24 Dec
2010
24 Dec
'10
5:33 p.m.
________________________________
On 12/24/10 5:40 PM, Gene Buckle wrote:
Unfortunately, nobody is kitting them.
They're dead easy to assemble. The only "problem" components are the
unreasonably large EPROM and SRAM chips, and even they aren't too tough
to find.
-Dave
--
Dave McGuire
Port Charlotte, FL
-----REPLY-----
Hi! Due to numerous builder requests, we fixed the RAM/ROM size issue with
SBC V2. SBC V2 accepts 28 and 32 pin ROMs (EPROM, PROM, EEPROM, or Flash)
which is pretty much anything from a 27256 to a 29C040 or 27C080. You could
probably go smaller if you want although I haven't tested it myself. Also
the RAM accepts 512KB or 128KB SRAMs. All set with config jumpers. There
are several other improvements which fixed several bugs from SBC V1.
Maybe what is needed for the C64 CP/M cart is a expansion port to ECB
interface board. Then plug in an N8VEM SBC, SBC-188, or whatever and use
the ECB peripheral boards along with your C64. We have a 6809/6802/6502 to
ECB bridge board that might be a basis. Make the C64 appear as an IO port
on the ECB and as a memory mapped IO on the C64. Plug in SBC V2 for your
CP/M-80 cartridge or SBC-188 V1 for the CP/M-86. Just an idea.
Back in the day, I used to have a C64 with a CP/M cartridge. It was great
-- Nevada FORTRAN rocks!
Thanks and have a nice day!
Andrew Lynch
I wouldn't mind a CP/M standalone if it was cheap.
Most of the CP/M
machines of the era are either recycled or need some repairs by now.
N8VEM is a really nice machine..
24 Dec
24 Dec
5:47 p.m.
I'll try to ask a different question here
What is needed to CREATE (and not build) a CP/M compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc discussing
it?
6:19 p.m.
On Fri, 24 Dec 2010, Alexandre Souza - Listas wrote:
What is needed to CREATE (and not build) a CP/M
compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc discussing
it?
You don't NEED 64K. CP/M can run in 16K. But, if you want a decent sized
TPA, you can have 50+ K TPA with 64K, or even 63K TPA with 128K.
8080. Yeah, a Z80 is NICE. and FAST! and some aps insist on it. But,
CP/M can run on 8080, 8085, and anything else, including emulators that
can handle 8080 code.
It's too bad that nobody brought up CP/M on the Kyoceras (Radio Shack
model 100, Nec 8200, etc). THAT would have been a project.
Or on the Epson RC20!
25 Dec
25 Dec
4:16 a.m.
It's too bad that nobody brought up CP/M on the
Kyoceras (Radio Shack
model 100, Nec 8200, etc). THAT would have been a project.
I think a lot of people (including myself, amazingly) tended to use the
built-in software of the M100, so no reall need for CP/M.
It probably wouldn't be that big a modification (the main part being
re-mappling the memory so there's RAM at location 0 -- in fact a
programmable invesion of A15 would do it). But you'd only have 32K RAM
and an 8085 processor which would limit what you could run.
Or on the Epson RC20!
The Epson TF20 (floppy drive for the HX20, PX4, PX8, etc) contains a Z80
and 64K RAM. Oh, and a boot ROM that is mapped out after booting. I have
never analyhsed the system disk in detail, but I have a feeling this
thing runs CP/M internally.
-tony
6:16 a.m.
On 12/25/2010 07:16 AM, Tony Duell wrote:
Exactly. Mine is modded to 64K ram, 32K "boot disk"
and a 16mb CF for main store.
80% of the software out there is 8080. Those few that are not
are often available in source form and can be patched if needed.
The PX-8 is a CP/M system based on Rom and Ram disk with
hooks for external extension for both greater rom/ram disk or
the floppy device.
Allison
It's too
bad that nobody brought up CP/M on the Kyoceras (Radio Shack
model 100, Nec 8200, etc). THAT would have been a project.
I think a lot of
people (including myself, amazingly) tended to use the
built-in software of the M100, so no reall need for CP/M.
It probably wouldn't be that big a modification (the main part being
re-mappling the memory so there's RAM at location 0 -- in fact a
programmable invesion of A15 would do it). But you'd only have 32K RAM
and an 8085 processor which would limit what you could run.
Or on the
Epson RC20!
The Epson TF20 (floppy drive for the HX20, PX4, PX8, etc) contains a
Z80
and 64K RAM. Oh, and a boot ROM that is mapped out after booting. I have
never analyhsed the system disk in detail, but I have a feeling this
thing runs CP/M internally.
-tony
10:11 a.m.
Or on the Epson RC20!
The Epson TF20 (floppy
drive for the HX20, PX4, PX8, etc) contains a Z80
and 64K RAM. Oh, and a boot ROM that is mapped out after booting. I have
never analyhsed the system disk in detail, but I have a feeling this
thing runs CP/M internally.
11:35 a.m.
On 12/25/2010 01:11 PM, Tony Duell wrote:
The PX-8 is a CP/M system based on Rom and Ram disk with
hooks for external extension for both greater rom/ram disk or
the floppy device. Sure, as is the PX4 (I have both). The HX20, of course is not
(it uses a
pair of 6303 microprocessors). But the 5.25" floppy drive appears to run
CP/M (or something close) on its intenral processor. The 3.5 drive, BTW,
doesn't...
Ah HUH?
Speaking for the PX-8 (I have three plus wedges and drives) the base PX8
runs CP/M with a menu program on boot. If you escape from the menu it's
stock CP/M. ti's clever for it's time in that the BIOS is complex and
complete
enough to handle a raft of storage devices be they based on R0M, RAM or
a real floppy. As it turns out they made the floppy smart so that some of
bios is in the floppy and the system can do directed calls to it so the main
unit is relieved of some of the work.
So the Floppy for the PX8 is specialized for a CP/M host it's not running
CP/M itself as there is not enough ram alone to qualify.
Allison
Or on the
Epson RC20!
The Epson TF20 (floppy drive for the HX20, PX4, PX8, etc) contains a
Z80
and 64K RAM. Oh, and a boot ROM that is mapped out after booting. I have
never analyhsed the system disk in detail, but I have a feeling this
thing runs CP/M internally.
-tony
12:05 p.m.
pair of 6303
microprocessors). But the 5.25" floppy drive appears to run
CP/M (or something close) on its intenral processor. The 3.5 drive, BTW,
doesn't...
Ah HUH?
Speaking for the PX-8 (I have three plus wedges and drives) the base PX8
runs CP/M with a menu program on boot. If you escape from the menu it's stock CP/M. ti's clever for it's time in that
the BIOS is complex and
complete
enough to handle a raft of storage devices be they based on R0M, RAM or
a real floppy. As it turns out they made the floppy smart so that some of
Not forgettign the microcassette tape,. which is trated somewhat like a
slow floppy drive by the system.
bios is in the floppy and the system can do directed
calls to it so the main
unit is relieved of some of the work.
So the Floppy for the PX8 is specialized for a CP/M host it's not running
CP/M itself as there is not enough ram alone to qualify.
Have you ever looked inside the TF20 daul 5.25" floppy unit? I beleive
schematics are on the web somewhere...
There is a heck of a lot of elkectronics inside. One large PCB containg a
Z80, 2K (IIRC) boot ROM ()with logic to switch it out after booting), 64K
DRAM, floppy controller, a duaghterboard with a 7201 serial chip, etc.
From what others have been saying that's easily
enough to run CP/M. Note,
I am not talking about the CP/.M that runs in the PX8 or
wherever. I am
talking about the Z80 in the TF20 itself.
And the TF20 is not specialised for a CP/M host. The original application
was for the HX20, which as I said has a poair of 6303 CPUs and does not
run CP/M.
-tony
1:28 p.m.
On 12/25/2010 03:05 PM, Tony Duell wrote:
Indeed...
I have the portable battery powered single 3.5" floppy (PF10) and it's
not
so full of ram. I also have the full manuals for the system, peripherals,
both tech and programming. On occasion I hack peripherals for them.
Allison
pair of 6303 microprocessors). But the 5.25"
floppy drive appears to run
CP/M (or something close) on its intenral processor. The 3.5 drive, BTW,
doesn't...
Ah HUH?
Speaking for the PX-8 (I have three plus wedges and drives) the base PX8
runs CP/M with a menu program on boot. If you escape from the menu it's stock CP/M. ti's clever for it's time in
that the BIOS is complex and
complete
enough to handle a raft of storage devices be they based on R0M, RAM or
a real floppy. As it turns out they made the floppy smart so that some of
Not
forgettign the microcassette tape,. which is trated somewhat like a
slow floppy drive by the system.
bios is in the floppy and the system can do
directed calls to it so the main
unit is relieved of some of the work.
So the Floppy for the PX8 is specialized for a CP/M host it's not running
CP/M itself as there is not enough ram alone to qualify.
Have you ever looked
inside the TF20 daul 5.25" floppy unit? I beleive
schematics are on the web somewhere...
There is a heck of a lot of elkectronics inside. One large PCB containg a
Z80, 2K (IIRC) boot ROM ()with logic to switch it out after booting), 64K
DRAM, floppy controller, a duaghterboard with a 7201 serial chip, etc.
From what others have been saying that's
easily enough to run CP/M. Note,
I am not talking about the CP/.M that runs in the
PX8 or wherever. I am
talking about the Z80 in the TF20 itself.
And the TF20 is not specialised for a CP/M host. The
original application
was for the HX20, which as I said has a poair of 6303 CPUs and does not
run CP/M.
-tony
26 Dec
26 Dec
10:32 a.m.
There is a
heck of a lot of elkectronics inside. One large PCB containg a
Z80, 2K (IIRC) boot ROM ()with logic to switch it out after booting), 64K
DRAM, floppy controller, a duaghterboard with a 7201 serial chip, etc.
I have the portable battery powered single 3.5" floppy (PF10) and it's
not
so full of ram. I also have the full manuals for the system, peripherals,
both tech and programming. On occasion I hack peripherals for them.
From what others have been saying that's
easily enough to run CP/M. Note,
I am not talking about the CP/.M that runs in the
PX8 or wherever. I am
talking about the Z80 in the TF20 itself.
4:12 p.m.
On 12/26/2010 01:32 PM, Tony Duell wrote:
In an earlier message I did mention that my comments did not apply to
r the 3.5" drives, which are very different inside. The PF10 (I have a
non-working one [1]) uses a 6303 microcontroller IIRC, with a program
ROM, small-ish RAM, and floppy controller chip (765 IIRC). One odd-ish
feature is that the floppy dirve mechanics (stepper, etc) are controlled
by the micrtocontroller (which takes the step signal from the 765 as an
interrupt IIRC), much of the logic you'd experct to find in a floppy
drive is absent.
It's also odd in that it's a 40cylinder (67.5 tpi) device.
its first generation 3.5", also runs slower data rate.
There is a heck of a lot of elkectronics inside. One
large PCB containg a
Z80, 2K (IIRC) boot ROM ()with logic to switch it out after booting), 64K
DRAM, floppy controller, a duaghterboard with a 7201 serial chip, etc.
I have the portable battery powered single 3.5" floppy (PF10) and it's
not
so full of ram. I also have the full manuals for the system, peripherals,
both tech and programming. On occasion I hack peripherals for them.
From what others have been saying that's
easily enough to run CP/M. Note,
I am not talking about the CP/.M that runs in the
PX8 or wherever. I am
talking about the Z80 in the TF20 itself.
[1] The microcontorller has failed. Even though
it's ROMless, this is not
much of a help. Finding the right version i nthe right package after all
thses years has proved impossible. I am darn sure it's the
microcontroller in that I pulled the ROM and forced NOPs onto the data
bus, and it still had an address bus that looked crazy.
6303 is easy to find, I have a few from old defunct HDs.
On inmportatn difference to the user between the
5.25" andf 3.5" drives
is that the former need a 'system disk' to do anything useful, the latter
do not, they have the operating software in EPROM.
Looks that way.
Allison
-tony
8:35 p.m.
New subject: Epson drives (Was: C64-C128 CP-M Cartridge Interest
On 26 Dec 2010 at 16:31, Fred Cisin wrote:
Did anybody besides Epson ever use the 40 cylinder 3.5" in a cmmercial
product?
Sure--a number of low-end word processors, such as Silver-Reed and
Brother used them. I think the Preis CP/M box used them also.
I find it more interesting that some of the 135 tpi users selected
odd values for the upper limit. HP used 66, I think and Jonos used
70. Of course, the Japanese HD application were all mirrors of 8"
usage and tended to be 77 cylinders for that reason.
--Chuck
It's also odd in that it's a 40cylinder (67.5
tpi) device.
its first generation 3.5", also runs slower data rate. 
27 Dec
27 Dec
1:17 a.m.
New subject: Epson drives
On Sun, 26 Dec 2010, Chuck Guzis wrote:
I find it more interesting that some of the 135 tpi
users selected
odd values for the upper limit. HP used 66, I think and Jonos used
HP used 35 for their 5.25" DS drives (e.g. HP82901) and 70 for their 3.5"
SS drives (e.g. HP9121).
Christian
9:27 a.m.
New subject: Epson drives
On 27 Dec 2010 at 10:17, Christian Corti wrote:
On Sun, 26 Dec 2010, Chuck Guzis wrote:
Perhaps, but the HP125 and HP150 Series I *used* 66. I don't know
why, just that's what I found. Also, the HP125 8" drives used 66.
--Chuck
I find it more interesting that some of the 135
tpi users selected
odd values for the upper limit. HP used 66, I think and Jonos used
HP used 35 for their 5.25" DS drives (e.g. HP82901) and 70 for their
3.5" SS drives (e.g. HP9121).
2:45 p.m.
New subject: Epson drives
> HP used 35 for their 5.25" DS drives (e.g.
HP82901) and 70 for their
> 3.5" SS drives (e.g. HP9121).
On Mon, 27 Dec 2010, Chuck Guzis wrote:
Perhaps, but the HP125 and HP150 Series I *used* 66.
I don't know
why, just that's what I found. Also, the HP125 8" drives used 66.
The inner (higher numbered) tracks were certainly less reliable in the
earlier drives.
Perhaps HP set some sort of reliability threshold, and stopped at the
point that their statisticians said was the max to perform [statistically]
to spec.
I saw multiple DIFFERENT upper limits (66 and 70) on discs (HP refused to
call them "disks") that were supposedly the same format.
--
Grumpy Ol' Fred cisin at xenosoft.com
28 Dec
28 Dec
10:47 a.m.
New subject: Epson drives
On 27 Dec 2010 at 14:45, Fred Cisin wrote:
Not really, this ia a drive unit and not system parameter.
An HP150-II will use the 9121 signle-sided drive, where it formats it to
70 cylinders (and probably gives you 66 of them). I think an HP150 can
use the later double-sided drives like the 9122 (and will format them to
77 cylinders), but it may need a BIOS ROM upgrade amd/or a later version
of MS-DOS.
-tony
I saw multiple DIFFERENT upper limits (66 and 70)
on discs (HP refused
to call them "disks") that were supposedly the same format.
Indeed, the HP150 Series II floppies used 77 cylinders for data. 
6:26 p.m.
New subject: Epson drives
HP did everything it could to shoot itself in the foot. After promising
compatibility among the 100-Series machines, the floppy format produced
by the 9114 driven by the HP110 Portable or the Portable Plus was
different from the format produced by the HP150. There were clever
format programs and device driver tricks to get around some of these
incompatibilities as well as to allow the PPlus to make and use 3.5"
floppies that could also be read and written by regular DOS machines.
Sometimes I find it amazing that HP has survived!
Jim
On 12/28/2010 1:47 PM, Tony Duell wrote:
On 27 Dec 2010
at 14:45, Fred Cisin wrote:
Not really, this ia a drive
unit and not system parameter.
An HP150-II will use the 9121 signle-sided drive, where it formats it to
70 cylinders (and probably gives you 66 of them). I think an HP150 can
use the later double-sided drives like the 9122 (and will format them to
77 cylinders), but it may need a BIOS ROM upgrade amd/or a later version
of MS-DOS.
-tony I saw multiple DIFFERENT upper limits (66 and 70)
on discs (HP refused
to call them "disks") that were supposedly the same format.
Indeed, the
HP150 Series II floppies used 77 cylinders for data.
7:52 p.m.
New subject: Epson drives
On 28 Dec 2010 at 21:26, Jim Scheef wrote:
HP did everything it could to shoot itself in the
foot. After
promising compatibility among the 100-Series machines, the floppy
format produced by the 9114 driven by the HP110 Portable or the
Portable Plus was different from the format produced by the HP150.
There were clever format programs and device driver tricks to get
around some of these incompatibilities as well as to allow the PPlus
to make and use 3.5" floppies that could also be read and written by
regular DOS machines. Sometimes I find it amazing that HP has
survived!
Let's see,
HP-150 Series I: 66 cylinders, 1 side, 16 sectors of 256 bytes
interleaved 4:1),
1,024 byte clusters, media byte FA,
128 root directory entries.
3 sectors per FAT
HP-150 Series II: 77 cylinders, 2 sides, 9 sectors of 512 bytes
interleaved 2:1),
1,024 byte clusters, media byte FA,
128 root directory entries.
3 sectors per FAT
HP-110: 77 cylinders,2 sides, 9 sectors of 512 bytes
interleaved 1:1), but only 8 used,
1,024 byte clusters, media byte FB,
176 root directory entries.
3 sectors per FAT
HP Portable Plus: 77 cylinders, 1 side, 5 sectors of 1024 bytes
interleaved 1:1),
1,024 byte clusters, media byte FC,
96 root directory entries.
2 sectors per FAT
Yup, all mutually incompatible.
--Chuck
29 Dec
29 Dec
12:03 p.m.
New subject: Epson drives
Let's see,
I don;t doubt what you say, as I know you have a lot more experience of
this than I do. But there are a few comments...
I have the Techncial Reference Manuals for the Portable+, HP150 and
HP150-II. I do not, alas, have the manual for the HP110.
Unfortuantely they don't say much about the physical disk formats. The
Portable+ manual skips over that entirely. The HP150 and 150-II manuals
have a very brief summary, roughlu what you've given below.
HP-150 Series I: 66 cylinders, 1 side, 16 sectors of 256 bytes
interleaved 4:1),
1,024 byte clusters, media byte FA,
128 root directory entries.
3 sectors per FAT
HP-150 Series II: 77 cylinders, 2 sides, 9 sectors of 512 bytes
interleaved 2:1),
1,024 byte clusters, media byte FA,
128 root directory entries.
3 sectors per FAT
This is not strictly accurate.
A single-sided drive can be connected to the HP150-II, at which point is
uses the 66 cylinder, etc, format, AFIK totally compatible with the
earlier HP150.
HP150s with Reve D firmware (I think only Rev D and Rev F were released)
can only use single sided drives using the Amigo protocol All HP150-IIs
have Rev F frimware, and from what I can determine, you could get Rev F
frimware as n upgrade for the HP150. Machines with Rev F frimware can use
SS/80 double-sided drives.
HP-110: 77 cylinders,2 sides, 9 sectors of 512 bytes
interleaved 1:1), but only 8 used,
1,024 byte clusters, media byte FB,
176 root directory entries.
3 sectors per FAT
HP Portable Plus: 77 cylinders, 1 side, 5 sectors of 1024 bytes
interleaved 1:1),
1,024 byte clusters, media byte FC,
96 root directory entries.
2 sectors per FAT
Yup, all mutually incompatible.
I have certianly read HP150-II double-sided disks on a portable+ with no
problems. I have also connected HP150-II SS/80 hard disk units to a
portable+ with no problems.
And as I mentioned in an earlier reply, the Portable+ manual has some
information about using Amigo drifves and formatting single-sided disks
on the Portable+ Which are then claimed to be compatible with the HP150,
I am suprised HP made such a mess of this. The LIF disk format was pretty
much universal. To the extent that you can use the same disk with an HP41
calculator, an H71 handhled computer and an HP9000/200 series machine.
Oh, and I believe quite a bit of test gear too.
-tony
10:55 a.m.
New subject: Epson drives
HP did everything it could to shoot itself in the foot. After promising
compatibility among the 100-Series machines, the floppy format produced
by the 9114 driven by the HP110 Portable or the Portable Plus was
different from the format produced by the HP150. There were clever
I will say more in a reply to a later message.
I don't doubt you are correct (I have tried to avoid MS-DOS machines
whether HP or not :-)), but I have never had any problems reading HP150
disks on a Portable+ or vice versa.
The Portable+ Technical Reference Manual (yes, there is one, and yse, it's
rarer than hen's teeth) contains no information on the physical disk
format at all. It does claim that an HP150B or C (that's one that supports
double sided drives) can read/write Portable+ disks. It also gives a
procedure for formatting a single-sided disk on the Portable+ which can
then beread/written on the HP150A (single-sided drives only).
The manual also says (and I have proved this to be correct) that you can
use HP150 SS/80 drives (hard ar floppy) on the Portable+ using an 82169
HPIL-HPIB translater, provided the latter has a late enough firmware
version (there's a serial number given in the manaula, after which the
82169 will work). You can also used Amigo drives (like the single-sided
floppies) if you load a device driver in config.sys on the Poerable+
(the driver is in the ROM).
-tony
12:11 p.m.
New subject: Epson drives
Sometimes I find it amazing that HP has survived!
They haven't, or at least the HP I knew and loved is no more. I do not
consider going from a manufacturer of some very nice measuring
instruments to a manufactuer of printers and scanners (and not always
particularly good ones) to be 'survival'
-tony
27 Dec
27 Dec
2:33 p.m.
New subject: HP drives
On Sun, 26 Dec 2010, Chuck Guzis wrote:
> I find it more interesting that some of the 135 tpi users selected
> odd values for the upper limit. HP used 66, I think and Jonos used
On Mon, 27 Dec 2010, Christian Corti wrote:
HP used 35 for their 5.25" DS drives (e.g.
HP82901) and 70 for their 3.5"
SS drives (e.g. HP9121).
35 was, of course the original Shugart SA400. I guess that HP never heard
about, nor believed in, the extension to 40 of the "standard".
Then, of course, 79 was simply doubling the SA400 "standard".
"Standards aew wonderful things; everybody can have a unique one of their
own." - George Morrow
--
Grumpy Ol' Fred cisin at xenosoft.com
28 Dec
28 Dec
10:45 a.m.
New subject: HP drives
On Sun, 26 Dec 2010, Chuck Guzis wrote:
> I find it more interesting that some of the 135 tpi users selected
> odd values for the upper limit. HP used 66, I think and Jonos used
On Mon, 27 Dec 2010, Christian Corti wrote:
HP used 35 for their 5.25" DS drives (e.g.
HP82901) and 70 for their 3.5"
SS drives (e.g. HP9121).
35 was, of course the original Shugart SA400. I guess that HP never heard
about, nor believed in, the extension to 40 of the "standard".
Then, of course, 79 was simply doubling the SA400 "standard".
I assume that;s a typo. Every time I shift 100011 left by one bit I get
1000110, or 70...
Howeer, was there ever a 70 cylinder 5.25" drive? That is one that was
only designed to use 70 cylinders ? I mean a raw drive, not a
drive+controller like the HP units we're discussing.
-tony
11:48 a.m.
New subject: HP drives
The original Shugart SA400 minifloppy (August 1976) only allowed 35
tracks. In early 1977 Wangco announced the Model 82 Micro Floppy that
allowed 40 tracks with new media.
Here is a photo of both disks.
http://commons.wikimedia.org/wiki/File:Verbatim_5.25_minidisk_tracks_197
8.jpg
Michael Holley
-----Original Message-----
From: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] On Behalf Of Tony Duell
Sent: Tuesday, December 28, 2010 10:45 AM
To: cctalk at classiccmp.org
Subject: Re: HP drives
heard
> On Sun, 26 Dec 2010, Chuck Guzis wrote:
> > I find it more interesting that some of the 135 tpi users selected
> > odd values for the upper limit. HP used 66, I think and Jonos
used
On Mon, 27 Dec 2010, Christian Corti wrote:
> HP used 35 for their 5.25" DS drives (e.g. HP82901) and 70 for their
3.5"
SS drives
(e.g. HP9121).
35 was, of course the original Shugart SA400. I guess that HP never about, nor believed in, the extension to 40 of the
"standard".
Years ago somebody tried to tell me that '35 track' 35.25" disks had a
smaller head slot than 40 track ones and they could never be formatted
to
40 tracks. Whs this ever true?
[Mind you, this was the same person who tried to convince me that
formatting the 'other side' of a single-sided disk could ruin the drive
head, although using it asa single-sided disk would cause no damage. I
cannot think how that could be true.]
Then, of course, 79 was simply doubling the SA400 "standard".
I assume that;s a typo. Every time I shift 100011 left by one bit I get
1000110, or 70...
Howeer, was there ever a 70 cylinder 5.25" drive? That is one that was
only designed to use 70 cylinders ? I mean a raw drive, not a
drive+controller like the HP units we're discussing.
-tony
12:17 p.m.
New subject: HP drives
On 28 Dec 2010 at 11:48, Holley, Michael wrote:
The original Shugart SA400 minifloppy (August 1976)
only allowed 35
tracks. In early 1977 Wangco announced the Model 82 Micro Floppy that
allowed 40 tracks with new media.
Here is a photo of both disks.
http://commons.wikimedia.org/wiki/File:Verbatim_5.25_minidisk_tracks_1
97 8.jpg
I was about to mention that I seem to recall that the SA400 could
indeed write and read 40 cylinders, to the best of my recollection,
but that some early media didn't permit it.
And, to add a bit of stuff to the mix, the Micropolis 100 tpi 5.25"
floppies used 77 cylinders, with cylinder 0 offset outward a bit from
where a 96 TPI drive would put it.
And I don't recall how many cylinders the Shugart/Dysan 3.25"
floppies supported. 40? I could format one up and see...
--Chuck
1:45 p.m.
New subject: HP drives
Tony Duell wrote:
Howeer, was there ever a 70 cylinder 5.25" drive?
That is one that was
only designed to use 70 cylinders ? I mean a raw drive, not a
drive+controller like the HP units we're discussing.
I haven't heard of
one, though Micropolis made 77-track 5.25" drives
with 100tpi (vs. 35 or 40 track at 48tpi and 80 track at 96tpi), and a
few other vendors offered compatible drives, e.g. the Tandon TM100-4M,
where the "M" designates it as a 100tpi model.
4:52 p.m.
New subject: HP drives
On 12/28/10 1:45 PM, "Eric Smith" <eric at brouhaha.com> wrote:
Tony Duell wrote:
ISTR: the 8" drives at least as used in tandy computers formatted out to 77
tracks....
Howeer, was there ever a 70 cylinder 5.25"
drive? That is one that was
only designed to use 70 cylinders ? I mean a raw drive, not a
drive+controller like the HP units we're discussing.
I haven't heard of
one, though Micropolis made 77-track 5.25" drives
with 100tpi (vs. 35 or 40 track at 48tpi and 80 track at 96tpi), and a
few other vendors offered compatible drives, e.g. the Tandon TM100-4M,
where the "M" designates it as a 100tpi model.
5:06 p.m.
New subject: HP drives
On 28 Dec 2010 at 16:52, Geoffrey Reed wrote:
ISTR: the 8" drives at least as used in tandy
computers formatted out
to 77 tracks....
That's standard for all 8" drives--and it's "cylinders", please!
8"
drives can be double-sided with 2 tracks per cylinder.
The "M" in the TM-100-4M stood for "Micropolis"--not only for the 100
tpi physical format, but also for the somewhat different pinout from
the 96 tpi TM-100-4.
--Chuck
29 Dec
29 Dec
10:48 a.m.
New subject: HP drives
On 28 Dec 2010 at 16:52, Geoffrey Reed wrote:
I thought that at least the earlier Tandy 8" machines had single-head
drives, in which case'track' == 'cylinder'.
I can't rememebr which machine it was, but one machine with 5.25" drives
claimed to have '80 track' drives. What they actually were were 40
cylinder, 2 head drives. So technically they did have 80 tracks, but it
caused a lot of confusion at the time.
-tony
ISTR: the 8" drives at least as used in
tandy computers formatted out
to 77 tracks....
That's standard for all 8" drives--and it's "cylinders", please!
8"
drives can be double-sided with 2 tracks per cylinder.
2:18 p.m.
New subject: HP drives
On 29 Dec 2010 at 18:48, Tony Duell wrote:
My point was that the OP stated "8" drives at least as used in tandy
computers..." and that I have a Model 16 for example, that uses 8" DS
drives.
On 28 Dec 2010 at 16:52, Geoffrey Reed wrote:
I thought that at least the earlier Tandy 8" machines had single-head
drives, in which case'track' == 'cylinder'. ISTR: the 8" drives at least as used in
tandy computers formatted
out to 77 tracks....
That's standard for all 8" drives--and it's "cylinders", please!
8"
drives can be double-sided with 2 tracks per cylinder. I can't rememebr which machine it was, but one
machine with 5.25"
drives claimed to have '80 track' drives. What they actually were were
40 cylinder, 2 head drives. So technically they did have 80 tracks,
but it caused a lot of confusion at the time.
I've seen many times where the characteristics of a 3.5" floppy are
given as "160 tracks". My take is that if you mean a physical head
position, it's best to say "cylinders".
Just as you would with a hard drive. For the life of me, I don't see
why some folks insist on treating the nomenclature for hard and
floppy drives differently.
--Chuck
30 Dec
30 Dec
10:35 a.m.
New subject: HP drives
I've seen many times where the characteristics of
a 3.5" floppy are
given as "160 tracks". My take is that if you mean a physical head
position, it's best to say "cylinders".
I absolutely agree, and I try to do this. Doubtless I sometimes get it
wrong, but ....
However, like the term 'DB9 connector' (which you and I both know nearly
always mean 'DE9 connector'), alas the misuse of 'track' has become
common usage. An awful lot of people would claim the 3.5" drive in a PC
was an '80 track' device. It is, of course, 160 tracks with 80 cylinders
each containing 2 tracks.
I guess what I am saying is while it helps to be correct, it also helps
to know common incorrect usage of such terms.
Just as you would with a hard drive. For the life of
me, I don't see
why some folks insist on treating the nomenclature for hard and
floppy drives differently.
Alas they don't :-(. I've heard the term 'track' used for hard drives
when they meant 'cylinders'. ARGH!!!
-tony
6 Jan
6 Jan
6:10 a.m.
New subject: HP drives
----- Original Message -----
From: "Tony Duell" <ard at p850ug1.demon.co.uk>
To: <cctalk at classiccmp.org>
Sent: Thursday, December 30, 2010 6:35 PM
Subject: Re: HP drives
Alas they don't :-(. I've heard the term
'track' used for hard drives
when they meant 'cylinders'. ARGH!!!
-tony
I thought you could only read/write to the top layer of the harddrive
platter, or do some drives allow writing to both sides?
Regards,
Andrew B
aliensrcooluk at yahoo.co.uk
7 Jan
7 Jan
3:39 p.m.
New subject: HP drives
On 6 Jan 2011 at 14:10, Andrew Burton wrote:
I thought you could only read/write to the top layer
of the harddrive
platter, or do some drives allow writing to both sides?
Writing to both sides of a platter has been the way since the IBM 350
RAMAC drive.
--Chuck
8 Jan
8 Jan
12:13 p.m.
New subject: HP drives
Alas they
don't :-(. I've heard the term 'track' used for hard drives
when they meant 'cylinders'. ARGH!!!
-tony
I thought you could only read/write to the top layer of the harddrive
platter, or do some drives allow writing to both sides?
12 Jan
12 Jan
12:07 p.m.
New subject: HP drives
----- Original Message -----
From: "Tony Duell" <ard at p850ug1.demon.co.uk>
To: <cctalk at classiccmp.org>
Sent: Saturday, January 08, 2011 8:13 PM
Subject: Re: HP drives
Ahh, I see. Thanks for that Tony.
You (or more accurately, I) learn something new every day - especially on
this list :)
Regards,
Andrew B
aliensrcooluk at yahoo.co.uk
I thought you could only read/write to the top layer of the harddrive
platter, or do some drives allow writing to both sides?
Except for a ferw very odd ones (e.g. hard drives which record analagoue
singals such as analogue video [1]) every hard disk I've worked on has
used both sides of all platters for something. Maybe not user data
storage (for excample, it may contain servo information only), but there
will be a head on it.
[1[] Yes such things did exist. They were used for action replays on TV,
for exmaple. And somewhere I have a vido output system for the PDP11
which uses a special hard disk as its video memeory).
In the context of a hard disk, though, a 'cylinder' means all the tracks
that can be accessed without moving the heads. So if you have 4 data
surfaces (such as the ST412), a cylinder contains 4 tracks. In an RK07,
which has 3 data surfaces and 1 servo surface, as cylinder it 3 tracks.
And so on.
-tony
1:24 p.m.
New subject: HP drives
> In the context of a hard disk, though, a
'cylinder' means all the tracks
> that can be accessed without moving the heads. So if you have 4 data
> surfaces (such as the ST412), a cylinder contains 4 tracks. In an RK07,
> which has 3 data surfaces and 1 servo surface, as cylinder it 3 tracks.
> And so on.
Ahh, I see. Thanks for that Tony.
The name 'cylinder' comes, of course, from the image of a cylinder
paralel to the drive spindle going trhough all the platters. The
intersecion of that 'cylinder' and the platters (both sides) are the
tracks that you can access without moving the heads.
Now, if you had a positioner that could only move half the distance from
outside rim of the platter to the spindle, but you had 2 heads per
surface suitbleyoffset (so that you could access 2 tracks on each surface
without moving the psoitoned) then I would claim that both of thos tracks
were part of the same sylinder. For example, if you had 4 data surfaces
(and thus 8 hards in this drive), I would say that a 'cylinder' contains
8 tracks, even though the arrangement of said tracks doesn't look like a
physical cylinder any more.
You (or more accurately, I) learn something new every
day - especially on
this list :)
Actually, I learn several things from this list every day.
-tony
3:25 p.m.
New subject: HP drives
On 12 Jan 2011 at 21:24, Tony Duell wrote:
Now, if you had a positioner that could only move half
the distance
from outside rim of the platter to the spindle, but you had 2 heads
per surface suitbleyoffset (so that you could access 2 tracks on each
surface without moving the psoitoned) then I would claim that both of
thos tracks were part of the same sylinder. For example, if you had 4
data surfaces (and thus 8 hards in this drive), I would say that a
'cylinder' contains 8 tracks, even though the arrangement of said
tracks doesn't look like a physical cylinder any more.
Well, drives like that certainly existed, as well as drives that used
a single positioner located between two or more spindles that moved
heads on all drives (think "push pull" or even an X-shaped positioner
arm.
The CDC 808 recorded 6 tracks in parallel (I still have a head from
one) on each head, with IIRC, servicing two spindles on on upper
positioner and two spindles on a lower positioner. Transfer was 12
bits in parallel.
I know that the 821 (a very rare drive; AFAIK they were specified in
some Special Systems RFPs and turned to trash as soon as the 844
drives became available) was similarly configured pretty much the
same in appearance. I don't know if any other parallel drives were
made by CDC--at least I never used any.
--Chuck
7:49 p.m.
New subject: HP drives
On 1/12/11 6:25 PM, Chuck Guzis wrote:
The CDC 808 recorded 6 tracks in parallel (I still
have a head from
one) on each head, with IIRC, servicing two spindles on on upper
positioner and two spindles on a lower positioner. Transfer was 12
bits in parallel.
I know that the 821 (a very rare drive; AFAIK they were specified in
some Special Systems RFPs and turned to trash as soon as the 844
drives became available) was similarly configured pretty much the
same in appearance. I don't know if any other parallel drives were
made by CDC--at least I never used any.
Cray (hey, they're at least related to CDC) made several
parallel-transfer drives; the DD-49 comes to mind in particular.
-Dave
--
Dave McGuire
Port Charlotte, FL
29 Dec
29 Dec
4:02 p.m.
New subject: HP drives
On 12/29/10 10:48 AM, "Tony Duell" <ard at p850ug1.demon.co.uk> wrote:
I really miss my tandy 16 and 6000. Had the whole house set up (when I was
younger) with terminals off my tandy xenix machines, plus a modem on one of
them so we could pull newsgroups :)
And the 2's that had SS drives, called them tracks. So it was habit also
IIRC, when you formatted the disks the trsdos and CP/M called them tracks,
(it has been a LONG time since I've had any tandy 2,12,16 or 6000 in my
posession so that it is foggy memory here.)
On 28 Dec 2010 at 16:52, Geoffrey Reed wrote:
I thought that at least the earlier Tandy 8" machines had single-head
drives, in which case'track' == 'cylinder'.
I can't rememebr which machine it was, but one machine with 5.25" drives
claimed to have '80 track' drives. What they actually were were 40
cylinder, 2 head drives. So technically they did have 80 tracks, but it
caused a lot of confusion at the time.
-tony ISTR: the 8" drives at least as used in
tandy computers formatted out
to 77 tracks....
That's standard for all 8" drives--and it's "cylinders", please!
8"
drives can be double-sided with 2 tracks per cylinder.
30 Dec
30 Dec
10:39 a.m.
New subject: TRS-80 Model II/12/16 (was Re : HP Drives)
I really miss my tandy 16 and 6000. Had the whole
house set up (when I was
younger) with terminals off my tandy xenix machines, plus a modem on one of
them so we could pull newsgroups :)
I am jealous :-)
I grew up with TRS-80s, I had (and still have, of course) a Model 1,
Model 3, Model 4 and varisu CoCos (including a CoCo 3, whcih is not
common in the UK). I still sue my Model 4 for disk conversions, etc.
However, alas I nver had any of the 2/12/16 machines. I would love to
find a Model 16 or similar, but realistically, it's unlikely to happen,
It would ahve to be somewher ethat I could collect it from after all...
-tony
12:13 p.m.
New subject: TRS-80 Model II/12/16 (was Re : HP Drives)
--- On Thu, 12/30/10, Tony Duell <ard at p850ug1.demon.co.uk> wrote:
Yeah... I've been steadily working towards getting a working TRS-80 Xenix machine. At
the moment, I have enough parts (I believe) to make a working Model II, and a line on the
68000 board. Unfortunately, the difficult part to find is a hard disk. I have the Model II
host adapter for the early TRS-80 Eight Meg drive (which is, of course, completely
incompatible with anything else), but no drive. I can't even cobble it to a regular
MFM drive, since the actual disk controller is in the drive cabinet (I have only the host
adapter). Unfortunately, you can't really run Xenix without a hard drive.
-Ian
I really miss
my tandy 16 and
6000. Had the whole house set up (when I was
younger) with terminals off my tandy xenix
machines,
plus a modem on one of
them so we could pull newsgroups :)
I am jealous :-)
1:11 p.m.
New subject: TRS-80 Model II/12/16 (was Re : HP Drives)
Yeah... I've been steadily working towards getting
a working TRS-80
Xenix machine. At the moment, I have enough parts (I believe) to make a
working Model II, and a line on the 68000 board. Unfortunately, the
difficult part to find is a hard disk. I have the Model II host adapter
for the early TRS-80 Eight Meg drive (which is, of course, completely
incompatible with anything else), but no drive. I can't even cobble it
to a regular MFM drive, since the actual disk controller is in the drive
cabinet (I have only the host adapter). Unfortunately, you can't really
run Xenix without a hard drive.
I wonder...
The Model 1/3/4 hard disk system consisted of a WD1001 controller board,
a disk drive and a very simple host adapter (just an address decoder
really). I would not be suprised if the M2 one was similar in concept.
Radio Shack were also very good about supplying technical/service manuals
(one reason I liked their machines, yes even back then I knew the value
of a schematic!), so I suspect some information on the M2 hard disk
system exists.
I wonder how hard itwould be to make up a devie to plug into the host
adapter, appear to have the smae M2-accessible registers, but using some
mode modern storage device. Probsbly a lot easier than replacing an ST412
hard disk, for example.
-tony
1:45 p.m.
New subject: TRS-80 Model II/12/16 (was Re : HP Drives)
On 12/30/10 12:13 PM, "Mr Ian Primus" <ian_primus at yahoo.com> wrote:
Yeah... I've been steadily working towards getting a working TRS-80 Xenix
machine. At the moment, I have enough parts (I believe) to make a working
Model II, and a line on the 68000 board. Unfortunately, the difficult part to
find is a hard disk. I have the Model II host adapter for the early TRS-80
Eight Meg drive (which is, of course, completely incompatible with anything
else), but no drive. I can't even cobble it to a regular MFM drive, since the
actual disk controller is in the drive cabinet (I have only the host adapter).
Unfortunately, you can't really run Xenix without a hard drive.
-Ian
IIRC, the 8MB disk unit used a SA 1000 series 8" hard drive, they had a high
failure rate :( I know that while I had my 16 and 6000 running I had my 8
meg replaced 4 times under warranty. The last time it blew while still under
warranty they replaced the controller and drive unit with a 15 meg 5 1/4"
unit.
29 Dec
29 Dec
9:41 p.m.
New subject: HP drives
From: ard at p850ug1.demon.co.uk
Hi
I always considered track and cylinder to mean the same
for single sided.
For doule sided, Thre would be two tracks per cylinder.
Dwight
On 28 Dec 2010 at 16:52, Geoffrey Reed wrote:
I thought that at least the earlier Tandy 8" machines had single-head
drives, in which case'track' == 'cylinder'.
I can't rememebr which machine it was, but one machine with 5.25" drives
claimed to have '80 track' drives. What they actually were were 40
cylinder, 2 head drives. So technically they did have 80 tracks, but it
caused a lot of confusion at the time.
-tony ISTR: the 8" drives at least as used in
tandy computers formatted out
to 77 tracks....
That's standard for all 8" drives--and it's "cylinders", please!
8"
drives can be double-sided with 2 tracks per cylinder.
30 Dec
30 Dec
10:46 a.m.
New subject: HP drives
I always considered track and cylinder to mean the
same
for single sided.
For doule sided=2C Thre would be two tracks per cylinder.
And I would agree with you...
However, I'll bet that if you ask a reasnablu 'techy' PC person 'how many
tracks are there on a 1.44M vfloppy disk' you will not get the answer
'160' (whcih is the technically correct answer).. Most will answer '80'.
I would probably say 'There are 80 cylinders, each of 2 tracks',
particularly if I suspected it was a trick question :-)
-tony
2:34 p.m.
New subject: HP drives
On Thu, 30 Dec 2010, Tony Duell wrote:
However, I'll bet that if you ask a reasnablu
'techy' PC person 'how many
tracks are there on a 1.44M vfloppy disk' you will not get the answer
'160' (whcih is the technically correct answer).. Most will answer '80'.
I would probably say 'There are 80 cylinders, each of 2 tracks',
particularly if I suspected it was a trick question :-)
Would you settle for "80 per side"?
31 Dec
31 Dec
11:25 a.m.
New subject: HP drives
On Thu, 30 Dec 2010, Tony Duell wrote:
Of course :-). But you, and many others here, are sufficiently clueful to
get it right
-tony
However, I'll bet that if you ask a reasnablu
'techy' PC person 'how many
tracks are there on a 1.44M vfloppy disk' you will not get the answer
'160' (whcih is the technically correct answer).. Most will answer '80'.
I would probably say 'There are 80 cylinders, each of 2 tracks',
particularly if I suspected it was a trick question :-)
Would you settle for "80 per side"?
30 Dec
30 Dec
7:09 p.m.
New subject: HP drives
From: ard at p850ug1.demon.co.uk
Hi
The disk steps to 80 tracks but has 160 tracks on
the media if double sided..
I think the confusion is that when stepping to a track,
it doesn't make any difference to the stepper which
side you start or end on. I'd guess that that is how
the concept of cylinders got mixed up. The media
and the action of the drive are two different things.
Dwight
I always considered track and cylinder to mean
the same
for single sided.
For doule sided=2C Thre would be two tracks per cylinder.
And I would agree with you...
However, I'll bet that if you ask a reasnablu 'techy' PC person 'how many
tracks are there on a 1.44M vfloppy disk' you will not get the answer
'160' (whcih is the technically correct answer).. Most will answer '80'.
I would probably say 'There are 80 cylinders, each of 2 tracks',
particularly if I suspected it was a trick question :-)
-tony
7:18 p.m.
New subject: HP drives
On 30 Dec 2010 at 19:09, dwight elvey wrote:
The disk steps to 80 tracks but has 160 tracks on
the media if double sided..
I think the confusion is that when stepping to a track,
it doesn't make any difference to the stepper which
side you start or end on. I'd guess that that is how
the concept of cylinders got mixed up. The media
and the action of the drive are two different things.
Does any hard disk literature refer to "cylinders" as "tracks"?
Most refer to the geometry as CHS.
--Chuck
31 Dec
31 Dec
11:45 a.m.
New subject: HP drives
However=2C
I'll bet that if you ask a reasnablu 'techy' PC person 'how ma=
ny=20
tracks are there on a 1.44M vfloppy disk' you
will not get the answer=20
'160' (whcih is the technically correct answer).. Most will answer '80'.=
=20
I would probably say 'There are 80
cylinders=2C each of 2 tracks'=2C=20
particularly if I suspected it was a trick question :-)
=20
-tony
Hi
The disk steps to 80 tracks but has 160 tracks on the media if double sided..
I think the confusion is that when stepping to a track=2C
it doesn't make any difference to the stepper which
side you start or end on. I'd guess that that is how
Correct, and that's really what a 'cylinder' means.
the concept of cylinders got mixed up. The media
and the action of the drive are two different things.
-tony
29 Dec
29 Dec
10:45 a.m.
New subject: HP drives
ISTR: the 8" drives at least as used in tandy
computers formatted out to 77
tracks....
Were there 8" drives that had something other than 77 cylinders? Perhaps
the very early IBM ones with the index holes around the outside edge?
Certainly all the ones I've come across were 77 cylinders.
-tony
2:26 p.m.
New subject: HP drives
On 29 Dec 2010 at 18:45, Tony Duell wrote:
Are you perhaps thinking of the Memorex 651 disks with outer-edge
holes and a large "notch" along one side? Those were 64 cylinders,
32-hard-sector, spinning at 375 RPM. Same FM data rate, though (250
Kbit/sec) as the Shugart-style 8" floppies.
I think some of the early IBM models (did they ever get into
production?) used 8 sector holes along the outer edge.
--Chuck
ISTR: the
8" drives at least as used in tandy computers formatted
out to 77 tracks....
Were there 8" drives that had something other than 77 cylinders?
Perhaps the very early IBM ones with the index holes around the
outside edge? Certainly all the ones I've come across were 77
cylinders.
4:19 p.m.
New subject: HP drives
Chuck Guzis wrote:
I think some of the early IBM [8-inch floppy drive]
models (did they ever get into production?)
used 8 sector holes along the outer edge.
The original 8-inch floppy drive was the IBM 23FD "Minnow", which was
used as a read-only device to load microcode into the 3330 controller,
and other System/370 control units and peripherals. Presumably IBM had
special drives in-house for writing the disks sent to customers.
The disks spun at 90 rpm (vs. 36), and used FM encoding at 33.3 Kbps.
They had 32 tracks at 32 tpi (vs 77 tracks at 48 tpi), eight sectors per
track, and the total formatted capacity was 81,664 bytes.
5:13 p.m.
New subject: HP drives
On 29 Dec 2010 at 17:19, Eric Smith wrote:
The original 8-inch floppy drive was the IBM 23FD
"Minnow", which was
used as a read-only device to load microcode into the 3330 controller,
and other System/370 control units and peripherals. Presumably IBM
had special drives in-house for writing the disks sent to customers.
Did IBM ever use the Memorex 651 drives?
--Chuck
5:47 p.m.
New subject: HP drives
On 12/29/10 5:26 PM, Chuck Guzis wrote:
Are you perhaps thinking of the Memorex 651 disks with outer-edge
holes and a large "notch" along one side? Those were 64 cylinders,
32-hard-sector, spinning at 375 RPM. Same FM data rate, though (250
Kbit/sec) as the Shugart-style 8" floppies.
I think some of the early IBM models (did they ever get into
production?) used 8 sector holes along the outer edge.
I have several boxes of disks (I seem to recall having discussed them
here before) that I *think* are IBM-labeled but I'm not certain, as I
last actually looked at them many years ago. I know I still have them
and I know where they are, but I'm several states away at the moment and
can't check. Anyway, they have a large notch along one side, and sector
holes around the outer perimeter of the disk.
The only thing I'm not 100% certain of is whether or not they're IBM.
I'll check when I can.
-Dave
--
Dave McGuire
Port Charlotte, FL
ISTR: the 8" drives at least as used in tandy
computers formatted
out to 77 tracks....
Were there 8" drives that had something other than 77 cylinders?
Perhaps the very early IBM ones with the index holes around the
outside edge? Certainly all the ones I've come across were 77
cylinders.
6:07 p.m.
New subject: HP drives
On 29 Dec 2010 at 20:47, Dave McGuire wrote:
I have several boxes of disks (I seem to recall
having discussed
them here before) that I *think* are IBM-labeled but I'm not certain,
as I last actually looked at them many years ago. I know I still have
them and I know where they are, but I'm several states away at the
moment and can't check. Anyway, they have a large notch along one
side, and sector holes around the outer perimeter of the disk.
The only thing I'm not 100% certain of is whether or not they're
IBM.
Those certainly sound like Memorex 651 floppies; there's
documentation for the 651 on bitsavers, if you want to compare notes.
--Chuck
28 Dec
28 Dec
5:13 p.m.
New subject: HP drives
On Tuesday (12/28/2010 at 02:45PM -0700), Eric Smith wrote:
Tony Duell wrote:
Yes... I can confirm these facts as I have two of each of those 100tpi
drives here and they connect to a Micropolis designed controller which
served as a reference design for that floppy subsystem. Vector Graphic
deployed it in one of their early floppy based systems as well.
--
Chris Elmquist
Howeer, was there ever a 70 cylinder 5.25"
drive? That is one that was
only designed to use 70 cylinders ? I mean a raw drive, not a
drive+controller like the HP units we're discussing.
I haven't heard of
one, though Micropolis made 77-track 5.25" drives
with 100tpi (vs. 35 or 40 track at 48tpi and 80 track at 96tpi), and a
few other vendors offered compatible drives, e.g. the Tandon TM100-4M,
where the "M" designates it as a 100tpi model.
27 Dec
27 Dec
11:19 a.m.
New subject: Epson drives (Was: C64-C128 CP-M Cartridge Interest
I find it more interesting that some of the 135 tpi
users selected
odd values for the upper limit. HP used 66, I think and Jonos used
the original HP 3.5" drive unit -- the 9121 was designed ot be exactly
compatible with the 82901 5.25" drive unit. The latter was a 35 cylinder
2 head deevce, so 70 physcial tracks. The 3.5" drive therefore only used
70 physicla tracks too. But HP also had this idea f having 'spare' tracks
so that slightly defective disks would appear perfect to the user. And
this was handled by the drive unit. There's soem kind of infromation
stored o nthe inntermost track whih gives the track replacemtn
indromation, disk usage count, etc. No hP maual that I've seen has any
real details of this.
The single-0sided drives in the earlier 9133 uniuts are the same as te
9121 (to the exptent that the firmware EPROM on the controler board is the
same)..The double-sided ones are different, I think you get something
like 77 user cylinders on thsoe (and still some bad track informaiton on
cylidner 79.).
-tony
11:42 a.m.
New subject: Epson drives (Was: C64-C128 CP-M Cartridge Interest
On 27 Dec 2010 at 19:19, Tony Duell wrote:
the original HP 3.5" drive unit -- the 9121 was
designed ot be exactly
compatible with the 82901 5.25" drive unit. The latter was a 35
cylinder 2 head deevce, so 70 physcial tracks. The 3.5" drive
therefore only used 70 physicla tracks too. But HP also had this idea
f having 'spare' tracks so that slightly defective disks would appear
perfect to the user. And this was handled by the drive unit. There's
soem kind of infromation stored o nthe inntermost track whih gives the
track replacemtn indromation, disk usage count, etc. No hP maual that
I've seen has any real details of this.
I wonder if it was universally implemented on HP gear. I've seen HP
150 Series I floppies with clusters marked bad the usual MS-DOS way.
You'd think that such a thing would be impossible with an active
"alternate track" scheme.
--Chuck
12:37 p.m.
New subject: Epson drives (Was: C64-C128 CP-M Cartridge Interest
On 27 Dec 2010 at 19:19, Tony Duell wrote:
The HP150 series 1 (I assume by that you mean the origianl one with the
9" CRT [1]) used the normal HP9121, etc, floppy drive units which
certainly did this.
[1] I could say 'like the one that's in lots of bits on my bench at the
moment, but that wouldn't help identify it really :-).
the original HP 3.5" drive unit -- the 9121
was designed ot be exactly
compatible with the 82901 5.25" drive unit. The latter was a 35
cylinder 2 head deevce, so 70 physcial tracks. The 3.5" drive
therefore only used 70 physicla tracks too. But HP also had this idea
f having 'spare' tracks so that slightly defective disks would appear
perfect to the user. And this was handled by the drive unit. There's
soem kind of infromation stored o nthe inntermost track whih gives the
track replacemtn indromation, disk usage count, etc. No hP maual that
I've seen has any real details of this.
I wonder if it was universally implemented on HP gear. I've seen HP
150 Series I floppies with clusters marked bad the usual MS-DOS way. You'd think that such a thing would be impossible
with an active
"alternate track" scheme.
AFAIK, the alternate tracks were only assifend when the disk was
formated, and it was totally transpararent to the user. You ended uup
witha 66 (or whatever) track disk, the drive unit took care of actaulyl
gettign the head to the right physical track. If you haf no bad tracks
during formating (which happens most, if not all, of the time), then it
used the outside 66 tracks so everything looks normal.
-tony
11:12 a.m.
[PF10]
Is it? I thought it turned at 300rpm and used 250kbps MFM recording like
most 3.5" drives of the time.
This, as I am sure you know, is a PQFP package. it also has to be the
right version to get the I/O ports it needs. I still have the dead chip,
I can post the number if anyone thinks they could get one.
-tony
It's also
odd in that it's a 40cylinder (67.5 tpi) device.
its first generation 3.5", also runs slower data rate. [1] The microcontorller has failed. Even though
it's ROMless, this is not
much of a help. Finding the right version i nthe right package after all
thses years has proved impossible. I am darn sure it's the
microcontroller in that I pulled the ROM and forced NOPs onto the data
bus, and it still had an address bus that looked crazy.
6303 is easy to find, I have a few from old defunct HDs. 
10 Feb
10 Feb
5:41 p.m.
On Friday 24 December 2010 09:19:41 pm Fred Cisin wrote:
On Fri, 24 Dec 2010, Alexandre Souza - Listas wrote:
Yes! I remember one time a guy who'd acquired an Osborne 1 with no floppies, so we
assisted in his acquiring some. There was a user group that met in the building as well.
He was astounded, coming from peecees, how small all those utilieis were.
And as reasonably zippy as those 4 MHz machines were, it should be interesting to see how
one would be with a faster part -- I remember hearing about a 20 or 25 MHz Z80 a while
back, and some other implementations that ran even faster.
What is needed to CREATE (and not build) a
CP/M compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc discussing
it?
You don't NEED 64K. CP/M can run in 16K. But, if you want a decent sized
TPA, you can have 50+ K TPA with 64K, or even 63K TPA with 128K.
8080. Yeah, a Z80 is NICE. and FAST! and some aps insist on it.
I think I have a couple.
But, CP/M can run on 8080, 8085, and anything else,
including emulators that
can handle 8080 code.
It's too bad that nobody brought up CP/M on the Kyoceras (Radio Shack
model 100, Nec 8200, etc). THAT would have been a project.
Or on the Epson RC20!
That would indeed have been interesting. The Epson PX-8 that I have is the furthest
I've been in that direction, a model 100 would have been interesting to play with as
well.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
24 Dec
24 Dec
7:54 p.m.
On 12/24/10 8:47 PM, Alexandre Souza - Listas wrote:
I'll try to ask a different question here
What is needed to CREATE (and not build) a CP/M compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc
discussing it?
Very little is required beyond a CPU and memory. First, a Z80 isn't
required; you can use an 8080 or an 8085. (or one of the later Z80
implementations) You also don't need 64K; CP/M will run in much less,
16KB I think was the minimum.
Beyond that, you need some sort of ASCII textual console device (like
a serial port, or a video subsystem + keyboard interface) and a
mass-storage device.
That's it!
-Dave
--
Dave McGuire
Port Charlotte, FL
8:13 p.m.
> Now it got me puzzled. I have some circuit
adaptations of MSX and TRS-80
> computers, as minimal computers to run CP/M code. But what is needed -
> beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc
> discussing it?
On Fri, 24 Dec 2010, Dave McGuire wrote:
Very little is required beyond a CPU and memory.
First, a Z80 isn't
required; you can use an 8080 or an 8085. (or one of the later Z80
implementations) You also don't need 64K; CP/M will run in much less,
16KB I think was the minimum.
Beyond that, you need some sort of ASCII textual console device (like
a serial port, or a video subsystem + keyboard interface) and a
mass-storage device.
That's it!
Well, . . . CP/M kinda assumes that it will have contiguous RAM for its
TPA, starting at location 0 (The "relocated" CP/Ms for UNMODIFIED TRS80
don't really count) That means that in general, the BIOS (usually ROM)
needs to be at the top, and I/O should stay out of the way (avoid memory
mapped I/O anywhere but the top)
The BIOS is the key to being able to use all kinds of bizarre hardware.
The BIOS needs to be able to do console I/O a character at a time, and
block I/O (disk "sectors" that don't really need to be disk nor sectors)
128 bytes at a time. Study the source of some BIOS's - if you can create
a CPM compatible BIOS for it, then it is CP/M compatible.
(the above statement is a little over-simplified)
8:24 p.m.
On 12/24/2010 9:54 PM, Dave McGuire wrote:
On 12/24/10 8:47 PM, Alexandre Souza - Listas wrote:
Is there a link handy to a minimal schematic. Alexandre's question
piqued my curiosity, so I went looking, thinking there would be a ez80
Acclaim + SRAM + SD, using the RS232 for a terminal. I can't find
anything that minimal.
If true (I'm happy to be wrong), I agree a small board with the minimal
components might be useful to a wide range of people. But, it would
require someone who can write a BDOS for it.
Jim
--
Jim Brain, Brain Innovations
brain at jbrain.com
www.jbrain.com
www.jbrain.net (eStore)
I'll try to ask a different question here
What is needed to CREATE (and not build) a CP/M compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc
discussing it?
Very little is required beyond a CPU and memory. First, a Z80 isn't
required; you can use an 8080 or an 8085. (or one of the later Z80
implementations) You also don't need 64K; CP/M will run in much less,
16KB I think was the minimum.
Beyond that, you need some sort of ASCII textual console device
(like a serial port, or a video subsystem + keyboard interface) and a
mass-storage device.
That's it!
-Dave
25 Dec
25 Dec
6:26 a.m.
On Fri, 24 Dec 2010, Jim Brain wrote:
Is there a link handy to a minimal schematic.
Alexandre's question piqued my
curiosity, so I went looking, thinking there would be a ez80 Acclaim + SRAM +
SD, using the RS232 for a terminal. I can't find anything that minimal.
The ez80 project appears to have been "closed source".
If true (I'm happy to be wrong), I agree a small
board with the minimal
components might be useful to a wide range of people. But, it would require
someone who can write a BDOS for it.
At the risk of appearing negative, I do not think the world needs another
single-board CP/M computer at this moment in time.
In any event, I think it's the BIOS that will be required. There are
plenty of free options for CCP, BDOS and utility suites.
Steve
--
26 Dec
26 Dec
7:33 p.m.
On 12/25/2010 8:26 AM, Steven Hirsch wrote:
On Fri, 24 Dec 2010, Jim Brain wrote:
That may very well be.
I figured if I developed a cart that could
function standalone, then I have two chances for usefulness.
If true (I'm happy to be wrong), I agree a
small board with the
minimal components might be useful to a wide range of people. But,
it would require someone who can write a BDOS for it.
At the risk of appearing negative, I do not think the world needs
another single-board CP/M computer at this moment in time.
In any event, I think it's the BIOS that will be required. There are
plenty of free options for CCP, BDOS and utility suites.
Yes, as others had noted.
I'll not feel bad, as I started my thread
noting I have a very weak understanding of CP/M.
That said, I think there is a market at least for an open source
schematic of a minimal CP/M system. Something that used SD cards, had 2
UARTS (one for terminal, one for modem/etc.). People like me could use
it as the basis for learning the environment without needing to find a
classic machine, and then not knowing how to prepare floppies for it. I
can attest that many more folks are interested in their CBM machines now
that there are CF and SD (FAT-compatible) drives for them.
Of course, I should point out at this point that I release all of my
designs as open source, to guard against orphan projects/products with
no ability to re-use. Thus, if someone is not a fan of open source, we
probably won't find much in common on such a project.
Jim
25 Dec
25 Dec
9:24 a.m.
On 12/24/10 11:24 PM, Jim Brain wrote:
Is there a link handy to a minimal schematic.
Alexandre's question
piqued my curiosity, so I went looking, thinking there would be a ez80
Acclaim + SRAM + SD, using the RS232 for a terminal. I can't find
anything that minimal.
I could draw one up.
For the eZ80Acclaim! chips, there's at least one port of CP/M that
I've seen floating around. Those chips are pretty easy to design with;
I've done a few commercial designs with the eZ80F91 (the member of the
family that runs at 50MHz and has an on-chip Ethernet controller).
If true (I'm happy to be wrong), I agree a small
board with the minimal
components might be useful to a wide range of people. But, it would
require someone who can write a BDOS for it.
BIOS, not BDOS...The BDOS is supplied as a standard, not-modifiable
part of CP/M. In any case, I have done several BIOSes (including two
for homebrew Z80 SBCs) and can help with that.
-Dave
--
Dave McGuire
Port Charlotte, FL
11:33 a.m.
On 12/25/2010 11:24 AM, Dave McGuire wrote:
On 12/24/10 11:24 PM, Jim Brain wrote:
I would appreciate that.
Is there a link handy to a minimal schematic.
Alexandre's question
piqued my curiosity, so I went looking, thinking there would be a ez80
Acclaim + SRAM + SD, using the RS232 for a terminal. I can't find
anything that minimal.
I could draw one up.
For the eZ80Acclaim! chips, there's at least one port of CP/M that
I've seen floating around. Those chips are pretty easy to design
with; I've done a few commercial designs with the eZ80F91 (the member
of the family that runs at 50MHz and has an on-chip Ethernet controller).
In any case, I have done several BIOSes (including two for homebrew
Z80 SBCs) and can help with that.
That would be grand. There is a small but
passionate group of folks who
would enjoy this, but I'm not a BIOS person, I barely know Z80 assembler.
26 Dec
26 Dec
6:12 p.m.
On 12/25/10 2:33 PM, Jim Brain wrote:
Writing a CP/M BIOS is surprisingly easy, but knowing Z80 assembler
would be a good start. :)
I learned CP/M BIOS hacking from "The Programmer's CP/M Handbook" by
Andy Johnson-Laird, perhaps the very best CP/M book I've ever seen.
-Dave
--
Dave McGuire
Port Charlotte, FL
In any case, I
have done several BIOSes (including two for homebrew
Z80 SBCs) and can help with that.
That would be grand. There is a small but passionate group of folks who
would enjoy this, but I'm not a BIOS person, I barely know Z80 assembler.
25 Dec
25 Dec
11:39 a.m.
On 12/25/10 9:24 AM, "Dave McGuire" <mcguire at neurotica.com> wrote:
On 12/24/10 11:24 PM, Jim Brain wrote:
How are you at Z80 disassembly? It would be nice to get the ATR8000 BIOS
disassembled and turned into something somewhat modular so it could be
redesigned with modern parts :)
Is there a link handy to a minimal schematic.
Alexandre's question
piqued my curiosity, so I went looking, thinking there would be a ez80
Acclaim + SRAM + SD, using the RS232 for a terminal. I can't find
anything that minimal.
I could draw one up.
For the eZ80Acclaim! chips, there's at least one port of CP/M that
I've seen floating around. Those chips are pretty easy to design with;
I've done a few commercial designs with the eZ80F91 (the member of the
family that runs at 50MHz and has an on-chip Ethernet controller).
If true (I'm happy to be wrong), I agree a
small board with the minimal
components might be useful to a wide range of people. But, it would
require someone who can write a BDOS for it.
BIOS, not BDOS...The BDOS is supplied as a standard, not-modifiable
part of CP/M. In any case, I have done several BIOSes (including two
for homebrew Z80 SBCs) and can help with that.
-Dave
26 Dec
26 Dec
6:07 p.m.
On 12/25/10 2:39 PM, Geoffrey Reed wrote:
Heh, I used to be pretty good at it, but it's probably been 20-22
years since I reverse-engineered any Z80 code. I'm not sure I'd tackle
something like that at this point. I'm doing plenty of Z80 *writing*
these days, but that's a different story entirely.
Now, if someone were to slide me an ATR8000, that'd provide some
serious motivation. ;)
-Dave
--
Dave McGuire
Port Charlotte, FL
BIOS, not
BDOS...The BDOS is supplied as a standard, not-modifiable
part of CP/M. In any case, I have done several BIOSes (including two
for homebrew Z80 SBCs) and can help with that.
How are you at Z80 disassembly? It would be nice to get the ATR8000 BIOS
disassembled and turned into something somewhat modular so it could be
redesigned with modern parts :)
25 Dec
25 Dec
11:52 a.m.
On 12/25/2010 12:24 PM, Dave McGuire wrote:
The CP/M system is three parts. CCP (console monitor or user interface),
BDOS (actual device and files system
logical interface and applications interface), BIOS( hardware
abstraction unique to the local hardware to match the abstracted
interface of the BDOS). User usually did not rewrite the CCP or BDOS
just relocate it with MOVCPM.
The BIOS was unique to the local system and matched the calls for
console in, console out, console status, same for printer, punch, reader
, and then the interface to the mass storage (setdma, SETtrack,
Setsector, SECTRAN(logical to physical translate for sector skewing on
floppies) READ, WRITE.
FOR 8085/Z80 the CCP and BDOS are a give block of binary (we officially
have sources now) and the BIOS is written
as needed. A BIOS can be complex or simple depending on hardware and
user desires. Oddly it takes more code to do something like a floppy as
you have to handhold the FDC chip all the way. smarter devices like
IDE, CF, SD
are easier and the only real task is to match the CP/M logical 128byte
block to the devices 512byte standard
block (physical sector). The latter is fairly easy, the code is even in
the CP/M alteration guide.
I've implemented CPm on Z80 with less than 16 pcs TTL, memory, EPROM and
CPU total and that was using
32K ram x2 (there are 64K parts!). The IDE, CF or SD interfaces can be
pretty trivial with SD and CF being
preferred as CF does do an 8bit mode, and SD is a serial interface with
only a few lines. That leaves the
more interesting IO, user console, and ports to run a printer or
communicate with maybe a PC. [note
one port with the right protocol can do all that if the host is playing
terminal, printer, storage].
Allison
On 12/24/10 11:24 PM, Jim Brain wrote:
I looked at that port, it's not Z80 compatible. CP/M and all had to be
re-assembled
for the eZ80. Same would be required to run many of the apps and utilities.
Is there a link handy to a minimal schematic.
Alexandre's question
piqued my curiosity, so I went looking, thinking there would be a ez80
Acclaim + SRAM + SD, using the RS232 for a terminal. I can't find
anything that minimal.
I could draw one up.
For the eZ80Acclaim! chips, there's at least one port of CP/M that
I've seen floating around. Those chips are pretty easy to design
with; I've done a few commercial designs with the eZ80F91 (the member
of the family that runs at 50MHz and has an on-chip Ethernet controller).
If true
(I'm happy to be wrong), I agree a small board with the minimal
components might be useful to a wide range of people. But, it would
require someone who can write a BDOS for it.
BIOS, not BDOS...The BDOS is supplied as a standard, not-modifiable
part of CP/M. In any case, I have done several BIOSes (including two
for homebrew Z80 SBCs) and can help with that.
12:11 p.m.
On 12/25/2010 1:52 PM, allison wrote:
I looked at that port, it's not Z80 compatible.
CP/M and all had to
be re-assembled
for the eZ80. Same would be required to run many of the apps and
utilities.
Well, that's a disappointment. It looked like such a time saver.
I've implemented CPm on Z80 with less than 16 pcs TTL, memory, EPROM
and CPU total and that was using
32K ram x2 (there are 64K parts!). The IDE, CF or SD interfaces can
be pretty trivial with SD and CF being
preferred as CF does do an 8bit mode, and SD is a serial interface
with only a few lines. That leaves the
more interesting IO, user console, and ports to run a printer or
communicate with maybe a PC. [note
one port with the right protocol can do all that if the host is
playing terminal, printer, storage].
Maybe I should just see if there is a schematic of a minimal Z80 system
(fewest ICs possible) and go from there.
Jim
12:34 p.m.
On 25 Dec 2010 at 14:52, allison wrote:
I looked at that port, it's not Z80 compatible.
CP/M and all had to
be re-assembled for the eZ80. Same would be required to run many of
the apps and utilities.
I've looked at the eZ80 CPU documentation; do you have any
information on what differences there are? From the manual, it seems
that one can opt for extended 16/24 bit mode or standard 64K Z80 mode-
-and switch between them.
I've also run across a couple of web posts that indicate that the
Acclaim does in fact run CP/M.
So I'm mystified.
--Chuck
4:11 a.m.
I'll try to ask a different question here
What is needed to CREATE (and not build) a CP/M compatible computer?
Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
computers, as minimal computers to run CP/M code. But what is needed -
beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc discussing
it?
Actaully, AFAIK you don't need either of those...
An 8080 procesosr will do, so will an 8085 to run CP/M. However some
application programs did use the Z80-only instructions so it probably
makes a lot of sense ot sue that processor
You don't need 64K RAM. There have been 32K or 48K CP/M machines, it
wouldn't suprise me if 16K was done too. Again, though, it's probably
sensible to have 64K these days.
Unless you want to run a modified CP/M and applications, you need RAM
starting at location 0. Since the Z80 starts at location 0 when you reset
it, the normal thing to do is to have a boot ROM [1] that replaces RAM at
locaiton 0 on resetm and is then switched out 9or relocated to the top fo
the memory map) once the machine has loaded the OS into RAM.
[1] Some machines -- I think Cromemco was one -- had a little bit of
hardware to force a jump instruction onto the processor data bus
immediately after rest, thus forcing th CPU to execute thr boot ROM at
the tom of the memory map. Whether you want to call this a 'ROM at
location 0' is up to you ;-)
You presumably need some kidn of mass storage. It doesn't have to be
floppy disks.
You need a text terminal or something equivelent to it. You can have a
seiral port to a normal terminal, or memory mapped video, or whatever
you like.
And I think that's about it.
-tony
6:27 a.m.
On 12/25/2010 07:11 AM, Tony Duell wrote:
>
> I'll try to ask a different question here
>
> What is needed to CREATE (and not build) a CP/M compatible computer?
>
> Now it got me puzzled. I have some circuit adaptations of MSX and TRS-80
> computers, as minimal computers to run CP/M code. But what is needed -
> beyond Z80 and 64K of RAM - to run CPM? Is there any kind of doc discussing
> it?
Minimal system:
8080, 8085, NsC800, Z80, or Z180 cpu. The prefered CPU is z80 as
it's easy to interface and there are some CP/M apps that require it.
64K ram, (technically cp/m runs in 20k or more but useful apps are
minimally 48K)
Boot rom(eprom/flash) that maps out of the 0000h startup address
Mass storage, this can be floppy, CF, SD, Or some form of rom/ram to
look like disk.
User IO, typically a serial line for a terminal can be a memory mapped video
and keyboard or similar.
CP/M is self contained, howeer for any hardware configuration there is
a thing
called BIOS and is unique to that hardware. The BIOS provides hardware
abstraction
layer from the physical to the logical interface that CP/M expects. For
most systems
it can be trivial.
The base TRS80 cannot do that as all the IO (keyboard, video) and rom
are mapped
into the low ram address space. it's possible to hack that to high
space and dump
the tTRS roms but thats a bit of work.
Allison
Actaully, AFAIK you don't need either of those...
An 8080 procesosr will do, so will an 8085 to run CP/M. However some
application programs did use the Z80-only instructions so it probably
makes a lot of sense ot sue that processor
You don't need 64K RAM. There have been 32K or 48K CP/M machines, it
wouldn't suprise me if 16K was done too. Again, though, it's probably
sensible to have 64K these days.
Unless you want to run a modified CP/M and applications, you need RAM
starting at location 0. Since the Z80 starts at location 0 when you reset
it, the normal thing to do is to have a boot ROM [1] that replaces RAM at
locaiton 0 on resetm and is then switched out 9or relocated to the top fo
the memory map) once the machine has loaded the OS into RAM.
[1] Some machines -- I think Cromemco was one -- had a little bit of
hardware to force a jump instruction onto the processor data bus
immediately after rest, thus forcing th CPU to execute thr boot ROM at
the tom of the memory map. Whether you want to call this a 'ROM at
location 0' is up to you ;-)
You presumably need some kidn of mass storage. It doesn't have to be
floppy disks.
You need a text terminal or something equivelent to it. You can have a
seiral port to a normal terminal, or memory mapped video, or whatever
you like.
And I think that's about it.
-tony
10:18 a.m.
On 25 Dec 2010 at 9:27, allison wrote:
8080, 8085, NsC800, Z80, or Z180 cpu. The prefered
CPU is z80 as it's
easy to interface and there are some CP/M apps that require it.
If you're starting off from scratch and want the smallest parts-count
system, why not an eZ80 Acclaim? 256KB of flash, 24-bit external
memory interface, built-in 100BaseT ethernet MAC, 2 UARTS, IrDA, lots
of GPIO pins, SPI, I2C and a PLL clock generator good to 50MHz as
well as ICE support and separately-powered RTC. Still Z80-compatible
(or the enhanced instruction set if you want it).
I suspect that you could dump most of what you need in the way of
CP/M programs into the on-chip flash. Add an external SRAM and
external interface "glue" and you're done, but for the programming.
--Chuck
11:30 a.m.
On 12/25/2010 01:18 PM, Chuck Guzis wrote:
On 25 Dec 2010 at 9:27, allison wrote:
You could if the EZ80 was compatible, in the extended mode its a
different beast. Since most CP/M apps are available as source
it's doable and could result in an interesting hybrid.
Really depends if you want to build with available through hole parts
and simple design or do a new and different thing. Neither is wrong,
just different paths.
Generally, a minimal cp/m system needs about 128K of "disk" for
common utilities and some tools. However thats a bit minimal.
I find none of the CP/M sytems I've used {still use} over the years
are limited if the disk is cramped. Whats cramped? Under about 500K.
At about 1MB it's decent, assuming you have at least 128K or more of
RW media. With large rams, CF or SD that can easily be large enough
to exceed the limits of standard CP/M2 (16 devices of 8MB each or
128MB. However I've been using large IDEs and CFs (16mb to 500mb)
for years and just mapping them in 8mb chunks (external partition table,
a new drive every +16384 LBA) and a simple utility to remap physical
to logical. the result is tiny, fast and runs anything CP/M.
The biggest issue is not the memory or mass storage system anymore.
It's IO, such as serial line for terminal (did that using FTDI serial to
usb)
I've done both bit bash serial and various serial devices (8250, SIO, 8251).
I've also tried memory mapped (16x64 like PT s100 board) video,
8Lx40char LCD (cramped), and so on. The ideal here would be USB
keyboard input (mac mini key board is very short) and VGA video where
the logic is I'm always near a VGA display and a USB keyboard. Then
with CMOS Z80 and and other parts, coompact portable is doable.
I believe something that is portable (battery or wall wart power)
is likely to appeal more to many with the right IO.
Allison
8080, 8085, NsC800, Z80, or Z180 cpu. The
prefered CPU is z80 as it's
easy to interface and there are some CP/M apps that require it.
If you're
starting off from scratch and want the smallest parts-count
system, why not an eZ80 Acclaim? 256KB of flash, 24-bit external
memory interface, built-in 100BaseT ethernet MAC, 2 UARTS, IrDA, lots
of GPIO pins, SPI, I2C and a PLL clock generator good to 50MHz as
well as ICE support and separately-powered RTC. Still Z80-compatible
(or the enhanced instruction set if you want it).
I suspect that you could dump most of what you need in the way of
CP/M programs into the on-chip flash. Add an external SRAM and
external interface "glue" and you're done, but for the programming.
--Chuck
12:10 p.m.
On 25 Dec 2010 at 14:30, allison wrote:
You could if the EZ80 was compatible, in the extended
mode its a
different beast. Since most CP/M apps are available as source
it's doable and could result in an interesting hybrid.
Since the eZ80 also has SPI, interfacing to a SDHC card is easy and
gives one gigabytes of storage--a virtual embarrassment of storage.
I seem to recall, however, that the single-density Osborne 1 existed
with two 90KB floppies.
The biggest issue is not the memory or mass storage
system anymore.
It's IO, such as serial line for terminal (did that using FTDI serial
to usb) I've done both bit bash serial and various serial devices
(8250, SIO, 8251). I've also tried memory mapped (16x64 like PT s100
board) video, 8Lx40char LCD (cramped), and so on. The ideal here
would be USB keyboard input (mac mini key board is very short) and VGA
video where the logic is I'm always near a VGA display and a USB
keyboard. Then with CMOS Z80 and and other parts, coompact portable
is doable. I believe something that is portable (battery or wall wart
power) is likely to appeal more to many with the right IO.
The eZ80 Acclaim has UARTs and I think there are more recent members
of the family with USB.
Of course, making use of the 100BaseT ethernet capabilities would
seem to make I/O support rather academic.
But you're probably right about people not wanting to deal with a 144
pin QFP and that's probably the root of the matter.
One could probably do a good job of running 8-bit CP/M with scarcely
more than a suitably-equipped ARM CPU running emulation code and
maybe some external flash device. But that would probably really rub
some semsitivities the wrong way.
All the best,
--Chuck
7:30 a.m.
From: ard at p850ug1.demon.co.uk
----snip---
Unless you want to run a modified CP/M and applications, you need RAM
starting at location 0. Since the Z80 starts at location 0 when you reset
it, the normal thing to do is to have a boot ROM [1] that replaces RAM at
locaiton 0 on resetm and is then switched out 9or relocated to the top fo
the memory map) once the machine has loaded the OS into RAM.
[1] Some machines -- I think Cromemco was one -- had a little bit of
hardware to force a jump instruction onto the processor data bus
immediately after rest, thus forcing th CPU to execute thr boot ROM at
the tom of the memory map. Whether you want to call this a 'ROM at
location 0' is up to you ;-)
---snip---
Hi
I suspect that the reason for this may have been that the original
CP/M was done on computers with 0 based RAM and methods to
load code at 0 on boot.
I don't recall how the MDS800s did this but I do know that early
work on CP/M was done with one of Torode's early controllers.
These were DMA and had there own state machine that
could auto transfer the first boot sector into RAM on reset, independent
of what processor was in the machines.
I have one of his later controllers in my IMSAI and that is how it
works.
Adding hardware to do the ROM/RAM transfer on boot is relatively
easy with the addition of a 7474. Most machines decode a I/O
address. You use the first I/O to toggle the 7474 and that turns
on the RAM and the ROM off. Reset is connected to the 7474.
Most ROM and RAM boards use some type of comparator to
determine the addess ( a 74688 was what I recall ). Normally
the input was tied to ground. This could be tied to the 7474.
On newer boards, there was the phantom signal.
The boot ROM would transfer all the needed ROM to RAM
and then switch to all RAM.
Dwight
10:25 a.m.
I suspect that the reason for this may have been that
the original
CP/M was done on computers with 0 based RAM and methods to
load code at 0 on boot.
I beleive the historay can be traced back to the MCS8i system. This,
AFAIK, never had disks, and never ran CP/M. But it's 8080-based and
amazingly it has a CPM-like IOBYTE at location 3.
This machine has RAM at location 0 by default. The thing is that the 8080
starts executing at location 0, so by puittign RAM there you could enter
a program (or at least a jJMP instruction) at location 0, hit reset, and
run it. That's what you did on the MCS8i. Since a jump is 3 bytes long,
it occupied locations 0.1.2, so location 3 was the first 'free' location,
hence used for IOBYTE. To run the ROM monitor on the MSC8i, you toggled
C3 00 38 (jump to 0x3800) into the first 3 locations, reset, and let it run.
The MDS800, for all it didn't have a front panel, kept the same sort of
memory map (for good reasons IMHO).
FWIW, I have both machines with manuals...
Adding hardware to do the ROM/RAM transfer on boot is
relatively
easy with the addition of a 7474. Most machines decode a I/O
I've seen it done in all sorts of ways. Including a little state machine
that picks up the first 3 read cycles after a reset, disables data
buffers, etc, so they never go outside the CPU board, and forces a jump
instruciton onto the processor data lines. Or, as you say, having a
flip-flop that's dset one way on reset to enable a ROM at location 0 and
the other way by some other method. One trick was to have the ROM always
accessible at the top of memory. On a reset, it was also eneables at
locaiton 0 (or it filled the memory map). The frirst instruction in the
ROM wasa jump to the 'real' ROM location (let's say it was 0xE000) and
the first read cycle with A15 set (or something) cleared the flip-flop,
thus making the ROM appear only at the top of the memory map and enabling
RAM elsewhere. And there are all sorts of other ways to do it.
-tony
11:11 a.m.
On 12/25/2010 01:25 PM, Tony Duell wrote:
This is the common non rom approach for most Z80 S100 cpus.
Another is to invert A15, 14, 13, 12 on reset and hold it with a
D-flop, then
the first bytes of Rom do a jump to F003 where the rom address decode
clear the FF with the PC holding F003 where the actual start of the
monitor is.
My favorite for 8085 and Z80 systems I build is to put EProm/flash/EEROM at
0000 and use a port with a latch to map a ram into 0000 space after loading
I suspect
that the reason for this may have been that the original
CP/M was done on computers with 0 based RAM and methods to
load code at 0 on boot.
I beleive the historay can be traced back to the MCS8i
system. This,
AFAIK, never had disks, and never ran CP/M. But it's 8080-based and
amazingly it has a CPM-like IOBYTE at location 3.
This machine has RAM at location 0 by default. The thing is that the 8080
starts executing at location 0, so by puittign RAM there you could enter
a program (or at least a jJMP instruction) at location 0, hit reset, and
run it. That's what you did on the MCS8i. Since a jump is 3 bytes long,
it occupied locations 0.1.2, so location 3 was the first 'free' location,
hence used for IOBYTE. To run the ROM monitor on the MSC8i, you toggled
C3 00 38 (jump to 0x3800) into the first 3 locations, reset, and let it run.
The MDS800, for all it didn't have a front panel, kept the same sort of
memory map (for good reasons IMHO).
FWIW, I have both machines with manuals...
Adding hardware to do the ROM/RAM transfer on
boot is relatively
easy with the addition of a 7474. Most machines decode a I/O
I've seen it done
in all sorts of ways. Including a little state machine
that picks up the first 3 read cycles after a reset, disables data
buffers, etc, so they never go outside the CPU board, and forces a jump
instruciton onto the processor data lines. Or, as you say, having a from rom to high ram. this works very nicely with 32K
Rom/Eprom and Ram
devices commonly around. it also allows the rom to be large as
32K (or
32K pages)
to allow things like rom disk and other address space stretching tricks.
There are many more ways to skin the problem of ram at zero and reset
starting the
PC at 0000.
Allison
11:31 a.m.
On 12/25/2010 1:11 PM, allison wrote:
There are many more ways to skin the problem of ram at zero and reset
starting the
PC at 0000.
I like the latch idea, where it resets to map the ROM into 0000, but
the
code can then latch the RAM in. As well, this would allow the code to
bring the ROM back into the map at the lower space if needed.
12:02 p.m.
Tony Duell wrote:
I beleive the historay can be traced back to the MCS8i
system. This,
AFAIK, never had disks, and never ran CP/M. But it's 8080-based and
amazingly it has a CPM-like IOBYTE at location 3.
It's not that the Intel development systems had a CP/M-link IOBYTE.
It's that CP/M has an Intel MDS-like IOBYTE. Remember that Gary Kildall
wrote software for Intel before writing CP/M, and that CP/M was written
to run on an MDS.
Eric
12:18 p.m.
On 25 Dec 2010 at 13:02, Eric Smith wrote:
Tony Duell wrote:
...and IOBYTE wasn't universally implemented among OEMs. It was an
optional feature of the BIOS in the early days of CP/M. I can
probably find some words in the System Alteration Guide to that
effect, if anyone's interested. Other than STAT and PIP, I'm not
aware of any standard CP/M utility that cares about it--or supports
the additional "extended" device names.
--Chuck
I beleive the historay can be traced back to the
MCS8i system.
This, > AFAIK, never had disks, and never ran CP/M. But it's
8080-based and > amazingly it has a CPM-like IOBYTE at location 3.
It's not that the Intel development systems had a CP/M-link IOBYTE.
It's that CP/M has an Intel MDS-like IOBYTE. Remember that Gary
Kildall wrote software for Intel before writing CP/M, and that CP/M
was written to run on an MDS
12:33 p.m.
Tony Duell wrote:
Sure. The MCS8i and DMS800 came before CP/M...
The only reason I said 'CPM-like' is that few people hwere have read the
MCS8i monitor soruces.... Rather more know CPM.
-tony
I beleive the historay can be traced back to the
MCS8i system. This,
AFAIK, never had disks, and never ran CP/M. But it's 8080-based and
amazingly it has a CPM-like IOBYTE at location 3.
It's not that the Intel development systems had a CP/M-link IOBYTE.
It's that CP/M has an Intel MDS-like IOBYTE. Remember that Gary Kildall
wrote software for Intel before writing CP/M, and that CP/M was written
to run on an MDS. 
12:38 p.m.
dwight elvey said the following on 12/25/2010 10:30 AM:
---snip---
Hi
I suspect that the reason for this may have been that the original
CP/M was done on computers with 0 based RAM and methods to
load code at 0 on boot.
I don't recall how the MDS800s did this but I do know that early
work on CP/M was done with one of Torode's early controllers.
These were DMA and had there own state machine that
could auto transfer the first boot sector into RAM on reset, independent
of what processor was in the machines.
I have one of his later controllers in my IMSAI and that is how it
works.
Adding hardware to do the ROM/RAM transfer on boot is relatively
easy with the addition of a 7474. Most machines decode a I/O
address. You use the first I/O to toggle the 7474 and that turns
on the RAM and the ROM off. Reset is connected to the 7474.
Most ROM and RAM boards use some type of comparator to
determine the addess ( a 74688 was what I recall ). Normally
the input was tied to ground. This could be tied to the 7474.
On newer boards, there was the phantom signal.
The boot ROM would transfer all the needed ROM to RAM
and then switch to all RAM.
Dwight
While I remember, at least I *think* I remember, I'll fill in one
question here. The MDS800 had a switch on the front panel called
"boot." When in the up position it switched a small EPROM (1702) into
low memory that had the initial boot code in it. That was enough to
load the first track of the floppy disk into RAM. That boot switch only
affected the *read* cycles. *Writes* were still deposited into the RAM
at low memory.
Once the boot code from the disk was read into low RAM the execution
from the 1702 went into a tight loop waiting. There
was an indication
of that in an LED on the front panel, INT2 as I recall. The user
would
toggle the "boot" switch and execution would continue but now from the
ram that was just loaded. The boot process would finish by bringing in
the operating system from the disk, either ISIS-II, CP/M, FORTH, or
whatever you were using.
Pretty sure I remember that correctly. Wish I never sold mine. It was
definitely a classic!
Dave
5:11 p.m.
Date: Sat, 25 Dec 2010 15:38:05 -0500
From: dmabry at mich.com
To:
Subject: Re: C64-C128 CP-M Cartridge Interest
dwight elvey said the following on 12/25/2010 10:30 AM:
Hi Dave
I think you are correct. It has been a long time since I booted an MDS800.
Still, it is well documented that the early Kildall's machines used Torode's
disk controller that would DMA the first sector into RAM on reset.
This was before he had a MDS800 as was used for later versions of
CP/M. I would guess that Kildall's first machines had no ROMs in them
at all.
Dwight
---snip---
Hi
I suspect that the reason for this may have been that the original
CP/M was done on computers with 0 based RAM and methods to
load code at 0 on boot.
I don't recall how the MDS800s did this but I do know that early
work on CP/M was done with one of Torode's early controllers.
These were DMA and had there own state machine that
could auto transfer the first boot sector into RAM on reset, independent
of what processor was in the machines.
I have one of his later controllers in my IMSAI and that is how it
works.
Adding hardware to do the ROM/RAM transfer on boot is relatively
easy with the addition of a 7474. Most machines decode a I/O
address. You use the first I/O to toggle the 7474 and that turns
on the RAM and the ROM off. Reset is connected to the 7474.
Most ROM and RAM boards use some type of comparator to
determine the addess ( a 74688 was what I recall ). Normally
the input was tied to ground. This could be tied to the 7474.
On newer boards, there was the phantom signal.
The boot ROM would transfer all the needed ROM to RAM
and then switch to all RAM.
Dwight
While I remember, at least I *think* I remember, I'll fill in one
question here. The MDS800 had a switch on the front panel called
"boot." When in the up position it switched a small EPROM (1702) into
low memory that had the initial boot code in it. That was enough to
load the first track of the floppy disk into RAM. That boot switch only
affected the *read* cycles. *Writes* were still deposited into the RAM
at low memory.
Once the boot code from the disk was read into low RAM the execution
from the 1702 went into a tight loop waiting. There was an indication
of that in an LED on the front panel, INT2 as I recall. The user would
toggle the "boot" switch and execution would continue but now from the
ram that was just loaded. The boot process would finish by bringing in
the operating system from the disk, either ISIS-II, CP/M, FORTH, or
whatever you were using.
Pretty sure I remember that correctly. Wish I never sold mine. It was
definitely a classic!
Dave
4830
days inactive
4878
days old
91 comments
20 participants
participants (20)
-
ajp166＠verizon.net -
aliensrcooluk＠yahoo.co.uk -
ard＠p850ug1.demon.co.uk -
brain＠jbrain.com -
cc＠informatik.uni-stuttgart.de -
cclist＠sydex.com -
chrise＠pobox.com -
cisin＠xenosoft.com -
dkelvey＠hotmail.com -
dmabry＠mich.com -
eric＠brouhaha.com -
geoffr＠zipcon.net -
ian_primus＠yahoo.com -
lynchaj＠yahoo.com -
mcguire＠neurotica.com -
michael_holley＠mentor.com -
pu1bzz.listas＠gmail.com -
rtellason＠verizon.net -
scheefj＠netscape.net -
snhirsch＠gmail.com