a
data
track Into smaller, Mquentlally numbered blocks
of
data
called sectors.
..
The
706/712'5
format is a soft
sectored
type. which
means
that
the
beginning
of
each
sector
is
defined by a prewrit-
ten identification
(10)
field which
contains
the
physical
Hctor
'address, plus cylinder
and
head
information.
The
10
field
is
then
followed
by
a
user
data
field. '
The
soft sectored format is a slightly modified version of
the
IBM system
34
double
density format commonly used
on 8-inch floppy disk drives. The
encoding
method
used here
Is
modified
&equency
modulation
(MFM).
Figure 6-1 shows
each
track divided Into
32
sectors. Each sector
has
a
data
field
256
bytes
in length. However,
if
I
sector interleaving
is
used,
Gap
4
can
be
reduced
to
32
bytes
minimum
to
accommodate
33
sectors
per
track. If
sector interleaving is
not
used,
and
33
sectors are desired,
Gap
3 is
25
bytes
and
Gap
4
becomes
26
bytes
minimum.
1~~fXJ1-----------------fl\..-------------------_rL
"-
(
,
G»'
SYM:
22
a '.e
2M
•
USER
DATA
~
10
....
..
..
-
.~
..
..
•.
.-
-
..
..
.
-
w~ITE
UPDATE
--------------
NOTES:
1.
MINIMUM
TRACK CAPACITY -
1CM11IYTES.
2. WRITE TO READ
RECOVERY
TIME • • MICROSECONDS.
3. HEAD SWITCHING TIME • • "MICROSECONDS•
FIGURE
8·1.
TRACK
FORMAT
The
beginnings
of.
both
the
10
field
and
the
data
field
are
flagged by
unique
characters
called
address
marks.
An
address mark is
two
bytes in length.
The
first byte
is
always
ali
"At"
data
pattern.
This
is
followed
by
either
an
"FE" pattern which is
the
pattern
used
to
define
an
10
address
mark.
or
an
MFS··
which
is
a
data
address
panern.
The
MAl"
pattern
violates
the
encode
rules for
MFM
by
omitting
on
'clock
transition
between
bits 4
and
5.
This
makes
the
address
mark
panen unique
to
any
other
serial bit combination.
AllID
and
data
field
are
followed
by
a 16·bit cyclic
redundancy
check
(CRC)
character
used for
data
verification.
Each
eRe
polynomial is
unique
for a particular
data
panern.
Surrounding
the
10
and
data
fields
are
gaps
called
Intenecord
gaps.
6.2
GAP LENGTH CALCULATIONS
C,
6.2.1 Gap 1
The
purpose
of
Gap 1 is
to
provide a
head
switching recovery period
10
that.
when
switching from
one
track to
another,
sequential sectors
may
be
read
Without waiting
the
rotational latency time.
In
~ddition.
Gap
1 allows
8·1