10 Sep
2012
10 Sep
'12
6:54 p.m.
As long as we're on the subject, here's an interesting article I
saw just today about the performance and data safety aspects of
SATA vs. SAS/FibreChannel/SCSI disks.
http://queue.acm.org/detail.cfm?id=2367378
It's by Marshall McKusick, who is a man who ought to know his
filesystem and disk stuff.
For those who don't like going to web articles, it basically
boils down to the fact that while SATA is perfectly capable of
supporting proper tag queueing, most bottom-dollar drives don't
and basically all other higher-end transports (including SAS,
FibreChannel, and parallel SCSI) do.
There's a lot of other neat stuff in the article, but I thought
it might be interesting while we're bashing PATA and giving
SATA a "meh" around these parts.
- Dave
10 Sep
10 Sep
7:34 p.m.
On 9/10/12 6:54 PM, David Riley wrote:
"From the time of their first availability in the 1950s until about 2010, the sector
size on disks has been 512 bytes."
This isn't even true for SCSI disks, much less all the different kinds before that.
8:26 p.m.
On 10 Sep 2012 at 19:34, Al Kossow wrote:
On 9/10/12 6:54 PM, David Riley wrote:
"From the time of their first availability in the 1950s until about
2010, the sector size on disks has been 512 bytes."
This isn't even true for SCSI disks, much less all the different kinds
before that.
Indeed. I've seen hard drives with sector sizes all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
The author's probably never even seen an IBM 1311 or Bryant 4000.
--Chuck
8:42 p.m.
On Sep 10, 2012, at 11:26 PM, Chuck Guzis wrote:
On 10 Sep 2012 at 19:34, Al Kossow wrote:
I mean, he wrote Berkeley FFS, so I suppose he should have some
idea that there were other sector sizes. I'd be willing to chalk
it up to a brain fart, though it seems like a big oversight. I
hadn't even noticed that particular line; I think the rest of the
article is fairly informative.
- Dave
On 9/10/12 6:54 PM, David Riley wrote:
"From the time of their first availability in the 1950s until about
2010, the sector size on disks has been 512 bytes."
This isn't even true for SCSI disks, much less all the different kinds
before that.
Indeed. I've seen hard drives with sector sizes all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
The author's probably never even seen an IBM 1311 or Bryant 4000. 
9:11 p.m.
On 09/10/2012 11:42 PM, David Riley wrote:
I mean, he wrote Berkeley FFS, so I suppose he should have some
idea that there were other sector sizes. I'd be willing to chalk
it up to a brain fart, though it seems like a big oversight. I
hadn't even noticed that particular line; I think the rest of the
article is fairly informative.
...and I just read who wrote the article. I am completely
flabbergasted. I wonder if he did this to see if anyone would call him
on it? If ANYONE knows, he does. WTF?
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
"From the time of their first availability in the 1950s until about
2010, the sector size on disks has been 512 bytes."
This isn't even true for SCSI disks, much less all the different kinds
before that.
Indeed. I've seen hard drives with sector sizes all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
The author's probably never even seen an IBM 1311 or Bryant 4000. 
9:37 p.m.
On 9/10/2012 8:42 PM, David Riley wrote:
I mean, he wrote Berkeley FFS, so I suppose he should
have some
idea that there were other sector sizes.
I am currently having to use fuseext2
because the linux driver can't
handle sector sizes not related to the page size of the system, or at
least does not handle the 16k size that the netgear / infarant readynas
uses. it is implmented on a sparc and has the 16k size, which Intel
linux will not mount.
11 Sep
11 Sep
12:18 a.m.
David Riley wrote:
I mean, he wrote Berkeley FFS, so I suppose he should
have some idea
that there were other sector sizes.
Berkeley was using the PDP-11 and VAX, both of which almost exclusively
used 512-byte sectors. It's possible that he didn't realize that this
was not universal practice on other contemporary machines. "All the
world's a VAX."
1:50 a.m.
On Tue, 11 Sep 2012, Eric Smith wrote:
David Riley wrote:
...or maybe his technical editor screwed up while cutting the article down
to publication size?
I mean, he wrote Berkeley FFS, so I suppose he
should have some idea
that there were other sector sizes.
Berkeley was using the PDP-11 and VAX, both of which almost exclusively
used 512-byte sectors. It's possible that he didn't realize that this
was not universal practice on other contemporary machines. "All the
world's a VAX." 
10 Sep
10 Sep
9 p.m.
Indeed. I've seen hard drives with sector sizes
all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
The author's probably never even seen an IBM 1311 or Bryant 4000.
Or most of IBM's big disks until the 1990s. CKD "sectors" are variable
length.
--
Will
9:09 p.m.
On 09/10/2012 11:26 PM, Chuck Guzis wrote:
He's absolutely clueless. Drives used in EMC storage arrays, for
example, which usually contain hundreds of drives, are usually (perhaps
always) formatted to 576 bytes/sector.
Most SCSI-protocol drives (whether they use the "SCSI" physical layer
or not) can be formatted to any sector size you want.
The same goes for SMD; the formatting is handled by the controller in
that case.
Still, somehow, people like that manage to get published. I just
don't get it.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
"From the time of their first availability in the 1950s until about
2010, the sector size on disks has been 512 bytes."
This isn't even true for SCSI disks, much less all the different kinds
before that.
Indeed. I've seen hard drives with sector sizes all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
The author's probably never even seen an IBM 1311 or Bryant 4000. 
9:51 p.m.
At 11:09 PM 9/10/2012, Dave McGuire wrote:
He's absolutely clueless. Drives used in EMC
storage arrays, for
example, which usually contain hundreds of drives, are usually (perhaps
always) formatted to 576 bytes/sector.
Extra bytes for sector-level error correction, interpreted outside
of the drive's own hardware?
- John
10:27 p.m.
On 09/11/2012 12:51 AM, John Foust wrote:
At 11:09 PM 9/10/2012, Dave McGuire wrote:
AFAIK it's used for externally-handled ECC.
I typed 576 bytes above; I meant 520 bytes. I'm sitting in a hotel at
the moment, pretty tired from driving, that's my excuse. ;)
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
He's absolutely clueless. Drives used in
EMC storage arrays, for
example, which usually contain hundreds of drives, are usually (perhaps
always) formatted to 576 bytes/sector.
Extra bytes for sector-level error correction, interpreted outside
of the drive's own hardware? 
11 Sep
11 Sep
12:03 p.m.
From: Dave McGuire
Sent: Monday, September 10, 2012 10:28 PM
On 09/11/2012 12:51 AM, John Foust wrote:
> At 11:09 PM 9/10/2012, Dave McGuire wrote:
>> He's absolutely clueless. Drives used in EMC storage arrays, for
>> example, which usually contain hundreds of drives, are usually (perhaps
>> always) formatted to 576 bytes/sector.
>
> Extra bytes for sector-level error correction, interpreted outside
> of the drive's own hardware?
AFAIK it's used for externally-handled ECC.
I typed 576 bytes above; I meant 520 bytes. I'm
sitting in a hotel at
the moment, pretty tired from driving, that's my excuse. ;)
576 sounded fine to me. After all, it's what was used on RA-81 drives
when connected via CFS to KL-10 systems with the CI option installed.
Rich Alderson
Vintage Computing Sr. Systems Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104
mailto:RichA at vulcan.com
mailto:RichA at LivingComputerMuseum.org
http://www.LivingComputerMuseum.org/
12:14 p.m.
On 09/11/2012 03:03 PM, Rich Alderson wrote:
Yup. :-) Freudian slip. EMC drives are formatted to 520, though.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
>>
He's absolutely clueless. Drives used in EMC storage arrays, for
>> example, which usually contain hundreds of drives, are usually (perhaps
>> always) formatted to 576 bytes/sector.
>
> Extra bytes for sector-level error correction, interpreted outside
> of the drive's own hardware?
AFAIK it's used for externally-handled ECC.
I typed 576 bytes above; I meant 520 bytes.
I'm sitting in a hotel at
the moment, pretty tired from driving, that's my excuse. ;)
576 sounded fine to me. After all, it's what was used on RA-81 drives
when connected via CFS to KL-10 systems with the CI option installed. 
10 Sep
10 Sep
9:35 p.m.
Indeed. I've seen hard drives with sector sizes
all over the place.
256 bytes and 1024 bytes weren't unknown on SCSI devices .
I've got a 3?" SCSI floppy drive (yes, a floppy drive with a SCSI
interface!) that reports itself as having 256-byte blocks.
I don't know whether it works as a floppy drive; the OS support for
SCSI I have is broken enough on devices with non-512-byte sectors that
I'm not really in a position to try it out. I also don't really have
much else left that uses 3?" floppies....
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
11 Sep
11 Sep
8:17 a.m.
On Tue, 11 Sep 2012, Mouse wrote:
I've got a 3?" SCSI floppy drive (yes, a
floppy drive with a SCSI
interface!) that reports itself as having 256-byte blocks.
Does it have any provision to reconfigure?
I don't know whether it works as a floppy drive;
the OS support for
SCSI I have is broken enough on devices with non-512-byte sectors that
I'm not really in a position to try it out.
Prob'ly because the guy who wrote that code had just read:
>> "From the time of their first
availability in the 1950s until about
>> 2010, the sector size on disks has been 512 bytes."
9:22 a.m.
I've got a
3?" SCSI floppy drive (yes, a floppy drive with a SCSI
interface!) that reports itself as having 256-byte blocks.
Does it have any
provision to reconfigure? [...]; the OS
support for SCSI I have is broken enough on devices
with non-512-byte sectors that [...]
Prob'ly because the guy who wrote that
code had just read:
>>> "From the time of their first availability in the 1950s until
>>> about 2010, the sector size on disks has been 512 bytes." 
11:48 a.m.
It appears to be an interface board in front of a more
or less ordinary
floppy drive, though I suspect appearances are deceptive, since I don't
see any power feed to the front-end board except possibly over what
looks like an ordinary floppy interface - but an ordinary floppy
interface does not supply power from the drive to the ribbon cable.
It's not unheard-of. Typically, the odd-numbered pins were used for
power. On a 'normal' (as you called it) floppy drive, all are grounds. On
a drive with power o nthe ribbon cable, theyr'e split into 3 groups for
+5V, ground and +12V.
The SOny drive used in the HP9114B (the half-height drive that shares
parts with an Apple 800K drive) has soldered SMD jumpers to select either
'all grounds' and a spearate power connector or power over the ribbon
vable, the 9144B uses the latter option. 'My' scheamtics for this unit
should show it.
-tony
12:18 p.m.
On Tue, 11 Sep 2012, Mouse wrote:
>> I've got a 3?" SCSI floppy drive
(yes, a floppy drive with a SCSI
>> interface!) that reports itself as having 256-byte blocks.
> Does it have any provision to reconfigure?
At first I thought there was no model number marking,
but on careful
inspection I do find a sticker indicating it's a TEAC FD-235HS.
ftp.rodents-montreal.org:/mouse/misc/scsi-floppy.jpg is a photo I just
now took of it;
I picked up a 5-1/4" variant of that drive (also a TEAC "S" model) several
years ago and have never been able to find a shred of technical
information on it.
--
2:06 p.m.
I like the odd angle of the chip peeking thru the hole at the bottom of
the photo which is on the floppy drive.
And as far as power, I see that it may have a standard looking 34 pin
connector present in the form of a second rank of connector pins which
appears to be J5 after the 50 for the scsi connector. Then above it is a
20 pin connector array which is labeled J7 I'd suspect pretty standard
pins on the J5, and power, etc. on the J7 20 pin connector, but who knows.
It is interesting. Probably saved design cost for integrating a floppy
controller and drivers into a workstation as someone else said, but you
have to wonder how efficient it was as far as SCSI with the bus
handling,etc. And if one needed to do any of the usual stunts accessing
the drive in odd ways other than sectors, life could get really complicated.
jim
On 9/11/2012 12:18 PM, Steven Hirsch wrote:
ftp.rodents-montreal.org:/mouse/misc/scsi-floppy.jpg
is a photo I just
now took of it;
5:40 p.m.
I like the odd angle of the chip peeking thru the hole
at the bottom
of the photo which is on the floppy drive.
I noticed that. :)
I took the SCSI interface board off the floppy drive and took two more
pictures. To better keep them together, I created a new directory.
ftp.rodents-montreal.org:/mouse/misc/scsi-floppy/ holds the pictures.
assembled.jpg is the same file I mentioned before, just renamed.
apart-1.jpg is a view of the floppy drive and the adapter board,
separated; the floppy drive is in approximately the same orientation
relative to the camera as in assembled.jpg, while the SCSI adapter
board has been rotated a half-turn about a left-right line (and
disconnected from the short ribbon cable to the drive). You can see
the other side of the three 2xN pin arrays from the first picture: 2x25
for the SCSI, 2x17 for the floppy, and 2x11 for...well, I'm
speculating, but my guess is SCSI termination resistors. That oddly
rotated chip is quite clearly visible in this picture.
apart-2.jpg is a closer view of the connector area of the two pieces.
Orientations have been changed so that, again, one piece needs to be
rotated 180 degrees about a left-right line to bring it into
approximately its assembled orientation, but the two connector areas
are much closer to one another.
And as far as power, I see that it may have a standard
looking 34 pin
connector present in the form of a second rank of connector pins
which appears to be J5 after the 50 for the scsi connector.
Yes. That goes to a very short ribbon cable to the floppy drive
(disconnected in the two new pictures, but it's obvious, especially in
about-2.jpg, how it goes together).
Then above it is a 20 pin connector array which is
labeled J7 I'd
suspect pretty standard pins on the J5, and power, etc. on the J7 20
pin connector, but who knows.
J7 isn't power; at least, the board works enough to appear on the SCSI
bus without anything connected to those sockets.
Actually, looking more closely, I think J7 is only half of that, with
J8 being the other half; otherwise, I can't account for seeing both J4
and J8 markings next to the SCSI connector. The RA1 and RA2 markings,
and the white silk-screened line down the middle, also seem to imply
that it's not really a 2x11 connector but rather two 1x11 connectors
that happen to be mounted adjacent to one another. I'm still guessing
optional SCSI termination resistors...especially in view of the etch
runs (some of which are visible, poorly, in assembled.jpg and some
others of which are visible, much better, in apart-2.jpg) connecting
those pins to the SCSI connector.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
6:17 p.m.
On 9/11/12 5:40 PM, Mouse wrote:
TC8569 is a floppy controller
http://www.ic-on-line.cn/view_download.php?id=1029891&file=0009\tc8569a…
there is a ram just above it (U3) (28 pins)
U1 is probably the scsi controller
U4 is probably a prom (28 pins)
guessing U5 is a microcontroller (40 pins)
I like the odd
angle of the chip peeking thru the hole at the bottom
of the photo which is on the floppy drive.
I noticed that. :)
I took the SCSI interface board off the floppy drive and took two more
pictures.
6:24 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 9/11/12 6:17 PM, Al Kossow wrote:
On 9/11/12 5:40 PM, Mouse wrote:
TC8569 is a floppy controller
http://www.ic-on-line.cn/view_download.php?id=1029891&file=0009\tc8569a…
there is a ram just above it (U3) (28 pins)
U1 is probably the scsi controller
U4 is probably a prom (28 pins)
guessing U5 is a microcontroller (40 pins)
I found the hardware spec for the board, which I've copied to bitsavers
I like
the odd angle of the chip peeking thru the hole at the bottom
of the photo which is on the floppy drive.
I noticed that. :)
I took the SCSI interface board off the floppy drive and took two more
pictures.
6:50 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
I found the hardware spec for the board, which
I've copied to
bitsavers
I went to look for it, and...
[Sparkle] 262> ftp ftp.bitsavers.org
Connected to bitsavers.org.
220 ProFTPD 1.3.3e Server (Classiccmp FTP Server) [209.145.140.19]
Name (ftp.bitsavers.org:mouse): anonymous
331 Password required for anonymous
Password:
530 Login incorrect.
Does ftp.bitsavers.org just not support anonymous FTP, or is something
broken?
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
7 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 9/11/12 6:50 PM, Mouse wrote:
Does ftp.bitsavers.org just not support anonymous FTP,
or is something
broken?
there is no anonymous ftp on bitsavers
http://bitsavers.org/pdf/teac/FC-1/
if you dig around, a bunch of folks have figured out the mode select commmands
to set the format. I did a quick search but couldn't find the teac document that
describes the actual SCSI commands supported.
12 Sep
12 Sep
11:39 a.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On Tue, Sep 11, 2012 at 10:00 PM, Al Kossow <aek at bitsavers.org> wrote:
On 9/11/12 6:50 PM, Mouse wrote:
Cool! Thanks! I have one of these but haven't ever tried to hook it
up to anything.
Does ftp.bitsavers.org just not support anonymous
FTP, or is something
broken?
there is no anonymous ftp on bitsavers
http://bitsavers.org/pdf/teac/FC-1/ if you dig around, a bunch of folks have figured out
the mode select
commmands to set the format. I did a quick search but couldn't find
the teac document that describes the actual SCSI commands supported.
From what you've posted, I see that it supports
three modes - described
(in *raw* capacity terms) as 1MB (on DD media), 1.6MB, and
2.0MB (on
HD media). It seems that the drive supports FM or MFM-formatted
sectors at 256 bytes per sector, 512, and 1024, with as many sectors
as appropriate per track (so a "normal" PC DD floppy would be 9
sectors of 512 bytes per track - called "720K" commonly).
Besides the "obvious" application of reading/writing an ordinary "DOS
Floppy" (FAT12 filesystem, "720K" or "1.44MB" capacities), and
the
uncommon higher-density floppy format, what sorts of disks from
what sort of systems would be useful to read or write with this drive?
CBM 8-bit-systems disks commonly use 256 bytes per sector, but
they aren't FM or MFM, they are GCR. Are there any CP/M systems
that used 3.5" media that this could read or write disks for?
I don't tend to use floppies on a daily basis, but I'm always
interested in ways to archive old disks or create new disks from
image files to use with old equipment.
-ethan
12:35 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 12 Sep 2012 at 14:39, Ethan Dicks wrote:
Besides the "obvious" application of
reading/writing an ordinary "DOS
Floppy" (FAT12 filesystem, "720K" or "1.44MB" capacities), and
the
uncommon higher-density floppy format, what sorts of disks from what
sort of systems would be useful to read or write with this drive? CBM
8-bit-systems disks commonly use 256 bytes per sector, but they aren't
FM or MFM, they are GCR. Are there any CP/M systems that used 3.5"
media that this could read or write disks for?
Yes, but you don't run into them a lot unless you're in the CNC or
lab equipment business. For example, the Hitachi HPC-6000 or the
Mitsubishi EDM-J.
--Chuck
12:49 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On Wed, 12 Sep 2012, Ethan Dicks wrote:
Besides the "obvious" application of
reading/writing an ordinary "DOS
Floppy" (FAT12 filesystem, "720K" or "1.44MB" capacities), and
the
uncommon higher-density floppy format, what sorts of disks from
what sort of systems would be useful to read or write with this drive?
CBM 8-bit-systems disks commonly use 256 bytes per sector, but
they aren't FM or MFM, they are GCR. Are there any CP/M systems
that used 3.5" media that this could read or write disks for?
I don't have convenient access today to my files; I don't remember for
sure which of these were CP/M, but there were a few:
Amacom/Amada ; used for some CNC, even currently!
Apricot F1 ?
Apricot DS ?
AT&T WP7700 SS ?
AT&T WP7700 DS ?
Bondwell 2 ?
CDI DOT ?
CPT CP/M-80 ?
Digitech ?
Epson Geneva PX-8 (40T/Side) (67.5 TPI You would need to double step,
and might not be able to RE-write (like using a 1.2M drive for 360K))
HP 3.5 CP/M
IBS BTCE ?
Jonos
Mitsubishi A6GPP ?
Mountain Side MSC-ICO ?
NEC 8201
Netway 1500 SSDD ?
Netway 1500 DSDD ?
Olivetti ETV-250 ?
Panasonic KXE700 ?
Sanyo MBC 4050 ?
Sears (Nakajima) WP ?
Silver Reed CS1000 ?
Sony M35 CP/M SS
Sony M35 CP/M DS
Sony SMC70 3.5
Spectravideo SV738 X'PRESS ?
Tandy 600 ?
Xscripe CP/M
3:52 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 09/12/2012 03:49 PM, Fred Cisin wrote:
On Wed, 12 Sep 2012, Ethan Dicks wrote:
Apricot was a Apple clone if memory serves.
Besides the "obvious" application of
reading/writing an ordinary "DOS
Floppy" (FAT12 filesystem, "720K" or "1.44MB" capacities), and
the
uncommon higher-density floppy format, what sorts of disks from
what sort of systems would be useful to read or write with this drive?
CBM 8-bit-systems disks commonly use 256 bytes per sector, but
they aren't FM or MFM, they are GCR. Are there any CP/M systems
that used 3.5" media that this could read or write disks for?
I don't
have convenient access today to my files; I don't remember for
sure which of these were CP/M, but there were a few:
Amacom/Amada ; used for some CNC, even currently!
Apricot F1 ?
Apricot DS ? AT&T WP7700 SS ?
AT&T WP7700 DS ?
Bondwell 2 ?
CDI DOT ?
CPT CP/M-80 ?
Depends, I"ve seen both CP/M and MSDOS version differing in cpu
used.
Digitech ?
Epson Geneva PX-8 (40T/Side) (67.5 TPI You would
need to double step,
and might not be able to RE-write (like using a 1.2M drive for 360K))
The PX-8 was
definitely a CP/M machine and I have three. the 3.5" disk
unit was
a "normal" format that could be read on PC hardware.
HP 3.5 CP/M
IBS BTCE ?
Jonos
Mitsubishi A6GPP ?
Mountain Side MSC-ICO ?
NEC 8201
It ran cpm and maybe MSX.
In the list..
There were a lot of wild flyers out there. Some were sorta clones and some
tweaked private label units.
Allison
Netway 1500 SSDD ?
Netway 1500 DSDD ?
Olivetti ETV-250 ?
Panasonic KXE700 ?
Sanyo MBC 4050 ?
Sears (Nakajima) WP ?
Silver Reed CS1000 ?
Sony M35 CP/M SS
Sony M35 CP/M DS
Sony SMC70 3.5
Spectravideo SV738 X'PRESS ?
Tandy 600 ?
Xscripe CP/M
5:12 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On Wed, 12 Sep 2012, allison wrote:
Then there may have been more than one company calling itself Apricot, or
more than one format, but what I have is a 3.5" MFM diskette.
I, and others, have some Epson Geneva PX-8 diskettes that are 3.5" MFM,
with FORTY tracks per side, at 67.5 TPI! I have read them (have not tried
REwriting) with a 135tpi drive, and I HAVE an Epson 67.5 TPI drive.
It is the only MFM 3.5" disk format that I have seen with 40 cylinders.
It is nice to hear that there was also an 80 cylinder model!
AND a "stand-alone BASIC" as do many
other NEC models.
Apricot F1 ?
Apricot DS ?
Apricot was a Apple clone if memory serves. Epson Geneva
PX-8 (40T/Side) (67.5 TPI You would need to double step,
and might not be able to RE-write (like using a 1.2M drive for 360K))
The PX-8 was
definitely a CP/M machine and I have three. the 3.5" disk
unit was
a "normal" format that could be read on PC hardware. NEC 8201
It ran cpm and maybe MSX. In the list..
There were a lot of wild flyers out there. Some were sorta clones and some
tweaked private label units.
Yes, indeed.
I don't have access to my personal files today (family medical crisis), so
I can't even identify for sure which on that list were CP/M style formats,
which were MS-DOS style formats, Stand-alone-BASIC, P-system, etc.
And many had more than one format available or possible.
In many cases, I have never even SEEN the computer itself, and only have
floppy disks from it.
--
Grumpy Ol' Fred cisin at xenosoft.com
5:41 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 12 Sep 2012 at 18:52, allison wrote:
There were a lot of wild flyers out there. Some were
sorta clones and
some tweaked private label units.
I think what some are missing is that that 1024 byte sector size was
customarily used with a 360 RPM spindle speed (1.23/1.3 MB 3.5"
format). AKA NEC PC98 format, although it wasn't restricted to NEC
machines.
LS120 drives and many USB 3.5" (e.g. Teac FD05) support this
recording mode. Also called a "3-mode" drive, in case you were
wondering what that setup choice in your BIOS referred to. I have a
Toshiba "legacy floppy interface" drive that has the smarts to
recognize the firmat and alter its speed. Bizarre.
The examples I gave were 1024 byte sector CP/M 360 RPM formats.
--Chuck
6:51 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On Wed, 12 Sep 2012, Chuck Guzis wrote:
I think what some are missing is that that 1024 byte
sector size was
customarily used with a 360 RPM spindle speed (1.23/1.3 MB 3.5"
format). AKA NEC PC98 format, although it wasn't restricted to NEC
machines.
It wasn't even restricted to the NEC PC98 format.
In "720K" mode (300 RPM, 250K bits per second), you could run 5 sectors of
1024 bytes each, and get 800K per disk.
What was the sector size of the "1.7M" "IMD" distribution formats?
7:21 p.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
On 12 Sep 2012 at 18:51, Fred Cisin wrote:
On Wed, 12 Sep 2012, Chuck Guzis wrote:
Or 10 sectors of 512, but you miss the point, I think. You can't
read or write PC98 disks in a garden-variety PC or Mac floppy drive,
as those aren't usually 3-mode. Before I could get reasonably-priced
3-mode drives, I'd just use a secondary FDC and change the 24 MHz
crystal to 20 MHz. Mohammed to the mountain and all that...
Are you thinking of the Microsoft distribution mode (DMF)? 21
sectors of 512 bytes. Far more interesting was IBM XDF. If memory
serves, 1 sector each of 8192, 2048, 1024 and 512 bytes per track.
That's 11,176 formatted bytes per track, so only 1,326 bytes per
track were used for headers, CRCs, gaps and sync bytes. IBM had a
lot of guts to try something like that as an "official" format.
--Chuck
I think what some are missing is that that 1024
byte sector size was
customarily used with a 360 RPM spindle speed (1.23/1.3 MB 3.5"
format). AKA NEC PC98 format, although it wasn't restricted to NEC
machines.
It wasn't even restricted to the NEC PC98 format.
In "720K" mode (300 RPM, 250K bits per second), you could run 5
sectors of 1024 bytes each, and get 800K per disk.
What was the sector size of the "1.7M" "IMD" distribution formats?
13 Sep
13 Sep
11:12 a.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
Apricot F1 ?
Apricot DS ?
Apricot was a Apple clone if memory serves. Epson Geneva
PX-8 (40T/Side) (67.5 TPI You would need to double step,
and might not be able to RE-write (like using a 1.2M drive for 360K))
The PX-8 was
definitely a CP/M machine and I have three. the 3.5" disk
unit was
a "normal" format that could be read on PC hardware.
10:59 a.m.
New subject: TEAC FC-1 (was Re: Regarding SATA vs. SAS/FibreChannel)
I don't have convenient access today to my files;
I don't remember for
sure which of these were CP/M, but there were a few:
[...]
Amstrad PCW9512+ (YEs, thsi had 3.5", not 3" drives and ran CP/M)
Apricot F1 ?
Apricot DS ?
Those are 'PC incompatibles'. I am pretty sure htt at least the original
Apricot had an 8086 (and not an 8088) processor, along with an 8089 'I/O
procesosor' (a fancy DMA chip) and an optional 8087. THey did not run
CP/M 80. Whether they ran CP/M 86 I do not recall, they normally ran a
special MS-DOS.
[...]
Epson Geneva PX-8 (40T/Side) (67.5 TPI You would
need to double step,
and might not be able to RE-write (like using a 1.2M drive for 360K))
Also on the Epson PX4, of course.
HP 3.5 CP/M
Likely to be a signle-sided disk, 135tpi, but with only 70 tracks used.
The drive unit is the nroaml HP9121. The machine would be an HP120 or
HP125 I would guess.
[...]
Tandy 600 ?
I thought that was an 8088 box?
-tony
11 Sep
11 Sep
9:31 a.m.
On 11 Sep 2012 at 8:17, Fred Cisin wrote:
On Tue, 11 Sep 2012, Mouse wrote:
If it's a drive used in a lot of workstations (e.g. Teac FD235S),
it's a standard FD235HF with a bolted on SCSI-to-Floppy
adapter/controller. Sector size is strictly a matter of firmware,
but I believe the Teac drives allowed for different sector sizes.
If it's one of the "flopticals", such as the old InSite or LS-120
drives, format choice is much less flexible, although most LS-120
drives (as well as most USB drives) will *read* (but not format) PC98
1024-byte sector floppies.
The big problem that makes SCSI unsitable for universal floppy use is
that it's strictly block-relative addressing--the entire disk is
assumed to contain tracks of the same number and size sectors. This
can create issues when a disk has, say, an FM first cylinder with the
remainder of the disk MFM (not as uncommon as one might think).
--Chuck
I've got a 3?" SCSI floppy drive (yes, a
floppy drive with a SCSI
interface!) that reports itself as having 256-byte blocks.
Does it have any provision to reconfigure?
4435
days inactive
4437
days old
35 comments
16 participants
participants (16)
-
aek＠bitsavers.org -
ajp166＠verizon.net -
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
cisin＠xenosoft.com -
eric＠brouhaha.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
jfoust＠threedee.com -
jws＠jwsss.com -
mcguire＠neurotica.com -
mouse＠Rodents-Montreal.ORG -
RichA＠vulcan.com -
snhirsch＠gmail.com -
tothwolf＠concentric.net -
wdonzelli＠gmail.com