17 Apr
2000
17 Apr
'00
5:22 p.m.
"Has anyone got sufficient experience with the
IDE in non-PC-compatible
applications to say, unequivocally, whether the 8-bit operating mode
described in section 6 of the standard for the ATA (IDE) interface, yes
SPELLED OUT, actually exists in drives of the pre-1996 vintage?" It was
dropped from the standard in 1996. There seem to be many folks with
suggestions about how to implement this extremely elementary interface.
There seem to be few who know whether the standard was every full
implemented.
Why would they include that? NONE of the PC hardware they were intended
for wants to run as 8bit bus. Even the crippled SL/SLC run as 16bit busses.
Allison seems to be the only one who's tried this,
and, I fear, it may be
in
a PC-compatible, where all bets are off as to what
really happens.
Obviously visually impaired! I don't hack IO in PCs nor have I tried that
yet.
I may add why even bother, IDE works there as is.
ALL of the IDE work I've done was with 8085, z80s either stand alone
(bus less) or S100 Z80s. Further I'm currently working on a Z280 system
with IDE (Zbus 16bit). This is where I need interfaces and so I can replace
older MFM or non-existent hard disks. I currently have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
Connor 2022
Connor 3044
WD2120
WD2420
St3096A
St3144A
St3250A
St3660
Fijitsu 528mb
Quantum LPS 80 and 120
Maxtor 124mb
and afew other sub 60mb WD, NEC, Seagate drives.
I now have two 2.5mb IDE in the 700-800mb region I may try one day.
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Allison
Dick
----- Original Message -----
From: Mike Ford <mikeford(a)socal.rr.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, April 17, 2000 7:22 PM
Subject: Re: 8-bit IDE
I don't know a hoot about this, but I wonder
if taking a look at the
Sequential Systems Focus card for the Apple II might be instuctive. Its a
controller and notebook IDE drive that all fits on a Apple II slot card.
17 Apr
17 Apr
7:11 p.m.
allisonp wrote:
ALL of the IDE work I've done was with 8085, z80s
either stand alone
(bus less) or S100 Z80s. Further I'm currently working on a Z280 system
with IDE (Zbus 16bit). This is where I need interfaces and so I can replace
older MFM or non-existent hard disks. I currently have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
Connor 2022
Connor 3044
WD2120
WD2420
St3096A
St3144A
St3250A
St3660
Fijitsu 528mb
Quantum LPS 80 and 120
Maxtor 124mb
and afew other sub 60mb WD, NEC, Seagate drives.
I now have two 2.5mb IDE in the 700-800mb region I may try one day.
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Jerome Fine replies:
But have you tried to get an IDE drive to work with a Qbus? That would seem
to be an even better goal? Well at least from my point of view! I do admit
to being strictly focused on a PDP-11 as well. I always wondered why none
of the 3rd party manufactures did not do an MSCP emulation. And if you are
still concerned about using MSCP until or before the patent expires, there are
a number of other possibilities that are much less difficult and still allow
a 32 bit block number. Can you at least suggest how easy or difficult it
might be?
Sincerely yours,
Jerome Fine
8:31 p.m.
My experience with Q-Bus I/O was always spoon-fed via a DRV11-WA card. That
serves to isolate my problems from "theirs." Would it be better done
without the DRV11-WA? It would seem to me that, since the DE timing spec is
typically faster than what the DRV11 requires, yet tolerant of slower and
more varied rates than what the DRV11 can do, it should be quite
straighforward, and probably every bit as fast as anything ANY DEC box with
a Q-Bus can manage. The data I/O comes through a single address, so only
that one address has to be word-wide, not to say it can't be. I do wonder
whether the Q-Bus can generate the appropriate addresses, however, without
hardware help. A bit of steering would, I believe, suffice to make the
interface work from a word-wide bus.
Dick
----- Original Message -----
From: Jerome Fine <jhfine(a)idirect.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, April 17, 2000 10:11 PM
Subject: Re: 8-bit IDE
replace
seem
allisonp wrote:
> ALL of the IDE work I've done was with 8085, z80s either stand alone
> (bus less) or S100 Z80s. Further I'm currently working on a Z280 system
> with IDE (Zbus 16bit). This is where I need interfaces and so I can older MFM or
non-existent hard disks. I currently have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
Connor 2022
Connor 3044
WD2120
WD2420
St3096A
St3144A
St3250A
St3660
Fijitsu 528mb
Quantum LPS 80 and 120
Maxtor 124mb
and afew other sub 60mb WD, NEC, Seagate drives.
I now have two 2.5mb IDE in the 700-800mb region I may try one day.
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Jerome Fine replies:
But have you tried to get an IDE drive to work with a Qbus? That would to be an even better goal? Well at least from my
point of view! I do
admit
to being strictly focused on a PDP-11 as well. I
always wondered why none
of the 3rd party manufactures did not do an MSCP emulation. And if you
are
still concerned about using MSCP until or before the
patent expires, there
are
a number of other possibilities that are much less
difficult and still
allow
a 32 bit block number. Can you at least suggest how
easy or difficult it
might be?
Sincerely yours,
Jerome Fine
18 Apr
18 Apr
4:33 a.m.
On Mon, 17 Apr 2000, Richard Erlacher wrote:
My experience with Q-Bus I/O was always spoon-fed via
a DRV11-WA card. That
I find that an awkward interface. Handy though.
serves to isolate my problems from "theirs."
Would it be better done
without the DRV11-WA? It would seem to me that, since the DE timing spec is
yes.
straighforward, and probably every bit as fast as
anything ANY DEC box with
a Q-Bus can manage. The data I/O comes through a single address, so only
that one address has to be word-wide, not to say it can't be. I do wonder
whether the Q-Bus can generate the appropriate addresses, however, without
hardware help. A bit of steering would, I believe, suffice to make the
interface work from a word-wide bus.
Qbus output address at Bsync/ (like 8085 ALE) and you latch the address
off the databus, plus BBS7 (bankselect 7 is the IOpage) and theres your
address. The nest part of the cycle is typically an IO_read with an
optional IO_write to follow. What makes doing IDE on PDP11 is the read
before write to the same address (plays havoc with device resgisters).
the simple solution is to map all reads to base+0 to base+n and writes
to base+N to Base+n+n so reads do not overlap writes.
Allison
6:13 a.m.
please see embedded comments below.
Dick
----- Original Message -----
From: <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 18, 2000 7:33 AM
Subject: Re: 8-bit IDE
On Mon, 17 Apr 2000, Richard Erlacher wrote:
> My experience with Q-Bus I/O was always spoon-fed via a DRV11-WA card.
That
<snip>
Qbus output address at Bsync/ (like 8085 ALE) and you latch the address
off the databus, plus BBS7 (bankselect 7 is the IOpage) and theres your
address. The next part of the cycle is typically an IO_read with an
optional IO_write to follow. What makes doing IDE on PDP11 is the read
before write to the same address (plays havoc with device resgisters).
the simple solution is to map all reads to base+0 to base+n and writes
to base+N to Base+n+n so reads do not overlap writes.
So you'd advocate essentially using the Write signal (forgive me if I
don't
remember which name that one wore) as an address line, kind of like the
AMPRO folks did on their FDC?
Allison
3:47 a.m.
But have you tried to get an IDE drive to work with a
Qbus? That would seem
No, not yet. The probelm there is more one of drivers than interface.
At the interface level, it's ben done by a russian chap. The problem with
that design is it's PIO driven and as a result slower than putty. A good
IDE interface wtih DMA has the potential to rival some of the faster
interfaces used on Qbus as it would not have all the protocal overhead.
I have adaquate devices for my PDP-11s but I'd love to build a small
T-11 powered box. Two limitations to doing that, an extreme shortage of
time and a free OS I can write a driver for (RT11 and unix don't count).
of the 3rd party manufactures did not do an MSCP
emulation. And if you are
still concerned about using MSCP until or before the patent expires, there are
Silly question, why bother with MSCP at all? A native drive like RL02,
DSD-880, RKnn would work just as well and offer better performance.
Answer, it's a standard that everyone seems to emulate.
Besides CP/M based systems are where I hardware hack the most.
Allison
17 Apr
17 Apr
8:21 p.m.
Well, I've got one drive, an ST351, I think it is, that responds to the
8-bit conditioning. It's a 1.25"-tall 3-1/2" drive, of 44 MB capacity.
That's not what I want for the "hard-card" on the S-100.
All your assertions about the relative folly of expending effort/resources
to make the drive talk-8-bits wide has no real purpose except to bind it to
its historical roots. I'd bet that the logic in the 1003-WAH uses the
data-width bit to enable the IOCS16- signal and thereby lets the AT bus
control whether the transfer is 8 or 16 bits wide, since the AT bus is
required to do that.
The early IDE drives made for COMPAQ by Seagate, were half-high 5-1/4"
drives from which I've saved one or two of the PCB's. These have all the
same logic on them that I recognize from the 1003-WA2 which is a WAH with
the FDC missing, I think. Maybe it's the other way around . . .
Those guys undoubtedly supported the 8-bit mode, since their host was, in
some cases, an XT-class machine.
Dick
----- Original Message -----
From: allisonp <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, April 17, 2000 8:22 PM
Subject: Re: 8-bit IDE
busses.
replace
> "Has anyone got sufficient experience
with the IDE in
non-PC-compatible
applications to
say, unequivocally, whether the 8-bit operating mode
described in section 6 of the standard for the ATA (IDE) interface, yes
SPELLED OUT, actually exists in drives of the pre-1996 vintage?" It was
dropped from the standard in 1996. There seem to be many folks with
suggestions about how to implement this extremely elementary interface.
There seem to be few who know whether the standard was every full
implemented.
Why would they include that? NONE of the PC hardware they were intended
for wants to run as 8bit bus. Even the crippled SL/SLC run as 16bit Allison seems to be the only one who's tried
this, and, I fear, it may be
in
a PC-compatible, where all bets are off as to what
really happens.
Obviously visually impaired! I don't hack IO in PCs nor have I tried that
yet.
I may add why even bother, IDE works there as is.
ALL of the IDE work I've done was with 8085, z80s either stand alone
(bus less) or S100 Z80s. Further I'm currently working on a Z280 system
with IDE (Zbus 16bit). This is where I need interfaces and so I can older MFM or non-existent hard disks. I currently
have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
Connor 2022
Connor 3044
WD2120
WD2420
St3096A
St3144A
St3250A
St3660
Fijitsu 528mb
Quantum LPS 80 and 120
Maxtor 124mb
and afew other sub 60mb WD, NEC, Seagate drives.
I now have two 2.5mb IDE in the 700-800mb region I may try one day.
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Allison
>
>Dick
>
>----- Original Message -----
>From: Mike Ford <mikeford(a)socal.rr.com>
>To: <classiccmp(a)classiccmp.org>
>Sent: Monday, April 17, 2000 7:22 PM
>Subject: Re: 8-bit IDE
>
>
>> I don't know a hoot about this, but I wonder if taking a look at the
>> Sequential Systems Focus card for the Apple II might be instuctive. Its
a
>> controller and notebook IDE drive that all
fits on a Apple II slot
card.
8:58 p.m.
I know someone has made an ide hard card for the 8 bit apple 2 bus that takes
modern IDE drives, including the 1 inch notebook drives. I'm unclear how it
was done, but you can buy one at www.allelec.com/harddiap.htm .
Well, I've got one drive, an ST351, I think it is, that responds to the
8-bit conditioning. It's a 1.25"-tall 3-1/2" drive, of 44 MB capacity.
That's not what I want for the "hard-card" on the S-100.
All your assertions about the relative folly of expending effort/resources
to make the drive talk-8-bits wide has no real purpose except to bind it to
its historical roots. I'd bet that the logic in the 1003-WAH uses the
data-width bit to enable the IOCS16- signal and thereby lets the AT bus
control whether the transfer is 8 or 16 bits wide, since the AT bus is
required to do that.
The early IDE drives made for COMPAQ by Seagate, were half-high 5-1/4"
drives from which I've saved one or two of the PCB's. These have all the
same logic on them that I recognize from the 1003-WA2 which is a WAH with
the FDC missing, I think. Maybe it's the other way around . . .
Those guys undoubtedly supported the 8-bit mode, since their host was, in
some cases, an XT-class machine.
Dick
----- Original Message -----
From: allisonp <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, April 17, 2000 8:22 PM
Subject: Re: 8-bit IDE
busses.
replace
--
Jim Strickland
jim(a)DIESPAMMERSCUMcalico.litterbox.com
-----------------------------------------------------------------------
BeOS Powered!
-----------------------------------------------------------------------
> "Has anyone got sufficient
experience with the IDE in
non-PC-compatible
applications to say, unequivocally, whether the 8-bit operating mode
described in section 6 of the standard for the ATA (IDE) interface, yes
SPELLED OUT, actually exists in drives of the pre-1996 vintage?" It was
dropped from the standard in 1996. There seem to be many folks with
suggestions about how to implement this extremely elementary interface.
There seem to be few who know whether the standard was every full
implemented.
Why would they include that? NONE of the PC hardware they were intended
for wants to run as 8bit bus. Even the crippled SL/SLC run as 16bit Allison seems to be the only one who's tried
this, and, I fear, it may be
in
a PC-compatible, where all bets are off as to what
really happens.
Obviously visually impaired! I don't hack IO in PCs nor have I tried that
yet.
I may add why even bother, IDE works there as is.
ALL of the IDE work I've done was with 8085, z80s either stand alone
(bus less) or S100 Z80s. Further I'm currently working on a Z280 system
with IDE (Zbus 16bit). This is where I need interfaces and so I can older MFM or non-existent hard disks. I
currently have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
Connor 2022
Connor 3044
WD2120
WD2420
St3096A
St3144A
St3250A
St3660
Fijitsu 528mb
Quantum LPS 80 and 120
Maxtor 124mb
and afew other sub 60mb WD, NEC, Seagate drives.
I now have two 2.5mb IDE in the 700-800mb region I may try one day.
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Allison
>
>Dick
>
>----- Original Message -----
>From: Mike Ford <mikeford(a)socal.rr.com>
>To: <classiccmp(a)classiccmp.org>
>Sent: Monday, April 17, 2000 7:22 PM
>Subject: Re: 8-bit IDE
>
>
>> I don't know a hoot about this, but I wonder if taking a look at the
>> Sequential Systems Focus card for the Apple II might be instuctive. Its
a
>> controller and notebook IDE drive that
all fits on a Apple II slot
card.
9:14 p.m.
There's no real difficulty involved in making the interface work. The one
thing I get out of doing this sort of thing myself is that I get to decide
how it SHOULD be done and the do it that way if possible. It's not
difficult with today's technology to build a simple IDE interface that
handles the bytes and words correctly.
However, it seems to me that, if there is supposed to be a means built into
the "standard" interface by means of which I can avoid having to deal with
the conversion to/from words on the data port, I should exploit that
feature. Finding the drives that support that feature is then the crux, but
it should be achievable. In no case will it be of any significant economic
value.
Unfortunately, there seems to be precious little information around about
whether this feature is available, and if so, on which drive models.
Dick
----- Original Message -----
From: Jim Strickland <jim(a)calico.litterbox.com>
To: <classiccmp(a)classiccmp.org>
Sent: Monday, April 17, 2000 11:58 PM
Subject: Re: 8-bit IDE
I know someone has made an ide hard card for the 8 bit
apple 2 bus that
takes
modern IDE drives, including the 1 inch notebook
drives. I'm unclear how
it
was done, but you can buy one at
www.allelec.com/harddiap.htm .
>
> Well, I've got one drive, an ST351, I think it is, that responds to the
> 8-bit conditioning. It's a 1.25"-tall 3-1/2" drive, of 44 MB
capacity.
> That's not what I want for the "hard-card" on the S-100.
>
> All your assertions about the relative folly of expending
effort/resources
> to make the drive talk-8-bits wide has no real
purpose except to bind it
to
> its historical roots. I'd bet that the logic
in the 1003-WAH uses the
> data-width bit to enable the IOCS16- signal and thereby lets the AT bus
> control whether the transfer is 8 or 16 bits wide, since the AT bus is
> required to do that.
>
> The early IDE drives made for COMPAQ by Seagate, were half-high 5-1/4"
> drives from which I've saved one or two of the PCB's. These have all
the
> same logic on them that I recognize from the
1003-WA2 which is a WAH
with
> the FDC missing, I think. Maybe it's the
other way around . . .
>
> Those guys undoubtedly supported the 8-bit mode, since their host was,
in
> some cases, an XT-class machine.
>
> Dick
>
> ----- Original Message -----
> From: allisonp <allisonp(a)world.std.com>
> To: <classiccmp(a)classiccmp.org>
> Sent: Monday, April 17, 2000 8:22 PM
> Subject: Re: 8-bit IDE
>
>
> > > "Has anyone got sufficient experience with the IDE in
> non-PC-compatible
> > >applications to say, unequivocally, whether the 8-bit operating mode
> > >described in section 6 of the standard for the ATA (IDE) interface,
yes
> > >SPELLED OUT, actually exists in drives
of the pre-1996 vintage?" It
was
> > >dropped from the standard in 1996.
There seem to be many folks with
> > >suggestions about how to implement this extremely elementary
interface.
> > >There seem to be few who know whether
the standard was every full
> > >implemented.
> >
> >
> > Why would they include that? NONE of the PC hardware they were
intended
> > for wants to run as 8bit bus. Even the
crippled SL/SLC run as 16bit
> busses.
> >
> > >Allison seems to be the only one who's tried this, and, I fear, it
may be
> > in
> > >a PC-compatible, where all bets are off as to what really happens.
> >
> >
> > Obviously visually impaired! I don't hack IO in PCs nor have I tried
that
> > yet.
> > I may add why even bother, IDE works there as is.
> >
> > ALL of the IDE work I've done was with 8085, z80s either stand alone
> > (bus less) or S100 Z80s. Further I'm currently working on a Z280
system
> > with IDE (Zbus 16bit). This is where I need
interfaces and so I can
> replace
> > older MFM or non-existent hard disks. I currently have one S100
system
> > running a connor3044A (40mb) IDE that will
be upgraded to a ST3250
> > (250mb) as it's a better drive.
> >
> > The drives I've tried include:
> >
> > Connor 2022
> > Connor 3044
> > WD2120
> > WD2420
> > St3096A
> > St3144A
> > St3250A
> > St3660
> > Fijitsu 528mb
> > Quantum LPS 80 and 120
> > Maxtor 124mb
> > and afew other sub 60mb WD, NEC, Seagate drives.
> >
> > I now have two 2.5mb IDE in the 700-800mb region I may try one day.
> >
> > I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
> >
> > I think this is a good cross section
> >
> > Allison
> > >
> > >Dick
> > >
> > >----- Original Message -----
> > >From: Mike Ford <mikeford(a)socal.rr.com>
> > >To: <classiccmp(a)classiccmp.org>
> > >Sent: Monday, April 17, 2000 7:22 PM
> > >Subject: Re: 8-bit IDE
> > >
> > >
> > >> I don't know a hoot about this, but I wonder if taking a look at
the
> > >> Sequential Systems Focus card for
the Apple II might be instuctive.
Its
a
--
Jim Strickland
jim(a)DIESPAMMERSCUMcalico.litterbox.com
-----------------------------------------------------------------------
BeOS Powered!
----------------------------------------------------------------------- >> controller and notebook IDE drive that
all fits on a Apple II slot
card.
>
>
10:29 p.m.
However, it seems to me that, if there is supposed to
be a means built into
the "standard" interface by means of which I can avoid having to deal with
the conversion to/from words on the data port, I should exploit that
feature. Finding the drives that support that feature is then the crux, but
it should be achievable. In no case will it be of any significant economic
value.
By day you're the main system architect at MicroSoft right? ;)
18 Apr
18 Apr
5:46 a.m.
No, I'm not, but I wish I had 10% of his weekly take . . .
Dick
----- Original Message -----
From: Mike Ford <mikeford(a)socal.rr.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 18, 2000 1:29 AM
Subject: Re: 8-bit IDE
>However, it seems to me that, if there is supposed
to be a means built
into
>the "standard" interface by means of
which I can avoid having to deal
with
>the conversion to/from words on the data port, I
should exploit that
>feature. Finding the drives that support that feature is then the crux,
but
>it should be achievable. In no case will it be of
any significant
economic
value.
By day you're the main system architect at MicroSoft right? ;)
4:36 a.m.
On Mon, 17 Apr 2000, Jim Strickland wrote:
I know someone has made an ide hard card for the 8 bit
apple 2 bus that takes
modern IDE drives, including the 1 inch notebook drives. I'm unclear how it
was done, but you can buy one at www.allelec.com/harddiap.htm .
It maps the high Byte to low byte like most do. There si also a coco-ide
mod that works the same way. the drawings for the COCO ide is suggest
a simple design as well but expects the IDE to be 16bits for data as do
most otehrs I've seen.
Allison
4:27 a.m.
On Mon, 17 Apr 2000, Richard Erlacher wrote:
Well, I've got one drive, an ST351, I think it is,
that responds to the
8-bit conditioning. It's a 1.25"-tall 3-1/2" drive, of 44 MB capacity.
That's not what I want for the "hard-card" on the S-100.
I have two of them. Never thought to try them as they are slow and very
old.
All your assertions about the relative folly of
expending effort/resources
to make the drive talk-8-bits wide has no real purpose except to bind it to
its historical roots. I'd bet that the logic in the 1003-WAH uses the
data-width bit to enable the IOCS16- signal and thereby lets the AT bus
control whether the transfer is 8 or 16 bits wide, since the AT bus is
required to do that.
Flog thyself if you care to. ;) Oh, the 1003WAH DOES NOT respond to
set_16bits. Before I ahd IDE drives to experiment with I tried it.
it still lost every other byte, just didn't generate the signal(IOCS16).
The 1003 was never useable in ISA-8 (xt bus).
What I did do instead was to use a WD1002WXA to make that interface
as it was already ISA-8 and easy to interface as a 8-bit controller.
Wrote an article for TCJ about it.
Allison
6:07 a.m.
plz see comments below.
Dick
----- Original Message -----
From: <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 18, 2000 7:27 AM
Subject: Re: 8-bit IDE
On Mon, 17 Apr 2000, Richard Erlacher wrote:
... and they're of about the right capacity for an old-style CP/M setup, if
you want to be true to the period. They make the typical
parallel-port-driven SMD or other setup typical of the 1976-1981 timeframe
look pretty bad, though.
Well, I've got one drive, an ST351, I think
it is, that responds to the
8-bit conditioning. It's a 1.25"-tall 3-1/2" drive, of 44 MB capacity.
That's not what I want for the "hard-card" on the S-100.
I have two of them. Never thought to try them as they are slow and very
old.
> All your assertions about the relative folly of expending
effort/resources
<snip>
The 1003 was never useable in ISA-8 (xt bus).
That's understandable since the IOCS16- is on the C-D connector. The
missing bit, BTW, is probably not on the 'WAH board, and it's SET for 8-bit
mode in the IDE standard. If you only use A-B, you won't see it and, by the
way, neither will the ISA bus, hence its standard (only because almost
everyone faithfully copied IBM) logic for steering the bytes in that
signal's absense, would try to shuffle the outgoing bytes into an 8-bit
window on the A-B connector only. On reads, you'd be in trouble, since half
the bytes would be missing. The ISA bus' response to IOCS16- is to enable
the high-order bus buffers on the little C-D connector.
What I did do instead was to use a WD1002WXA to make that interface
as it was already ISA-8 and easy to interface as a 8-bit controller.
Wrote an article for TCJ about it.
I have one old data book from WD that has some of the XT-type controllers in
it, but lacks that one. What are the characteristics of this one?
Allison
7:40 a.m.
... and they're of about the right capacity for an
old-style CP/M setup, if
you want to be true to the period. They make the typical
parallel-port-driven SMD or other setup typical of the 1976-1981 timeframe
look pretty bad, though.
I do have one of those 8inch monsters, runs too. However you seem to
assume that many of the CP/M systems I have are required to be of "period"
design. FOO on that. I have many that are simply restored and they stay
that way. However I have several that are upgraded where possible to get
more performance for actual practical use. Those systems run faster Z80s
or have SCSI, big IDE (by CPM standards) and the like.
I have one old data book from WD that has some of the
XT-type controllers in
it, but lacks that one. What are the characteristics of this one?
So do I, but I built a useable controller from it without that. Actually
there are three versions of the 1002 for the Xt bus, all look different
but operate the same due to the base chip, the 1100 series. I used some
PC source software , plus PC programming books to figure out the
registers. It was a very black box approach but I got a working MFM
subsystem that works well. The end result looks like a 1002-HDO in many
respects.
Allison
11:06 a.m.
NO! It's not a requirement at all, but I do find it aesthically preferable
to use same-generation hardware together. That's just a personal
preference, however. If you want to run your IMSAI with an ESDI interface,
and use a flat-panel display for the console and talk to other systems on
the MBEthernet, it's entirely up to your. It is, after all, YOUR computer,
not mine. I find it strange somehow, to see a living room furnished in
'70's Spanish style with a '50's danish-modern coffee table and a
strictly
art-deco set of lamps, vases, and pictures, but if one can do it, one can do
it tastefully.
I have a selection of 8" drives and controllers, and several of the early
5-1/4" full-height models as well. A few years back, I gave away my 20 MB
Century Data Marksman. which was one of those 14" drives with a parallel
port interface on it (driven by a couple of ports in a 6821 PIA.)
I made an S-100 board for a co-worker to use in combination with one of the
1/3-card WD HDC's from the XT generation, together with a 1/3-card
Hercules-style graphics/monochrome text board. Using what he knew about the
keyboard, he built an interface to an XT-class keyboard and hooked his
monitor up to the mono card, thereby giving himself access to quite a bit of
hardware with very little hassle. The driver code (maybe not the best of
it, though) is in the IBM tech ref manual. The HDC is quite
straightforward. The XT-keyboard interface was a piece of programmable
logic, possibly a MOT MCU.
Dick
----- Original Message -----
From: <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 18, 2000 10:40 AM
Subject: Re: 8-bit IDE
controllers in
> ... and they're of about the right capacity
for an old-style CP/M setup,
if
> you want to be true to the period. They make the
typical
> parallel-port-driven SMD or other setup typical of the 1976-1981
timeframe
look pretty
bad, though.
I do have one of those 8-inch monsters, runs too. However you seem to
assume that many of the CP/M systems I have are required to be of "period"
design. FOO on that. I have many that are simply restored and they stay
that way. However I have several that are upgraded where possible to get
more performance for actual practical use. Those systems run faster Z80s
or have SCSI, big IDE (by CPM standards) and the like.
> I have one old data book from WD that has some of the XT-type it, but lacks
that one. What are the characteristics of this one?
So do I, but I built a useable controller from it without that. Actually
there are three versions of the 1002 for the Xt bus, all look different
but operate the same due to the base chip, the 1100 series. I used some
PC source software , plus PC programming books to figure out the
registers. It was a very black box approach but I got a working MFM
subsystem that works well. The end result looks like a 1002-HDO in many
respects.
Allison
11:34 a.m.
preference, however. If you want to run your IMSAI
with an ESDI interface,
If I had an IMSAI it would be restored. Hwoever I have two NS* Horizons
one virgin, teh other my 1978 unit hacked six ways from sunday mostly
during the winter of 1980, that in itself if art too.
the MBEthernet, it's entirely up to your. It is,
after all, YOUR computer,
not mine. I find it strange somehow, to see a living room furnished in
'70's Spanish style with a '50's danish-modern coffee table and a
strictly
To me a compupro crate is not anything but an empty room and has no style
other than good construction rules. What I choose to put in it should
have order "fung shui" as it were but that doesn't rule mixing old and new
if done sanely.
I can do that as I have multiples of many of the machines that are
actively hacked. Nothing is lost, knowledge is both preserved and
added to.
Allison
12:22 p.m.
I've never seen a CompuPro ANYTHING I liked. However, there's no accounting
for taste.
You mustn't forget, BTW, that I'm not a collector, except in the sense that
I haven't thrown some things away because I remember what they cost me.
The S-100 stuff was pretty easy to deal with back when you could make a
living with that sort of thing. Though I found the INTEL boards for the
multibus-1 somewhat better designed, they were so much more costly, I
couldn't, in good conscience, recommend them if an S-100 board, otherwise
equivalent, cost 10% as much, as was often the case.
Of course, I have a real bunch of Multibus-1 stuff because it was always up
to the mark.
Dick
----- Original Message -----
From: <allisonp(a)world.std.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 18, 2000 2:34 PM
Subject: Re: 8-bit IDE
> preference, however. If you want to run your IMSAI with an ESDI
interface,
If I had an IMSAI it would be restored. Hwoever I have two NS* Horizons
one virgin, teh other my 1978 unit hacked six ways from sunday mostly
during the winter of 1980, that in itself if art too.
> the MBEthernet, it's entirely up to you. It is, after all, YOUR
computer,
> not mine. I find it strange somehow, to see a
living room furnished in
> '70's Spanish style with a '50's danish-modern coffee table and a
strictly
To me a compupro crate is not anything but an empty room and has no style
other than good construction rules. What I choose to put in it should
have order "fung shui" as it were but that doesn't rule mixing old and new
if done sanely.
I can do that as I have multiples of many of the machines that are
actively hacked. Nothing is lost, knowledge is both preserved and
added to.
Allison
17 Apr
17 Apr
9:49 p.m.
On 17 Apr 2000, at 22:22, allisonp wrote:
Whoa, good cross section, and missed the ST351A/X? This is one
of rare drives that can be jumpered for either physical 16bit or 8bit
mode.
Allison seems
to be the only one who's tried this, and, I fear, it may be
in
a PC-compatible, where all bets are off as to what
really happens.
Obviously visually impaired! I don't hack IO in PCs nor have I tried that
yet.
I may add why even bother, IDE works there as is.
ALL of the IDE work I've done was with 8085, z80s either stand alone
(bus less) or S100 Z80s. Further I'm currently working on a Z280 system
with IDE (Zbus 16bit). This is where I need interfaces and so I can replace
older MFM or non-existent hard disks. I currently have one S100 system
running a connor3044A (40mb) IDE that will be upgraded to a ST3250
(250mb) as it's a better drive.
The drives I've tried include:
I do have two WD PS/2m30 compatable 8bit IDE 20meg drives.
I think this is a good cross section
Allison
>
>Dick
Wizard
18 Apr
18 Apr
4:44 a.m.
I think this
is a good cross section
Whoa, good cross section, and missed the ST351A/X? This is one
of rare drives that can be jumpered for either physical 16bit or 8bit
mode.
8862
days inactive
8863
days old
20 comments
6 participants
participants (6)
-
allisonp＠world.std.com -
edick＠idcomm.com -
jhfine＠idirect.com -
jim＠calico.litterbox.com -
jpero＠cgocable.net -
mikeford＠socal.rr.com