27 Jan
2013
27 Jan
'13
11:23 p.m.
Hi,
I have been reading about LIF format floppies, in particular as used on the HP Series 80
and 200.
I found this description:
ftp://ftp.hpmuseum.org/lif/lifutil/lif_over.txt
an overview on MoHPC:
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/articles.cgi?read=24
and the HPDir utility:
http://www.hp9845.net/9845/projects/hpdir/index.html
Can anyone explain please how bad sectors are marked as 'spare' under LIF? The LIF
overview states:
"LIFUTIL cannot recognize "spared tracks" on a LIF disk. To explain ...
when you format a LIF disk, not all the tracks may be good; so the
formatting process marks them as "bad" or "spared" and they are not
used.
LIFUTIL is incapable of distinguishing a spared track from any other;
so it will suck in the spared track along with any other tracks in the
file or directory, leading to corrupted files or catalog reads."
What I can't find is a description of the bytes encoded on the disk to mark a track as
spare. One possibility is this is dealt with by the HP disk controller firmware, and so
hidden to the operating system, which sees a LIF directory as per the above descriptions,
without seeing the spare tracks.
I am trying to find some floppies with spare tracks and then analyse them on a PC,
hopefully someone has been there already and can shed some light on the mechanism used.
Regards,
John
28 Jan
28 Jan
10:22 a.m.
On Sun, 27 Jan 2013, John S wrote:
What I can't find is a description of the bytes
encoded on the disk to
mark a track as spare. One possibility is this is dealt with by the HP
disk controller firmware, and so hidden to the operating system, which
sees a LIF directory as per the above descriptions, without seeing the
spare tracks.
Now that's simple. The disc (HP don't say disk) format is described in
some HP-IB floppy drive manual. A bad track is marked by $FF as the
cylinder value in the ID field of each sector of a track. When reading,
you just skip to the next track as long as the cylinder value is $FF. For
example, if you've read (physical) cylinder 37 side 0 and want to read
side 1, when the cyl. value if $FF, you skip to cylinder 38 side 0. But
attention, the information in the ID field says cylinder 37 head 1
instead. So you have to distinguish between physical accesses and logical
tracks.
I am trying to find some floppies with spare tracks
and then analyse
them on a PC, hopefully someone has been there already and can shed some
light on the mechanism used.
You are right with your above assumption that the drive's firmware is
automatically doing bad track handling. I have several 3.5" floppies that
have one or two bad tracks and are otherwise unusable as normal DOS
disks. I use those floppies with the 9122C drive that I have on the
HP1000.
Christian
2:54 p.m.
On Mon, Jan 28, 2013 at 2:22 AM, Christian Corti
<cc at informatik.uni-stuttgart.de> wrote:
Now that's simple. The disc (HP don't say disk) format is described in some
HP-IB floppy drive manual. A bad track is marked by $FF as the cylinder
value in the ID field of each sector of a track. When reading, you just skip
to the next track as long as the cylinder value is $FF. For example, if
you've read (physical) cylinder 37 side 0 and want to read side 1, when the
cyl. value if $FF, you skip to cylinder 38 side 0. But attention, the
information in the ID field says cylinder 37 head 1 instead. So you have to
distinguish between physical accesses and logical tracks.
http://www.hpmuseum.net
9895A_DiscDrive_ServiceManual_09895-90030_173pages_Feb81.pdf
2-12 Theory of Operation
Track Numbering. Each track has a physical address as previously
described. There is also a logical track address associated with each
good track. The logical track address is written in the ID field of
each sector on the track. If a disc has no bad tracks, the logical
address of a track is the same as the physical address.
A disc with N bad tracks can be made to look like a 77-N track disc
with no bad tracks. To do this, the logical track address stored in
the ID field of each sector of every bad track is set to !FF. Tracks
of this type are known as invisible tracks. All visible tracks are
then sequentially assigned logical track numbers. Logical Track 0 is
the outermost good track, not necessarily physical Track 0.
29 Jan
29 Jan
12:19 a.m.
I am trying to find some floppies with spare tracks
and then analyse them o=
n a PC, hopefully someone has been there already and can shed some light on=
the mechanism used.
I'd also be interested in knowign this. One of the docuemtned bugs in my
LIF utilities for linus is that they don't handle the bad track
repalcements. Quite simply all the floppies I've seen have been perfect.
-tony
4364
days inactive
4366
days old
3 comments
4 participants
participants (4)
-
ard＠p850ug1.demon.co.uk -
cc＠informatik.uni-stuttgart.de -
glen.slick＠gmail.com -
john_a_s2004＠hotmail.com