7 Jul
2009
7 Jul
'09
10:33 a.m.
(Catching up on my email after the holidays)
Programs like Media Master can read and write Kaypro disks on a PC, but in order for a
disk to be readable, it must be formatted on the PC (same goes for a Zorba disk).
The following is the explanation from the Media Master Read.Me file:
<quote>Both the Kaypro 2x,4,10 (DSDD) and Zorba 40 trk (DSDD) formats have a
peculiarity that prevents you from using a disk that is created on
the "native" computer.? In both cases, the only way to transfer files
is to first format a transfer disk in the Kaypro or Zorba format using
MEDIA MASTER and your computer.? Then use the Kaypro or Zorba computer
to transfer files from the "native" disks to the disk you created with
MEDIA MASTER.? Then using MEDIA MASTER, you can transfer the files to
any format you wish.
Another way of viewing this problem is that disks created on the
Kaypro 2x, 4, or 10 can only be used on the Kaypro 2x, 4, or 10, but
a Kaypro 2x, 4, or 10 disk created on your computer with MEDIA MASTER
can be used on both your computer and the Kaypro computer.</quote>
BTW, I have copies (for MS-DOS) of Media Master and Convert (1984, Selfware, Inc., Charles
W Woodford). Contact me off-list.
Bob
?
Message: 7
Date: Thu, 02 Jul 2009 23:20:37 -0700
From: "Chuck Guzis" < cclist at sydex.com >
Subject: RE: Kaypro System disks?
On 2 Jul 2009 at 22:19, David Griffith wrote:
Here's a point I was never quite clear on...?
Are disks readable
across different machines?? For instance, take a disk that was
formatted and written on using a Kaypro 10.? Is that disk readable on
a Kaypro IV or Kaypro 4?? How about reading a single-sided Kaypro II
disk on one of the double-sider machines?
Natively, it was highly ROM-dependent as others have noted.? However,
I seem to remember a version of one of the interchange programs
(Media Master?) for the Kaypro that could be used to read a wide
range of formats.? If you had a 96 tpi drive, the possibilities were
even greater.
Fred might remember...
--Chuck 
7 Jul
7 Jul
12:38 p.m.
On 7 Jul 2009 at 17:33, feldman.r at comcast.net wrote:
<quote>Both the Kaypro 2x,4,10 (DSDD) and Zorba
40 trk (DSDD) formats
have a peculiarity that prevents you from using a disk that is created
on the "native" computer.? In both cases, the only way to transfer
files is to first format a transfer disk in the Kaypro or Zorba format
using MEDIA MASTER and your computer.? Then use the Kaypro or Zorba
computer to transfer files from the "native" disks to the disk you
created with MEDIA MASTER.? Then using MEDIA MASTER, you can transfer
the files to any format you wish.
Well, yes and no. The Kaypro formats, at least in its non-Advent ROM
incarnations, 10 512-byte sectors per track, leaving out the IDAM on
each track. The NEC 765 and its derivatives go blind for a period
after the index (the original 765 and 8272 are blind for longer than
the later versions of the chip) long enough to avoid seeing the IDAM.
The problem with the Kaypro is that there is no IDAM on Kaypro-
formatted diskettes, so the IAM of the first sector isn't seen and
only 9 sectors instead of 10 are visible to the 765.
There are several ways around this. One, as mentioned, is to format
the Kaypro diskette on the PC and write files to it, but that assumes
that you still have the Kaypro around to write the files. Another
commonly used method is to tweak the spindle speed a bit slower so
that the first sector IAM just slips past the "window". Another is
to isolate/disconnect the INDEX/ signal from the floppy drive so that
it isn't seen at all and the 765 always has its eyes open.
Finally, one can replace the uPD765 with a uPD7265, which leaves out
the IDAM nonsense and Kaypro disks will read just fine.
--Chuck
12:56 p.m.
On 7 Jul 2009 at 17:33, feldman.r at comcast.net
wrote:
> <quote>Both the Kaypro 2x,4,10 (DSDD) and Zorba 40 trk (DSDD) formats
> have a peculiarity that prevents you from using a disk that is created
> on the "native" computer.? In both cases, the only way to transfer
> files is to first format a transfer disk in the Kaypro or Zorba format
> using MEDIA MASTER and your computer.? Then use the Kaypro or Zorba
> computer to transfer files from the "native" disks to the disk you
> created with MEDIA MASTER.? Then using MEDIA MASTER, you can transfer
> the files to any format you wish.
I think that the problem that he is referring to is the invalid head
number in the sector headers on the second side. It's ignored by WD179x,
but requires going below INT13h on PC.
On Tue, 7 Jul 2009, Chuck Guzis wrote:
Well, yes and no. The Kaypro formats, at least in its
non-Advent ROM
incarnations, 10 512-byte sectors per track, leaving out the IDAM on
each track. The NEC 765 and its derivatives go blind for a period
after the index (the original 765 and 8272 are blind for longer than
the later versions of the chip) long enough to avoid seeing the IDAM.
The problem with the Kaypro is that there is no IDAM on Kaypro-
formatted diskettes, so the IAM of the first sector isn't seen and
only 9 sectors instead of 10 are visible to the 765.
There are several ways around this. One, as mentioned, is to format
the Kaypro diskette on the PC and write files to it, but that assumes
that you still have the Kaypro around to write the files. Another
commonly used method is to tweak the spindle speed a bit slower so
that the first sector IAM just slips past the "window". Another is
to isolate/disconnect the INDEX/ signal from the floppy drive so that
it isn't seen at all and the 765 always has its eyes open.
A small piece of opaque tape (write protect tab) over the index hole will
get around it, but then most errors are mis-reported as 80h ("Not ready")
Finally, one can replace the uPD765 with a uPD7265,
which leaves out
the IDAM nonsense and Kaypro disks will read just fine.
1:24 p.m.
On 7 Jul 2009 at 12:56, Fred Cisin wrote:
I think that the problem that he is referring to is
the invalid head
number in the sector headers on the second side. It's ignored by
WD179x, but requires going below INT13h on PC.
I've got a copy of Media Master somewhere, but I don't think it used
INT 13h services. OTOH, the 10x512 format is a problem no matter
what you use on the PC.
--Chuck
8 Jul
8 Jul
2:27 a.m.
On Tue, 7 Jul 2009, Chuck Guzis wrote:
Well, yes and no. The Kaypro formats, at least in its
non-Advent ROM
incarnations, 10 512-byte sectors per track, leaving out the IDAM on
each track. The NEC 765 and its derivatives go blind for a period
after the index (the original 765 and 8272 are blind for longer than
the later versions of the chip) long enough to avoid seeing the IDAM.
I think this is wrong. First, noone needs the Index Address Mark, second I
can read my Kaypro II and Kaypro 4 disks without problems on an old 1986
HP Vectra with ?PD765 FDC. At least ImageDisk doesn't seem to have any
problem with these disks.
The Kaypro disks are 10 sectors/512 bytes per sector.
The Kaypro uses an interleave of 1:4 with sectors numbered
0,8,3,6,1,9,4,7,2,5; the Kaypro 4 uses DS disks and continues with
10,18,13,16,11,19,14,17,12,15 on the other side.
The problem with the Kaypro is that there is no IDAM
on Kaypro-
formatted diskettes, so the IAM of the first sector isn't seen and
only 9 sectors instead of 10 are visible to the 765.
The 765 does not need the IDAM. And the original IBM floppy format (3742
and System/34) doesn't even mention an IDAM. Personally I have not seen a
system that actually needs an IDAM.
Christian
9:26 a.m.
On 8 Jul 2009 at 11:27, Christian Corti wrote:
The 765 does not need the IDAM. And the original IBM
floppy format
(3742 and System/34) doesn't even mention an IDAM. Personally I have
not seen a system that actually needs an IDAM.
Your 1988 PC is too late.. I did not say that the 765 needs the
IDAM, but that it was "blind" during the IDAM period.
The original 16-64K 5150 was equipped with an Intel 8272 FDC, which
is essentially Intel's rebranding of the NEC uPD765. I still have
the FDC from my 5150 and there it is--marked "P8272".
Let's go to the datasheet for the 8272--I refer you to the 1984 Intel
"Microprocessor and Peripheral Handbook", page 6-226. Look at the
section labeled "8272A Enhancements":
"On the 8272A, after detecting the index Pulse, the VCO Sync output
stays low for a shorter period of time. See figure 4A."
On the 8272A, the time given for the length of the VCO sync is 527 uS-
-I seem to recall (but I can check) that it was over 1 mS on the
8272.
The long VCO sync period was a headache for reading 10 sector formats
on the original 5150. Note that it's not a problem for disks
formatted on the PC--the 8272/uPD765 always writes an IDAM--there's
no way to suppress it.
But that's not true of the WD17xx/27xx floppy controllers. One can
just about put whatever one wants down as a track format. Indeed, in
WD179x datasheet (page 18) it is stated that "In addition, the Index
Address Mark is not required for operation by the WD179x."
Not all versions of the Kaypro format fail on the original
8272/uPD765, but enough do to warrant Media Master's caution.
10x512 sector Cromemco diskettes are perhaps the worst in this
regard; indeed many can't be read without "tweaking" even on
controllers equipped with the 8272A/765AC parts.
As far as the head ID being the cause for the statement, I don't
think that was the issue. There are other formats that involve non-
standard head ID fields (e.g. Xerox 16-8 and 820) that woiuld have
tripped Media Master up but IIRC didn't.
Since one had to lay out the whole track to be formatted on the
WD179x chips, it was a shortcut to use the same layout for both sides
of a diskette and so was quite common.
--Chuck
10:56 a.m.
Chuck Guzis wrote:
10x512 sector Cromemco diskettes are perhaps the worst
in this
regard; indeed many can't be read without "tweaking" even on
controllers equipped with the 8272A/765AC parts.
This is actually a side-note to the topic, since I have no idea what
the controller chip is, but...
I have an old Compaq Presario (75MHz Pentium) whose on-board FDC at
least reads Cromemco diskettes without problems and writes readable,
bootable diskettes for a Kaypro 2 with factory ROMs. I can look up the
model # tonight if anyone wants to know.
Doc
12:02 p.m.
On 8 Jul 2009 at 12:56, Doc Shipley wrote:
I have an old Compaq Presario (75MHz Pentium) whose
on-board FDC at
least reads Cromemco diskettes without problems and writes readable,
bootable diskettes for a Kaypro 2 with factory ROMs. I can look up the
model # tonight if anyone wants to know.
I count something like 7 common Cromemco formats--and, as Fred notes,
there was tremendous variability among them.
On a P75 system, the FDC has been buried in the entrails of of "Super
I/O" chip or even buried in a corner of a "Southbridge chip", such as
an Intel 82371.
--Chuck
1:05 p.m.
On Wed, 8 Jul 2009, Doc Shipley wrote:
I have an old Compaq Presario (75MHz Pentium) whose
on-board FDC at
least reads Cromemco diskettes without problems and writes readable,
bootable diskettes for a Kaypro 2 with factory ROMs. I can look up the
model # tonight if anyone wants to know.
Those "NEW" ones (it's all relative) have a lot fewer problems. The
"flash blindness" after index is much shorter. Also, some of them have
single density support enabled. The original 765 based IBM disk
controller has hard wired for Double Density. Some of the later ones,
such as the 37C65 can be switched in software. Unfortunately, some new
machines have very demented BIOS, and in some cases won't even support a
second floppy.
--
Grumpy Ol' Fred cisin at xenosoft.com
11:33 a.m.
On Wed, 8 Jul 2009, Chuck Guzis wrote:
10x512 sector Cromemco diskettes are perhaps the worst
in this
regard; indeed many can't be read without "tweaking" even on
controllers equipped with the 8272A/765AC parts.
I found LOTS of variation in Cromemco diskettes from different machines.
Since I didn't have the machines, I wasn't able to do much checking.
But, some of the Cromemco diskettes were the worst that I've seen.
As far as the head ID being the cause for the
statement, I don't
think that was the issue. There are other formats that involve non-
standard head ID fields (e.g. Xerox 16-8 and 820) that woiuld have
tripped Media Master up but IIRC didn't.
Since one had to lay out the whole track to be formatted on the
WD179x chips, it was a shortcut to use the same layout for both sides
of a diskette and so was quite common.
The fact that MM mentioned DSDD Kaypro and DSDD Zorba, but did NOT mention
SSDD Kaypro, nor even Cromemco, prompts me to disagree.
BUT, . . .
We are opining over a fact, speculating about somebody else's actions.
Why isn't the author HERE? Anybody in touch?
IIRC, his name was something like Mark? Graybill??
BTW, I was lucky. My first 5150 (~November 1981 - ~January 1982)
read single sided Kaypro with very little difficulty. My B: drive was
an old beat up Tandon, running at 300.5 RPM. It was the first
format that I succeeded at on the PC. (First alien format on TRS80 was
Osborne SSSD)
--
Grumpy Ol' Fred cisin at xenosoft.com
11:56 a.m.
On 8 Jul 2009 at 11:33, Fred Cisin wrote:
The fact that MM mentioned DSDD Kaypro and DSDD Zorba,
but did NOT
mention SSDD Kaypro, nor even Cromemco, prompts me to disagree.
The Zorbas that I've seen (both 48 and 96 tpi DSDD) have 0 as the
head ID for both sides.
http://www.rednil.de/pool/ag/ag6303.htm indicates that MM was
available in versions for Osborne and Kaypro as well as for PC.
We are opining over a fact, speculating about somebody
else's actions.
Why isn't the author HERE? Anybody in touch? IIRC, his name was
something like Mark? Graybill??
Good point. It seems to yet be available through Echoscan:
http://www.echoscaninc.com/product.cgi?code=MTkwLUQwMS0zNg==
--Chuck
12:49 p.m.
On Wed, 8 Jul 2009, Chuck Guzis wrote:
The Zorbas that I've seen (both 48 and 96 tpi
DSDD) have 0 as the
head ID for both sides.
To further add to the confusion, . . .
Several authors of such programs would use the competing products to
generate sample disks to analyze. I had a couple of disks from a
"Goupil". But the guy's handwriting was such that I thought that he had
written "Groupil". It was most of a year before I corrected the spelling.
Hmmm. A couple of programs support "Groupil"; where did they get the
disk from? The guy that sent them to me SAID that he hadn't sent any to
anybody else. There was also somebody who made an I.N.W. disk based on my
L.N.W. In addition, there are a very small number of formats that I put
in specifically for the developers of the machines, some of which never
made it to commercial market! Not a DELIBERATE Mountweazel, but, . . .
Since WD didn't CARE about the head number field, all of the disks
formatted by XenoCopy had the correct content in the field, "poached"
formats might not always have all of the correct weirdities.
--
Grumpy Ol' Fred cisin at xenosoft.com
1:44 p.m.
On 8 Jul 2009 at 12:49, Fred Cisin wrote:
Several authors of such programs would use the
competing products to
generate sample disks to analyze. I had a couple of disks from a
"Goupil". But the guy's handwriting was such that I thought that he
had written "Groupil". It was most of a year before I corrected the
spelling. Hmmm. A couple of programs support "Groupil"; where did
they get the disk from? The guy that sent them to me SAID that he
hadn't sent any to anybody else. There was also somebody who made an
I.N.W. disk based on my L.N.W. In addition, there are a very small
number of formats that I put in specifically for the developers of the
machines, some of which never made it to commercial market! Not a
DELIBERATE Mountweazel, but, . .
I hear you. That's why, for example, 22Disk has *two* 400K formats
for the Kaypro 2X. One with head IDs of 0/0 and the other with 0/1.
Both legitimate Kaypro system disks as far as I can tell.
There's little rhyme or reason here. The Future FX-20 uses head ID 0
for the first side and 4 for the second side. The Sharp MZ80B has
head IDs of 0 and 1, but starts the data on the "top" side of the
floppy, rather than the bottom side.
That's not by any means the extent of the whackiness (some would say
originality) in diskette formats that has come through here. We
offered a *special* 22Disk that employs not the static table of the
shareware version, but one that uses a table of procedure calls to
perform the necessary mapping for those really odd formats (e.g.
different head IDs on the same side with sector numbers that aren't
the same from track to track).
You wondered if any of those programmers had a life outside of the
workplace...
--Chuck
9 Jul
9 Jul
2 a.m.
On Wed, 8 Jul 2009, Chuck Guzis wrote:
Your 1988 PC is too late.. I did not say that the 765
needs the
IDAM, but that it was "blind" during the IDAM period.
My PC is over 2 years older (1985/86), and it actually has an i8272A.
You could have been more specific about the different revisions of that
IC ;-)
Christian
5646
days inactive
5648
days old
13 comments
5 participants
participants (5)
-
cc＠informatik.uni-stuttgart.de -
cclist＠sydex.com -
cisin＠xenosoft.com -
doc＠vaxen.net -
feldman.r＠comcast.net