7 Aug
2011
7 Aug
'11
7:57 p.m.
replying to ard at p850ug1.demon.co.uk (Tony Duell)
> Nice! There's been a resurgence in
appreciation of the "classics".
The 5160 I just mentioned with 4 floppy drive
How did you accomplish that?
- a controller that honors all 4 drive selects, not just 2
- primary & secondary disk controller
- Compaticard or other disk controller
at "non-standard" address & IRQ?
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
Not 'till SATA have we been able to un-do that mistake :-(
2 hard drives
Awww, come on, you can do better than that :-)
I have a 80386 DOS system running MFM, ATAPI/IDE and SCSI.
The "4drives" shareware was the first DOS support I found
for the hard drive controller at alternate addresses & IRQ.
MDA and CGA cards,
Dual display? What programs support that?
4 serial ports, 24 line user I/O (8255-based), etc.
Not all original IBM of course
Ah yes, that reminds me of the excitement of reading
Computer Shopper and all the computer shows,
always with new upgrades: higher res graphics, more memory, etc.
Every month or so, ALL the vendors had a different brand
of add-on card. For a while it was a box with a king playing card
on the cover. Then Pine Technologies. Then something else.
For those of us USING the stuff at the time,
we were always upgrading it piece by piece.
That's why it's so hard to find one with all original parts.
A PDP8/e with 32K words of real core
Do the boards have clear plastic for you to see the cores?
That's such a treat: true non-volatile memory!
> My high school started everyone with programming
BASIC
> on the HP 9820A:
I find that hard to believe for one good reason.
The HP9820 never ran BASIC...
You got me there.
It was kinda almost similar to BASIC.
It's so close that I totally forgot
when I converted to BASIC.
Since BASIC has so many "dialects",
I guess that fuzzed my memory a little.
Thanks for the clarification.
The HP9820 was HP's first infix-notation
(as opposed to RPN/postfix) calculator
and used a language that developed into the 'HPL' of the 9825 etc.
It was such a pleasure using an instant response interactive machine!
No waiting for carriage-return for the interpreter to handle the line
or waiting for the compiler for an all-or-nothing result.
You may have guessed I have something of a soft spot
for these machines.
As an undergrad, I never had the budget for a HP calculator
nor access to a HP desktop like my high school's.
Somehow a TI-55 sufficed for me as an EE undergrad.
Looking back, I just can't see how. It's *SO LAME*.
I got a HP-28C when it was new and it's still my primary calculator.
I really need that 4 line display to use RPN so I can see the stack.
-- jeffj
8 Aug
8 Aug
2:21 a.m.
On Sun, 7 Aug 2011, Jeff Jonas wrote:
WHICH INCLUDED THE STOCK
IBM 5150 FDC.
ANYTHING that you write.
(debugging mode in XenoCopy often made use of alternate display, and some
outputs to a P.O.S.T. card, but MOST of that code was compiled out on
release versions. We once tried a release where the HELP screens were
ALWAYS activated on alternate display, but the users that NEEDED help
wanted handholding instead.)
Early Windoze development software supported two videos.
I think that some early (pre-megapixel) CAD used both IBM displays.
The 5160 I
just mentioned with 4 floppy drive
How did you accomplish that?
- a controller that honors all 4 drive selects, not just 2 - primary & secondary disk controller
- Compaticard or other disk controller
at "non-standard" address & IRQ?
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
NO.
You are thinking of the AT (5170)
The 5150 AND the 5160 used a controller card that directly supported 4
floppy drives. (and NO hard disks. THAT was on another card). 2 of the
floppies connected to a 34 pin card edge on the FDC board. The other 2
connected to a DC37 connector hanging out the back of the case.
With the 5160, there was the addition of a hard^H^H^H^H FIXED disk
controller. IBM would not call it a hard disk, because that had negative
connotations, particularly with "ease of use". They thought that
"FIXED"
didn't have negative connotations (except in a veterinary context?) They
also called the main circuit board "system" board, rather than
"motherboard", s'posedly due to the use of "mother" as a prefix in
widely
publicized speeches by Black Panthers at Merritt College.
The 5160 HD controller (made by Xebec) handled 2 st412 style MFM drives.
It was set for 10M drives. That could be changed to 5, 16? and 24?
through soldering jumpers in a block in the middle of the board.
Further changes could easily be accomplished through editing the ROM, or
tellinf the OS to ignore the ROM values in software.
What YOU were thinking of was the 5170 controller (WD?) that handled 2
ST412 style MFM drives (one 34pin (2x17) berg and one 20?) and 2 SA450
style floppies (one 34 pin berg).
MDA and CGA
cards,
Dual display? What programs support that? 
3:25 p.m.
didn't have negative connotations (except in a
veterinary context?) They
also called the main circuit board "system" board, rather than
"motherboard", s'posedly due to the use of "mother" as a prefix
in widely
I've also seen it called the 'planar board'. I am pretty sure the
backplane in the expansio box is labeleld 'I/O Planar' or something very
similar.
What YOU were thinking of was the 5170 controller (WD?) that handled 2
ST412 style MFM drives (one 34pin (2x17) berg and one 20?) and 2 SA450
That's 1 34 pin (cotnrol cable) and _2_ 20 pin (data cabels from each of
2 drives).
Alas the disk cotnroller is one thing that's been changed in the 5170 I
am typing this on. I had to use a hacked IDE interface card to get a
reasonalbe size hard disk.
-tony
9 Aug
9 Aug
5:55 p.m.
On 08/08/2011 02:21 AM, Fred Cisin wrote:
>> MDA and CGA cards,
> Dual display? What programs support that?
...
> I think that some early (pre-megapixel) CAD used
both IBM displays.
In fact, AutoCAD supported this, with the command line interface on
the MDA and everything else on the graphics-capable card.
AutoCAD's command line interface, with all of those original (and
indispensable) commands is still there in both the Wintoy and UNIX
versions, but one can no longer redirect the command line I/O to a
different display, for obvious reasons.
-Dave
--
Dave McGuire
Port Charlotte, FL
8 Aug
8 Aug
12:22 p.m.
On 7 Aug 2011 at 19:57, Jeff Jonas wrote:
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
Not 'till SATA have we been able to un-do that mistake :-(
The 5150 and 5160 both supported 4 floppy drives with the original
controller--what do you think the DC37 on the bracket of the
controller board was for? As both machines shipped with full-height
5.25" drives, there wasn't room for 4 drives in the primary box, so
the added ones were external. I know of lots of 8-bit CP/M boxes
with the same arrangement.
The cable "twist" was a very clever (but not original) solution to a
thorny problem of how one controlled the spindle motors on two drives
when there is only one motor control line. The fact that one could
also set all floppies to the second drive select was a bonus.
I have Ultrastor controllers that use two unassigned lines in the
floppy cable to support a third drive on the cable. And a few DTC
controllers could be jumpered to allow for a 4-drive "flat" cable at
the expense of having all drive motors come on at the same time.
The "4drives" shareware was the first DOS
support I found
for the hard drive controller at alternate addresses & IRQ.
You've led a sheltered life. All of my disk products allowed for
configuration of port, IRQ and DMA channel.
Dual display? What programs support that?
The do exist, in the professional software market.
Do the boards have clear plastic for you to see the
cores?
That's such a treat: true non-volatile memory!
True non-volatile memory is punched cards--until the termites and
mice get to them.
TI is currently shipping its MSP430 uC in a version that uses FRAM
for memory. Since it's a von Neumann (unified address space),
provessor, you can execute code out of FRAM. There's also MRAM, but
I don't know if that's being incorporated into any uC designs.
As an undergrad, I never had the budget for a HP
calculator
nor access to a HP desktop like my high school's.
Somehow a TI-55 sufficed for me as an EE undergrad.
Looking back, I just can't see how. It's *SO LAME*.
As undergrads, many of us had access to a slide rule and the "rubber"
book.
--Chuck
3:35 p.m.
The cable "twist" was a very clever (but not
original) solution to a
thorny problem of how one controlled the spindle motors on two drives
when there is only one motor control line. The fact that one could
also set all floppies to the second drive select was a bonus.
I always liked the way thatHP did it in the 9836 machines. Those have 2
5.25" floppy drives, permanenytly enabled (MX jumper fitted) amd linked
by separate ribbon cables to the controller board. The latter is fitted
acrsso the case behind the flopy drives which means the ribbon cable
layout is quite tidy.
This arrangement means that not only can the motors be controlled
separately, but also the write protect lines can be monitored in
hardware. Since chanigny a disk will always result in some transitions on
the QP line, this is used to provide a 'disk changed' signal.
You've led a sheltered life. All of my disk
products allowed for
configuration of port, IRQ and DMA channel.
I am pretty use I used 22disk and/or anadisk with the secondard
controller in thae 5160 under discussion (this controller goes to a pair
of 8" drives) before I started writing my own stuff.
As undergrads, many of us had access to a slide rule
and the "rubber"
book.
FWIW, I still do. I would not want to be without a Rubber Bible actually,
there's a lot of very useful information in there. OK, mine is now nearly
40 years old, but while the physical constants are now known more
accurately, the differences are unlikely to affect anything I am doing :-)
As for slide rules, I have several and know how to use them. I was lucky
enough to find an Otis King (helical slide rule).
-tony
1:40 p.m.
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
Who ever used more than 2 floppies on a PC anyway? Or more than 2 hard
disks (back then, before PC based servers)?
--
Will
2:06 p.m.
On Mon, 8 Aug 2011, William Donzelli wrote:
Who ever used more than 2 floppies on a PC anyway? Or
more than 2 hard
disks (back then, before PC based servers)?
2 360K 5.25"
720K (96tpi) 5.25" ("quad")
micropolis 100tpi
3"
3.25"
2 720K 3.5", REPLACED with 2 1.4M 3.5"
2 1.2M 5.25"
Epson 67.5 TPI 3.5"
8" (Flagstaff, Vista, Maynard, etc.)
2.8M (I used MicroSoulutions "Backpack" parallel port)
floptical (SCSI)
LS120
ZIP
(and also a few obscure others)
MOST people could combine the 720K and the 1.2M 5.25" (such as Teac 55GF),
but sometimes it just made more sense to have the individual models. And
SOME people didn't need all of those types.
Using MORE THAN four drives was awkward (cabling switches, etc.), so
"multitasking" meant turning on numerous machines as part of the normal
"boot" process. LOTS of power strips.
I had most of the prototype external cases and power supplies from
Elcompco.
Stock power supply was not happy trying to power 2 4096s, so some machines
had extra power supplies.
2:22 p.m.
2 360K 5.25"
720K (96tpi) 5.25" ("quad")
micropolis 100tpi
3"
3.25"
2 720K 3.5", REPLACED with 2 1.4M 3.5"
2 1.2M 5.25"
Epson 67.5 TPI 3.5"
8" ?(Flagstaff, Vista, Maynard, etc.)
2.8M ? (I used MicroSoulutions "Backpack" parallel port)
floptical ?(SCSI)
LS120
ZIP
(and also a few obscure others)
During the 5150/5160 era? The time frame the PC design team was planning for?
Think realistically, please. Not a retrocomputing wet dream.
Even in the 386 era, BY FAR most machines had only two floppies and
one hard disk.
--
Will
3:53 p.m.
> 2 360K 5.25"
> 720K (96tpi) 5.25" ("quad")
> micropolis 100tpi
> 3"
> 3.25"
> 2 720K 3.5", REPLACED with 2 1.4M 3.5"
> 2 1.2M 5.25"
> Epson 67.5 TPI 3.5"
>
> 8" ?(Flagstaff, Vista, Maynard, etc.)
> 2.8M ? (I used MicroSoulutions "Backpack" parallel port)
> floptical ?(SCSI)
> LS120
> ZIP
> (and also a few obscure others)
On Mon, 8 Aug 2011, William Donzelli wrote:
During the 5150/5160 era? The time frame the PC design
team was planning for?
YES.
ALL of those, with the exception of the 1.2M and the 1.4M, MOST before the
5160 came out, I used on my 5150, which I ordered from IBM in 1981 with
16K of RAM. In a separate order at the same time, I bought an FDC card, a
serial card, and a CGA card. My 5150 has been in my shed for ten years
since I closed my office. When I put it into storage, If I remember
correctly, it had 2 360s (SA455, one of which had write-protect jumpering
reversed to write notchless disks), a 96tpi 5.25" (SA465? or Teac 55F),
and an external 8", with CGA, AST multifunction, and aftermarket
(Maynard?) FDC.
XenoCopy was designed for 5150 and 5160. Although, in general, it would
work with 5170, etc., in the documentation, I explicitly disclaimed any
promise of working on bizarre newer stuff. It was for a specific purpose,
and if your new fancy machine couldn't handle it, then do THAT task on a
REAL 5150 or 5160. In fact, my publisher (may they ROT IN PIECES) sent
out prerelease copies for review that still had in them explicit tests to
prevent them from running on anything other than an IBM 5150/5160.
BTW, I did the documentation for XenoCopy on a Corona Data Systems/Cordat
laser printer controller, whose software REFUSES to work on a 286.
Think realistically, please.
Make THAT request
of somebody who is more prone to do so.
Unreasonable?? Damn straight.
The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. - George Bernard Shaw
I asked Gary Kildall to designate an arbitrary set of parameters
as the "standard 5.25" format for CP/M. He responded, the standard format
for CP/M is 8" SSSD. I repect[ed] his refusal, but IF he had done so,
MS-DOS would not have been the platform for XenoCopy.
Not a retrocomputing wet dream.
Even in the 386 era,
Ya mean when PCs weren't FUN anymore??
I got out of computers in 1970 during the aerospace collapse, with a self
promise to return when individual computers became feasable.
I was late getting into MICROcomputers (TRS80 4K level1), and never had
much motivation to keep up with the "progress".
What has "improved" in the last 10 or 20 years? Sure, they've gotten
faster, but MICROS~1 PLANS on Moore's law, and the software takes away any
performance gains of the faster hardware.
Other than the higher data transfer rates needed for HD (and ED), the 5150
was better for floppy use than anything that came later. The 5150/5160
tech ref had full source code for Int 13h. When the 5170 came out, its
techref did not have comments on the code for density/DTR change! As a
beginning programmer (which I still am), I had a helluva struggle figuring
out how to over-ride 250K/300K/500K DTRs.
BY FAR most machines had only two floppies and
one hard disk.
Staring with 286, having a reasonable number of floppies (more than 2) was
becoming increasingly awkward and even sometimes difficult. And many of
the HD/floppy controllers that purported to support 4 drives were
bordering on unusable.
--
Grumpy Ol' Fred cisin at xenosoft.com
4:45 p.m.
BTW, I did the documentation for XenoCopy on a Corona
Data Systems/Cordat
laser printer controller, whose software REFUSES to work on a 286.
Was that one of the cards that used the PC to create the page bitmap nd
set it driectly to the printer engine (probably a Canon CX-VDO)? I know
that printer fairly well, having created one to use on my PERQ (which has
a similar printer interface). Pity you can no longer get toner cartridges
for it :-(
Think realistically, please.
Make THAT
request of somebody who is more prone to do so.
Unreasonable?? Damn straight.
The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. - George Bernard Shaw
'Therefore all progress depends on the unreasonable man'
What has "improved" in the last 10 or 20
years? Sure, they've gotten
faster, but MICROS~1 PLANS on Moore's law, and the software takes away an=
y
performance gains of the faster hardware.
As I have said here before, I can't think of one development in the last
30 years which has actaully improved my life ,or made it easier to do the
things I want to do (I can think of many that have had the reverse effect
though).
-tony
5:31 p.m.
> BTW, I did the documentation for XenoCopy on a
Corona Data Systems/Cordat
> laser printer controller, whose software REFUSES to work on a 286.
On Mon, 8 Aug 2011, Tony Duell wrote:
Was that one of the cards that used the PC to create
the page bitmap nd
set it driectly to the printer engine (probably a Canon CX-VDO)? I know
that printer fairly well, having created one to use on my PERQ (which has
a similar printer interface).
That's the one.
I also cabled the same one to an Eiconscript board and a Jlaser
Pity you can no longer get toner cartridges
for it :-(
But, until the drum, etc. wear out,you just need some bulk toner and a
set of torx bits
The reasonable
man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. - George Bernard Shaw
'Therefore all progress depends on the unreasonable man'
As I have said here before, I can't think of one development in the last
30 years which has actaully improved my life ,or made it easier to do the
things I want to do (I can think of many that have had the reverse effect
though).
7 p.m.
[CX toner cartridges]
> But, until the drum, etc. wear out,you just need some bulk toner and a
> set of torx bits
On Tue, 9 Aug 2011, Tony Duell wrote:
I've dismantled said cartridges (and put them back
together), but I've
never refilled one. What toner do you use, and where do you find it?
I don't know. 20+ years ago, we bought toner intended for them.
The staff in the computer lab learned the hard way that PC and CX
cartridges were NOT the same, didn't fit the other machine, and the toner
was "of reversed polarity"
10 years ago, a "friend" took more than a dozen PC and CX empties that I
had that had never been refilled before, and traded them for one of each
cartridges that had been refilled about a dozen times by drilling holes
in them!
What should I do with the ones that I have left? I couldn't even find
anybody who wanted 4 CX machines (one of which was the Fedex model with
different specs)
7:25 p.m.
On 08/09/2011 07:00 PM, Fred Cisin wrote:
10 years ago, a "friend" took more than a
dozen PC and CX empties that I
had that had never been refilled before, and traded them for one of each
cartridges that had been refilled about a dozen times by drilling holes
in them!
I did that for a long time. A company I worked for did that as a
service. It was messy, but profitable. One has to be work hard to
prevent plastic chips from drilling from getting into the toner
reservoir; they will ruin a drum in short order.
What should I do with the ones that I have left? I
couldn't even find
anybody who wanted 4 CX machines (one of which was the Fedex model with
different specs)
I can't take 'em, but it's a shame to see them go. Hmm, come to
think of it, I'd probably give a home to a clean original HP LaserJet
Series I.
-Dave
--
Dave McGuire
Port Charlotte, FL
10 Aug
10 Aug
12:01 a.m.
> 10 years ago, a "friend" took more than
a dozen PC and CX empties that I
> had that had never been refilled before, and traded them for one of each
> cartridges that had been refilled about a dozen times by drilling holes
> in them!
On Tue, 9 Aug 2011, Dave McGuire wrote:
I did that for a long time. A company I worked for
did that as a
service. It was messy, but profitable. One has to be work hard to
prevent plastic chips from drilling from getting into the toner
reservoir; they will ruin a drum in short order.
But, those cartridges didn't NEED any holes! With a torx bit, you could
disassemle it PROPERLY, clean it PROPERLY, refill it, and restore it to
use.
Was it REALLY that much easier to poke holes in it, instead of opening it?
12:14 a.m.
On 08/10/2011 12:01 AM, Fred Cisin wrote:
You know, this is activating some very dusty neurons as this was the
late 1980s when we were doing this, but I don't recall there being any
other way to access the toner reservoir. Sure, there were torx screws
in there, but I don't remember any of them opening up the reservoir. If
there had been, we'd have done it; we refilled many hundreds of CX
cartridges.
Dunno. *shrug*
-Dave
--
Dave McGuire
Port Charlotte, FL
> 10 years
ago, a "friend" took more than a dozen PC and CX empties that I
> had that had never been refilled before, and traded them for one of each
> cartridges that had been refilled about a dozen times by drilling holes
> in them!
On Tue, 9 Aug 2011, Dave McGuire wrote:
I did that for a long time. A company I
worked for did that as a
service. It was messy, but profitable. One has to be work hard to
prevent plastic chips from drilling from getting into the toner
reservoir; they will ruin a drum in short order.
But, those cartridges didn't NEED any holes! With a torx bit, you could
disassemle it PROPERLY, clean it PROPERLY, refill it, and restore it to
use.
Was it REALLY that much easier to poke holes in it, instead of opening it?
2:12 p.m.
Was it REALLY
that much easier to poke holes in it, instead of opening it?
You know, this is activating some very dusty neurons as this was the
late 1980s when we were doing this, but I don't recall there being any
other way to access the toner reservoir. Sure, there were torx screws
in there, but I don't remember any of them opening up the reservoir. If
there had been, we'd have done it; we refilled many hundreds of CX
cartridges. From what I rememebr (and I may be mixing it up with
the SX cartridge),
you can get the outer cassing off with the torx screws and some
plastic
barbs, then the thing spits in half at the drum. One half is the waste
toner collector -- which you empty [1], the other is the toner reservoir
which you can somehow get inside and refill. If anyone is seriously
interested, I'll dig out a cartridge and take it apart.
[1] This was certainly an SX story... A friend and I took apart his SX
cartridge, and decided to empty the waste toner out of it. He did so into
a toilet. We then discovered that toner flats on water, and the darn
stuff would not flush away. I then suggested adding a little washing-up
detergent, This worked after a number of flushes, you get this odd black
foam in the process...
-tony
11 Aug
11 Aug
8:22 p.m.
On 08/10/2011 02:12 PM, Tony Duell wrote:
Logically, there has to be a way to fill the toner reservoir, since that
was done at the factory. The plastic wasn't moulded with the toner inside
it :-)
From what I rememebr (and I may be mixing it up with the SX cartridge),
you can get the outer cassing off with the torx screws and some plastic
barbs, then the thing spits in half at the drum. One half is the waste
toner collector -- which you empty [1], the other is the toner reservoir
which you can somehow get inside and refill. If anyone is seriously
interested, I'll dig out a cartridge and take it apart.
I don't disagree, but nevertheless, in those days, this was a pretty
big business. We did a LOT of them, and so did a lot of other
companies. Refill kits, complete with a drill bit and little plastic
plugs, were available commercially all over the place. This was a
common and widespread thing.
There must've been some compelling reason to do it this way. I'll be
damned if I can figure out what it was.
-Dave
--
Dave McGuire
Port Charlotte, FL
Was it REALLY that much easier to poke holes in it,
instead of opening it?
You know, this is activating some very dusty neurons as this was the
late 1980s when we were doing this, but I don't recall there being any
other way to access the toner reservoir. Sure, there were torx screws
in there, but I don't remember any of them opening up the reservoir. If
there had been, we'd have done it; we refilled many hundreds of CX
cartridges.
8:35 p.m.
On Thu, 11 Aug 2011, Dave McGuire wrote:
I don't disagree, but nevertheless, in those
days, this was a pretty
big business. We did a LOT of them, and so did a lot of other
companies. Refill kits, complete with a drill bit and little plastic
plugs, were available commercially all over the place. This was a
common and widespread thing.
The first kit that I saw had a torx [compatible] screwdriver. After doing
it that way, I wasn't about to start poking holes. ("baby duck
syndrome"?)
There must've been some compelling reason to do
it this way. I'll be
damned if I can figure out what it was.
Yes.
You can get qo times as many done per unit of time by poking a hole
instead of disassembly. And you could USUALLY get away without the
additional cleaning, etc. inside.
BUT, can you remember where to get some of that toner, so that Tony can
rebuild cartridges?
13 Aug
13 Aug
3:03 p.m.
On 08/11/2011 08:35 PM, Fred Cisin wrote:
Multiple refills, yes, that makes sense.
One other thing that comes to mind is that the blacks were much
better AFTER the first refill. It was widely believed at the time that
the drum was just getting "broken in" by that time.
There
must've been some compelling reason to do it this way. I'll be
damned if I can figure out what it was.
Yes.
You can get qo times as many done per unit of time by poking a hole
instead of disassembly. And you could USUALLY get away without the
additional cleaning, etc. inside. BUT, can you remember where to get some of that toner,
so that Tony can
rebuild cartridges?
Oh hell no, if it's even still available, that was the 1980s.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
3:49 p.m.
On Sat, 13 Aug 2011, Dave McGuire wrote:
I have no idea even of how many different parameters there are in types of
toners, but surely SOMETHING in current availability uses a reasonably
compatible toner (toner itself - no chance of cartridge
interchangability!)
One other thing that comes to mind is that the
blacks were much
better AFTER the first refill. It was widely believed at the time that
the drum was just getting "broken in" by that time.
There was also a finite useful life of the drum, and although the
"sweet-spot" was AFTER the first refill, the quality after many refills
was NOT good. I would guess that it would vary enormously depending on
what kind of content was being done (average density?). But there was
also no question that the CX drum had a life far greater than the amount
of toner in the cartridge. That "problem" was fixed in the SX, by using a
much smaller drum, with a shorter drum life.
HP did not seem to be very receptive to the concept of refilling their
cartridges. "We will pay the postage for you to send it back so that we
can "recycle" it, by crushing it and retireving some amount of the raw
materials in order to get that cartridge out of availability!"
BUT, can you
remember where to get some of that toner, so that Tony can
rebuild cartridges?
Oh hell no, if it's even still available, that was the
1980s. 
3:03 p.m.
Tony Duell wrote:
Obvious ones would include :
Particle size
Fusing tempearture
Electrical conductivity
Permitivity?
Any others (and which of those really matter?)
How about magnetic properites? The old DEC 03 toner was magnetically
attracted within the printer onto a roller before being
electrostatically attracted onto the drum. Presumably, it
contained some iron or iron compound.
--
John Honniball
coredump at gifford.co.uk
3:41 p.m.
On 08/14/2011 03:03 PM, John Honniball wrote:
Tony Duell wrote:
Yes, and in fact in one engine I've seen (might have actually been
the CX or the SX, or maybe the Ricoh, I don't recall) the toner is
"splayed out" by a linear magnet underneath a roller, like whiskers, to
make contact with the drum.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
Obvious ones would include :
Particle size
Fusing tempearture
Electrical conductivity
Permitivity?
Any others (and which of those really matter?)
How about magnetic properites? The old DEC 03 toner was magnetically
attracted within the printer onto a roller before being
electrostatically attracted onto the drum. Presumably, it
contained some iron or iron compound.
15 Aug
15 Aug
4:18 p.m.
On 08/14/2011 03:03 PM, John Honniball wrote:
Yes, you're right. I think both the CX and SX enginers do this.
Tony Duell wrote:
Yes, and in fact in one engine I've seen (might have actually been
the CX or the SX, or maybe the Ricoh, I don't recall) the toner is Obvious ones would include :
Particle size
Fusing tempearture
Electrical conductivity
Permitivity?
Any others (and which of those really matter?)
How about magnetic properites? The old DEC 03 toner was magnetically
attracted within the printer onto a roller before being
electrostatically attracted onto the drum. Presumably, it
contained some iron or iron compound. "splayed out" by a linear magnet underneath
a roller, like whiskers, to
make contact with the drum.
I thought that in at least one case the roller itself was magnetic.
I would think that while some printer engines would require a magnetic
toner, there would be none that would rquire a non-magnetic one (if all
oterh paramters were correct -- of course it may be that all materilas
that meet said parameters turn out to be non-magnetic, but that's not the
issue).
I am going to have to dig out my copy of 'Electrophotography' and see
what it says about the toner.
-tony
10 Aug
10 Aug
2:04 p.m.
I've
dismantled said cartridges (and put them back together), but I've
never refilled one. What toner do you use, and where do you find it?
I don't know. 20+ years ago, we bought toner intended for them. The staff in the computer lab learned the hard way
that PC and CX
cartridges were NOT the same, didn't fit the other machine, and the toner
was "of reversed polarity"
I am not sure what that last bit means. Certainly at least one of the
corona wires is of the opposite polarity between the printer and
photocopier (that's how you get write-for-black in the printer nad
write-for-white in the copier). I beleiuve the toner power is different,
but I am not sure how,
anybody who wanted 4 CX machines (one of which was the
Fedex model with
different specs)
IIRC, the FedEx model was a silly colour (but that's just the casing
parts) and had different clock crystals so that the scanner motor ran
faster. I think it did 400dpi or so (as against the 300 dpi of the normal
model).
I suspec the only people still running such things now are those who have
classic machines with the direct video interface and can't use a more
modern printer.
-tony
4:27 p.m.
On Wed, 10 Aug 2011, Tony Duell wrote:
In those days, it was readily available in USA. We bought a "kit" that
consisted of some toner, a torx compatible screwdrive (not made by Torx),
and a single page of instructions for disassembly, clean, refill, and
assembly. We bought the CX version, and the lab staff used it on PC
cartridge, with a reverse video "solarized" like effect. Later, we bought
the toner by itself in small bulk quantities.
I have no idea what other copiers use toner that might be compatible. If
you can solve THAT issue, then many copiers distribute their toner in
cartridges that are little more than a plastic jar.
That's why I put their statement in quotes. Technically presumably
totally inaccurate, but quite adequately expressive.
I don't
know. 20+ years ago, we bought toner intended for them.
I've never seen such
stuff over here :-( The staff in
the computer lab learned the hard way that PC and CX
cartridges were NOT the same, didn't fit the other machine, and the toner
was "of reversed polarity"
I am not sure what that last bit means.
Certainly at least one of the
corona wires is of the opposite polarity between the printer and
photocopier (that's how you get write-for-black in the printer nad
write-for-white in the copier). I beleiuve the toner powDer is
different,
but I am not sure how,
When you find out the difference, share it with us, and then we will have
correct terminology, rather than "of reversed polarity".
IIRC, the FedEx model was a silly colour (but
that's just the casing
parts) and had different clock crystals so that the scanner motor ran
faster. I think it did 400dpi or so (as against the 300 dpi of the normal
model).
That is exactly my recollection as well. Why was it purple? I could see
UPS wanting BROWN, but I don't recall Fedex obsessing over purple.
I suspec the only people still running such things now
are those who have
classic machines with the direct video interface and can't use a more
modern printer.
Exactly.
IF any of them happen to turn up, should I flag for you:
Eiconscript (imitation Postscript and HPLJ emulation, with memory for 286)
JLASER (multiple versions, the "plus" model includes an interface for
Canon scanner mechanism)
Cordata/Corona Data Systems (software, which is probably no longer
anywhere near the board) refuses to run on 286, easy to use YAFIYGI
interface, and support for it (and LJ) by the Poems Font Editor
DC37 ABC switch
I was unable to stop the college from dumpstering the LJII (SX), which had
an add-in board for JLASER support, which PROBABLY would have permitted
upgrading most CX external interface to SX. SX had a very noticably
better black than CX, apparently due to having a toner blob larger than
(circumscribed) the addreeable box, rather than inscribed. The IIID found
a happy enthusiastic home.
5:42 p.m.
That is exactly my recollection as well.? Why was it
purple?? I could see
UPS wanting BROWN, but I don't recall Fedex obsessing over
purple.
Purple is a major thing at FedEx. The Fed in all of
the different FedEx logos is purple*. Years ago their
planes were mostly purple. When I worked there, they
had an initiative that encouraged going the extra mile
called the Purple Promise.
* The colors of the Ex indicate the operating company.
The original orange is still used for express, green
is for ground (and home delivery), light blue is
for custom critical, red is for freight, silver/grey
is for services which contains the technology and
marketing that support the other operations. I forget
the name of the operating company that is represented
by yellow, but I do think I remember correctly that
they focus on international logistics, particularly
customs.
BLS
11 Aug
11 Aug
2:54 p.m.
On Wed, 10 Aug 2011, Tony Duell wrote:
In those days, it was readily available in USA. We bought a "kit" that
I sorted out my CX about 20 years ago, and I never remembr seeing such
refil kits over here/. Perhaps I wasn't looking on the right places
I
don't know. 20+ years ago, we bought toner intended for them.
I've never
seen such stuff over here :-( Certainly at
least one of the
corona wires is of the opposite polarity between the printer and
photocopier (that's how you get write-for-black in the printer nad
write-for-white in the copier). I beleiuve the toner powDer is
different,
but I am not sure how,
When you find out the difference, share it with us, and then we will have
correct terminology, rather than "of reversed polarity". IF any of them happen to turn up, should I flag for
you:
Eiconscript (imitation Postscript and HPLJ emulation, with memory for 286)
JLASER (multiple versions, the "plus" model includes an interface for
Canon scanner mechanism)
Cordata/Corona Data Systems (software, which is probably no longer
anywhere near the board) refuses to run on 286, easy to use YAFIYGI
interface, and support for it (and LJ) by the Poems Font Editor
DC37 ABC switch
Those do all sound fairly interesting. But I don't think I could afford
shipping (and I don't have space for) another CX printer itself.
I was unable to stop the college from dumpstering the LJII (SX), which had
an add-in board for JLASER support, which PROBABLY would have permitted
upgrading most CX external interface to SX. SX had a very noticably
IIRC there were 2 ways to get a CX-VDO interface (or something close to
one) on an SX printer.
The first was a buffer board that fitted in place of the formatter. It
included things like the differnetial driver/receiver for the laser
control and sync signals. This was an officia Canon product I think.
Somewhere I have a scheamtic of it, it's quite simple
The second was a board that fitted in the optional interface slot in an HP
LJ2 and others that used the same formatter). AFAIK, this was a
third-party product, and I have never seen one.
-tony
8:14 p.m.
On Thu, 11 Aug 2011, Tony Duell wrote:
I sorted out my CX about 20 years ago, and I never
remembr seeing such
refil kits over here/. Perhaps I wasn't looking on the right places
About 10 years ago, the markets around here were full of Inkjet refill
kits. Now those are disappearing, along with the associated DIY carpet
cleaning systems.
The second was a board that fitted in the optional
interface slot in an HP
LJ2 and others that used the same formatter). AFAIK, this was a
third-party product, and I have never seen one.
I believe that that was what the JLASER interface was. But Talltree is
long gone. I wonder if PrinterWorks would have access to such? At least
they would be the a good authority about what is/was available.
8:35 p.m.
On 08/11/2011 08:14 PM, Fred Cisin wrote:
On Thu, 11 Aug 2011, Tony Duell wrote:
I'm pretty sure the JLASER used a modified SX engine, as well. It
definitely did use a big cable going from an ISA card to the printer,
but they sold the card WITH the engine, which I believe was special in
some way. I don't recall ever having seen them selling the ISA card +
video interface card separately, i.e. without the engine.
-Dave
--
Dave McGuire
Port Charlotte, FL
I sorted out my CX about 20 years ago, and I
never remembr seeing such
refil kits over here/. Perhaps I wasn't looking on the right places
About 10 years ago, the markets around here were full of Inkjet refill
kits. Now those are disappearing, along with the associated DIY carpet
cleaning systems.
The second was a board that fitted in the
optional interface slot in an HP
LJ2 and others that used the same formatter). AFAIK, this was a
third-party product, and I have never seen one.
I believe that that was what the JLASER interface was. But Talltree is
long gone. I wonder if PrinterWorks would have access to such? At least
they would be the a good authority about what is/was available.
9:16 p.m.
On Thu, 11 Aug 2011, Dave McGuire wrote:
I'm pretty sure the JLASER used a modified SX
engine, as well. It
definitely did use a big cable going from an ISA card to the printer,
but they sold the card WITH the engine, which I believe was special in
some way. I don't recall ever having seen them selling the ISA card +
video interface card separately, i.e. without the engine.
This was a box-stock LJII from another vendor. The Jlaser was a RAM plus
laser printer interface card, a big cable (DC37 for CX), and a small
adapter card that fit in the back of the LJII. The LJII was still capable
of switching back and forth between JLASER and "normal" HP
"Centronics"
PCL transparently under software control, so we had two computers cabled
to it. For CX LaserJet "minus" and the like, Talltree sold some sort of
adapter board (to go IN the printer?) that was switchable between
JLASER and "normal" LJ HP RS232 PCL. On CX VDO units, such as Cordata,
JLASER was a simple DC37 cable to the "video" input.
9:24 p.m.
On 08/11/2011 09:16 PM, Fred Cisin wrote:
On Thu, 11 Aug 2011, Dave McGuire wrote:
Yes, I remember it. At least, I think I do! ;) Is it possible there
was another product that was as I described above, using a modified SX
engine to achieve higher resolution (not the LJ-III style "resolution
enhancement" trick) and sold as a package with ISA interface and
printer? I'm 100% certain that there was one, but I am NOT 100% certain
that the one I'm remembering was the JLASER. We never got one of those
at my old company in the late 1980s, but we wanted one for in-house use.
(My company sold and installed quite a few similar systems, mostly
the one from QMS that used the raw video interface.)
-Dave
--
Dave McGuire
Port Charlotte, FL
I'm pretty sure the JLASER used a
modified SX engine, as well. It
definitely did use a big cable going from an ISA card to the printer,
but they sold the card WITH the engine, which I believe was special in
some way. I don't recall ever having seen them selling the ISA card +
video interface card separately, i.e. without the engine.
This was a box-stock LJII from another vendor. The Jlaser was a RAM plus
laser printer interface card, a big cable (DC37 for CX), and a small
adapter card that fit in the back of the LJII. The LJII was still capable
of switching back and forth between JLASER and "normal" HP
"Centronics"
PCL transparently under software control, so we had two computers cabled
to it. For CX LaserJet "minus" and the like, Talltree sold some sort of
adapter board (to go IN the printer?) that was switchable between
JLASER and "normal" LJ HP RS232 PCL. On CX VDO units, such as Cordata,
JLASER was a simple DC37 cable to the "video" input.
12 Aug
12 Aug
2:18 p.m.
The second was
a board that fitted in the optional interface slot in an HP
LJ2 and others that used the same formatter). AFAIK, this was a
third-party product, and I have never seen one.
I believe that that was what the JLASER interface was. But Talltree is
long gone. I wonder if PrinterWorks would have access to such? At least
they would be the a good authority about what is/was available.
8 Aug
8 Aug
5:36 p.m.
ALL of those, with the exception of the 1.2M and the
1.4M, MOST before the
5160 came out, I used on my 5150, which I ordered from IBM in 1981 with
16K of RAM.
Do you honestly think that the IBM PC design team thought that people
would hang all this different stuff off their machines?
--
Will
6:14 p.m.
On Mon, 8 Aug 2011, William Donzelli wrote:
Do you honestly think that the IBM PC design team
thought that people
would hang all this different stuff off their machines?
I honestly don't know what they intended. They made provivion for
expansion, such as building in support for [optional] four floppies,
[optional] serial, [optional] parallel, provision for both video systems
to co-exist, etc. They were obviously well aware of the propensity for
expansion of TRS80 and Apple][. Their internal expansion options look
like they were modeled after Apple, but I'm sure that absolutely
everything was completely original.
And having only 5 expansion slots was to demonstrate that they were big
enough that they didn't have to learn from anybody else's mistakes.
But, they included support for cassette write-only storage, RF modulator,
and some early commercials were of a toddler playing with it on the living
room rug (which the Lisa also did).
They explicitly made provision for multiple serial and parallel ports!
They did NOT make provision for multiple FDC, although they also didn't
take steps to block it.
You are, however, absolutely right that I KNOW that I have never been
their target market.
--
Grumpy Ol' Fred cisin at xenosoft.com
6:43 p.m.
On 8 Aug 2011 at 15:14, Fred Cisin wrote:
And having only 5 expansion slots was to demonstrate
that they were
big enough that they didn't have to learn from anybody else's
mistakes.
One thing that really stuck out to me as a design flaw that we are
still living with was the physical design of the expansion slots.
Far too Apple-like (i.e. cheap). Contrast it with most
"professional" buses, such as Multibus, Eurocard, etc) that "cage"
cards and are able to secure cards and ensure a good flow of cooling
air between the cards.
The single-screw bracket with no other retention allows for cards to
pop out of their slots and receive very little in the way of coolig
is just plain wrong, even if it is cheap.
I've seen industrial rackmount PC chassis with hold-down brackets and
rubber "bumpers" to hold cards in place, but it's a mess because of
there being no standard profile for a PC expansion card.
From that I can only conclude that the PC was intended
as a home
entertainment system.
--Chuck
7:48 p.m.
One thing that really stuck out to me as a design flaw
that we are
still living with was the physical design of the expansion slots.
Far too Apple-like (i.e. cheap). ?Contrast it with most
"professional" buses, such as Multibus, Eurocard, etc) that "cage"
cards and are able to secure cards and ensure a good flow of cooling
air between the cards.
I am no fan of the design either, but was the physical connection or
cooing ever really an issue for cards that followed the design specs?
It seems "good enough" - which is just where the bean counters like
things.
I've seen industrial rackmount PC chassis with
hold-down brackets and
rubber "bumpers" to hold cards in place, but it's a mess because of
there being no standard profile for a PC expansion card.
IBM made a line of industrial PCs - beefier supports, with more cooling.
--
Will
8:29 p.m.
On 8 Aug 2011 at 19:48, William Donzelli wrote:
I am no fan of the design either, but was the physical
connection or
cooing ever really an issue for cards that followed the design specs?
It seems "good enough" - which is just where the bean counters like
things.
Just so--I point this out because there was speculation as to what
the target market might be. But even early on, some cards, such as
the PGC (1984), taxed the design. I was mildy surprised to find that
three-PCB sandwich didn't fail after a few hundred hours.
The NEC PC9801, introduced in 1982 had a much superior card system.
I remember thinking that the 9801 was probably a better system for
the money than the 5150. (better graphics, 8086 instead of 8088,
etc.) The only problem was that NEC kept the system closed--
including details about C-Bus.
Certainly PC98 architecture found its way into lots of industrial
applications and is still in use today.
Chuck
8:52 p.m.
Just so--I point this out because there was
speculation as to what
the target market might be. ?But even early on, some cards, such as
the PGC (1984), taxed the design. ?I was mildy surprised to find that
three-PCB sandwich didn't fail after a few hundred hours.
Well, maybe the IBM geeks did the math right.
--
Will
7:54 p.m.
On Mon, 8 Aug 2011, Chuck Guzis wrote:
One thing that really stuck out to me as a design flaw
that we are
still living with was the physical design of the expansion slots.
Far too Apple-like (i.e. cheap). Contrast it with most
"professional" buses, such as Multibus, Eurocard, etc) that "cage"
cards and are able to secure cards and ensure a good flow of cooling
air between the cards.
In addition to the poor mounting, they did not even include the
addressability that Apple had had.
OK, one slot on the 5160 had a difference, but they didn't seem to make
much effort to document that.
From that I can only conclude that the PC was intended
as a home
entertainment system.
Which merely shows an incompetence when it comes to home entertainment.
8:52 p.m.
Come to think of it, didn't the IBM 5551 (JX) host 3 floppy drives in
the main box? Better expansion card design than the 5150 also.
That would be a real collectible system, at any rate, along with a
Mitsubishi Multi-16.
--Chuck
9:21 p.m.
On 8 Aug 2011 at 21:02, William Donzelli wrote:
Yes, but that was a PC sort-of with a word processor
bolted on the
side. Very kludgey in a Japanese sort of way.
Marketed in a very strange way also.
I remember reading about the JX in ComputerWeek and picking up the
phone and calling the local sales office. "How can I get one?" "--
You can't". Supposedly only for the Japanese market, but also
marketed (IIRC) in Australia and NZ, those bastions of the Japanese
language...
--Chuck
9:20 p.m.
Yea,
Those models pop up here in New Zealand now and again and are not THAT rare.
Maybe we were a bit of a test bed for the machine (western country, small
isolated population so if something bombed it didn't take the whole brand
with it..this has certainly been the case in past with other electronic
devices)?
Terry Stewart (Tezza)
----- Original Message -----
From: "Chuck Guzis" <cclist at sydex.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, August 09, 2011 12:52 PM
Subject: Re: Tony Duell's maxed out 5160 and PDP
Come to think of it, didn't the IBM 5551 (JX) host
3 floppy drives in
the main box? Better expansion card design than the 5150 also.
That would be a real collectible system, at any rate, along with a
Mitsubishi Multi-16.
--Chuck
9 Aug
9 Aug
2:03 p.m.
The single-screw bracket with no other retention
allows for cards to
pop out of their slots and receive very little in the way of coolig
is just plain wrong, even if it is cheap.
At least IBM procided a hold-down bracket (and a guide for the front edge
of full-length cards). Unlike the Apple ][ where te expansion cards were
supported by the edge conenctor only, and could wobble and break contact.
-tony
5:08 p.m.
On 08/08/2011 06:43 PM, Chuck Guzis wrote:
One thing that really stuck out to me as a design flaw
that we are
still living with was the physical design of the expansion slots.
Far too Apple-like (i.e. cheap). Contrast it with most
"professional" buses, such as Multibus, Eurocard, etc) that "cage"
cards and are able to secure cards and ensure a good flow of cooling
air between the cards.
The single-screw bracket with no other retention allows for cards to
pop out of their slots and receive very little in the way of coolig
is just plain wrong, even if it is cheap.
PCIe has the little latch on the other end for x16 cards. That seems
to help a bit with the popping out problem.
From that I can only conclude that the PC was
intended as a home
entertainment system.
...a status that it has finally achieved. ;)
-Dave
--
Dave McGuire
Port Charlotte, FL
6:06 p.m.
On 9 Aug 2011 at 17:08, Dave McGuire wrote:
PCIe has the little latch on the other end for x16
cards. That
seems to help a bit with the popping out problem.
Something doubtless learned from the AGP poppers before someone got
the idea of putting a latch on the motherboard.
For me, the idea of cards with little fans mounted to them (or even
CPU heatsinks with fans on them) points to poor design.
I've got an old server with two P3/1GHz Slot 1 CPUs in it. They're
located too close together to get good circulation from heatsink-
mounted fans (besides, I don't trust the little buggers). I took
some 3/8" Plexiglas and worked up ounting for a (Papst) 120mm fan
that sits right over both CPUs. The highest temp I've observed thus
far has been a toasty 38C.
I think IBM did something similar with their Slot 1 Netfinity
servers.
--Chuck
2 p.m.
They did NOT make provision for multiple FDC, although
they also didn't
take steps to block it.
The buffer to drive the DMA requrst line on the IBM FDC card is a 3-state
drvice, whit the eneable software-controllaable. Thus you can disable it
and allow some other device (like a second FDC) to use that DMA channel.
When the PC firsrt came out, the FDC was justabotu the only thing that
needed a DMA cahnnel. So the 3-start bufer was fitted by IBM to allow you
to do somthing they'd not really considered (more DMA devices, inclduing
a second FDC).
-tony
3:47 p.m.
On 9 Aug 2011 at 19:00, Tony Duell wrote:
The IBM FDC was made perfectly capable of sharing the bus with
another FDC; all that is needed is a separate range of I/O ports.
Interrupt and DMA both can be shared. Note that a more=or-less de
facto standard for secondary FDC ports (370h-378h) was well-
established early on. Most floppy controllers respond to a RESET
signal as being in a disabled state.
The XT hard disk controller also used DMA channel 3 and a few network
cards also used ISA DMA.
Once the 80286 came along with its "string" I/O instructions (INS,
OUTS), programmed I/O became quite a bit faster than ISA 8-bit legacy
DMA.
8-bit DMA using what amounts to the DMA controller for an 8085 gets
pretty clumsy, having to be aware of 64K physical address boundaries
(8237 DMA 'wraps" a 64K physical boundary). Later CPUs with
integrated DMA (e.g. 80186, V40 or even the often-forgotten 8089 I/O
cotntroller) all enjoy a continous 20-bit DMA address space.
--Chuck
They did NOT
make provision for multiple FDC, although they also
didn't take steps to block it.
The buffer to drive the DMA requrst line on the IBM FDC card is a
3-state drvice, whit the eneable software-controllaable. Thus you can
disable it and allow some other device (like a second FDC) to use that
DMA channel.
4:13 p.m.
...a bit more on DMA...
The Intel 440BX/GX chipsets were the last ones that I'm aware of to
fully support ISA devices, including those that use legacy DMA.
My experience with later Intel chipsets, such as the 820 that employ
a bridge chip such as the National PC87200 seldom support ISA DMA or
do so with "gotchas".
So it's a good idea not to toss that PII motherboard if you want to
use DMA-oriented ISA devices.
--Chuck
10 Aug
10 Aug
1:55 p.m.
On 9 Aug 2011 at 19:00, Tony Duell wrote:
The IBM FDC was made perfectly capable of sharing the bus with
another FDC; all that is needed is a separate range of I/O ports.
Absolutely, that's waht I said.
They did
NOT make provision for multiple FDC, although they also
didn't take steps to block it.
The buffer to drive the DMA requrst line on the IBM FDC card is a
3-state drvice, whit the eneable software-controllaable. Thus you can
disable it and allow some other device (like a second FDC) to use that
DMA channel. Interrupt and DMA both can be shared. Note that a
more=or-less de
My point is that as the IRQ and DMA request lines are active high, the
only way to share them is to drive them with 3-state buffers that can be
disabled under software control when some other device wishes to use that
IRQ or DMA channel.
IBM kindly used such buffers on the origianl FDC card (and AFAIK all
clone FDCs, even single-chip ones, did likewise). If the had not done so,
having 2 FDCs in the same machine would have been a bit harder.
facto standard for secondary FDC ports (370h-378h) was
well-
established early on. Most floppy controllers respond to a RESET
signal as being in a disabled state.
If it's an exact clone of thr IBM one, it has to. The control signals
come from a '273 latch at U17, this is cleared by the hardware reset
signal. One of the outputs enables the DMA request and IRQ buffers (so
these are disabled after a hrdware reset, and have to be enabled by
software writing to this latch), another bit is inverted and used to
reset the 765 FDC chip itself (so this chip is held in the reset state
until the latch is laoded with an appropraite value).
The XT hard disk controller also used DMA channel 3
and a few network
cards also used ISA DMA.
When the PC first came out, did any IBM expansion card other than the
FDC use DMA?
8-bit DMA using what amounts to the DMA controller for
an 8085 gets
pretty clumsy, having to be aware of 64K physical address boundaries
That;s the polite way of putting it!. IBM was even worse than some othr
8088 machines in that they only had page registers (to provide the top 4
bits of the address), one of which served for 2 channels.
Of course they should have used an 8089 line Apricot did, but that's
another story...
-tony
3:40 p.m.
> > > They did NOT make provision for
multiple FDC, although they also
> > > didn't take steps to block it.
> > The buffer to drive the DMA requrst line on the IBM FDC card is a
> > 3-state drvice, whit the eneable software-controllaable. Thus you can
> > disable it and allow some other device (like a second FDC) to use that
> > DMA channel.
> The IBM FDC was made perfectly capable of sharing the bus with
> another FDC; all that is needed is a separate range of I/O ports.
Tony: Absolutely, that's waht I said.
Fred: Likewise. They didn't take steps to prevent it, but unlike the LPT
and COM boards, they did not provide an "official" alternate address, BIOS
support (could have been an INT1Eh parameter), nor solder pads for
jumpering.
That;s the polite way of putting it!. IBM was even
worse than some othr
8088 machines in that they only had page registers (to provide the top 4
bits of the address), one of which served for 2 channels.
From a BEGINNING programmer perspective: If your
buffer for INT13h I/O
(even single sector), happens to straddle one of the 64K
"boundaries",
then your INT13H call will fail with an error code. That was not a BUG,
just one more poorly documented caveat before using INT13h. It is up to
your code to find whether the memory location assigned to your program
puts your buffer at risk (or just TRY IT), find another buffer location,
and redo the I/O. Many commercial programs failed to do that test, often
resulting in code that "tested" fine, but failed at the customer's
location, depending on where in memory the program happened to get loaded.
From a USER's perspective: FORMAT was one such
program! IF its buffer
happened to straddle one of the boundary lines, it would
FAIL, and FORMAT
would proceed to report a COMPLETY TOTALLY incorrect error message,
blaming the diskette. THAT is a bug. The cause might be the fault of the
original design (Microsoft called it a hardware bug!), but it was FORMAT's
responsibility to move the buffer, or at least correctly identify what
went wrong. (just as it was SMARTDSK's responsibility to recover from a
sector error in delayed writes, rather than locking up on an an
unterminatable R[etry]? condition) Whether the buffer straddled the
boundary could often be easily changed by adding or removing a TSR or two.
Since the "experts" had no clue what was really happening, that led to
much stupid incorrect journalism that concluded that certain TSRs were
incompatible with FORMAT, or even in at least one case, reports that a
certain TSR was NEEDED for FORMAT to work!
--
Grumpy Ol' Fred cisin at xenosoft.com
11 Aug
11 Aug
2:16 p.m.
[Second disk controller in an IBM PC]
Right. I rememebr when I was directly accessing the disk controller, I
would allocate a 512byte buffer (I ws handling 256 byte sectors), thus
ensuring that at leat one half of the allocated area didn't cross a page
boundary. I'd use the thrst half if it was OK, the scond half otherwise.
I've never been bitten by this, but it sounds ridiculous.
Tony: Absolutely, that's waht I said.
Fred: Likewise. They didn't take steps to prevent it, but unlike the LPT
and COM boards, they did not provide an "official" alternate address, BIOS
support (could have been an INT1Eh parameter), nor solder pads for
jumpering.
True.
Talking of solder pads, I assume _everyone_ spotted the unneceessary vias
on the original parallel (printer) card. If you cut the track between
them and jumper one of the vias to another nearby via, you control th OE/
line of the data latch by the othewise unused bit of the control port,
thus making it a bidirectional port. This is not docuemted, not even in
th TechRef. Alas the same traces were not present on the MDA card (which
has a built-in parallel port, of course).
That;s the polite way of putting it!. IBM was
even worse than some othr
8088 machines in that they only had page registers (to provide the top 4
bits of the address), one of which served for 2 channels.
From a BEGINNING programmer perspective: If your
buffer for INT13h I/O
(even single sector), happens to straddle one of the 64K
"boundaries",
then your INT13H call will fail with an error code. That was not a BUG, just one more poorly documented caveat before using
INT13h. It is up to
your code to find whether the memory location assigned to your program
puts your buffer at risk (or just TRY IT), find another buffer location,
The linux fdrawcmd ioctl() allocates its own buffer which will not cross
a page bondary, and transfers the data to/from the array you pass to it
(which can be anywhere). In other words th OS handles the details fro
you, as it should do.
and redo the I/O. Many commercial programs failed to
do that test, often
resulting in code that "tested" fine, but failed at the customer's
location, depending on where in memory the program happened to get loaded.
Ouch!
From a USER's perspective: FORMAT was one such
program! IF its buffer
happened to straddle one of the boundary lines, it would
FAIL, and FORMAT
would proceed to report a COMPLETY TOTALLY incorrect error message,
blaming the diskette. THAT is a bug. The cause might be the fault of the original design (Microsoft called it a hardware bug!),
but it was FORMAT's
Well, I woill happily agree that the hardware design is a right kludge
(the language I would like to use to describe it is not suitable for a
public list :-)), but it is surely the program writer's responsibility to
ensre the program runs on said hardare under all conditions.
-tony
8:23 p.m.
On Thu, 11 Aug 2011, Tony Duell wrote:
Well, I woill happily agree that the hardware design
is a right kludge
Not being a Brit, I'd call it a WRONG kludge.
(the language I would like to use to describe it is
not suitable for a
public list :-)),
You would be shocked at what is considered acceptable here and
not there,
and vice versa.
but it is surely the program writer's
responsibility to
ensre the program runs on said hardare under all conditions.
The insidious aspect of it was that most of us did not understand the
"inner workings" well enough to know that there could be a problem until
it bit us. In my case, I was lucky enough to have been curious enough to
go through all of the error return codes of INT13h, and ask myself, "what
the hell is THAT about??"
12 Aug
12 Aug
1:50 p.m.
On Thu, 11 Aug 2011, Tony Duell wrote:
Ah, but is a 'wrong kludge' something that's not a kludge and therefore a
good design ?
Well, I woill happily agree that the hardware
design is a right kludge
Not being a Brit, I'd call it a WRONG kludge. but it is
surely the program writer's responsibility to
ensre the program runs on said hardare under all conditions.
The insidious aspect of it was that most of us did not understand the
"inner workings" well enough to know that there could be a problem until
it bit us. In my case, I was lucky enough to have been curious enough to
go through all of the error return codes of INT13h, and ask myself, "what
the hell is THAT about??"
3:39 p.m.
On 12 Aug 2011 at 18:50, Tony Duell wrote:
Things get more complicated when later (e.g. Win NT...) MS operating
systems are used. Not having a Microsoft-conforming DOS boot
sector results in a "General Failure" error from the driver, even
when the virtualized INT 13H services are used. Note that the
floppy in question can be perfectly readable otherwise.
Another Microsoft abomination was the volume-tracking driver in Win95-
-it would attempt to rewrite the boot sector of any floppy being read
or written. Really smart planning there.
--Chuck
The insidious
aspect of it was that most of us did not understand
the "inner workings" well enough to know that there could be a
problem until it bit us. In my case, I was lucky enough to have
been curious enough to go through all of the error return codes of
INT13h, and ask myself, "what the hell is THAT about??"
Err, yes... Anotehr good reason to read the TechRef from cover to
cover.
10 Aug
10 Aug
5:29 p.m.
On 10 Aug 2011 at 18:55, Tony Duell wrote:
That;s the polite way of putting it!. IBM was even
worse than some
othr 8088 machines in that they only had page registers (to provide
the top 4 bits of the address), one of which served for 2 channels.
Of course they should have used an 8089 line Apricot did, but that's
another story...
Remember that the 8237 DMAC furnished dual functionality--regular DMA
services and also memory refresh. With an 8089 having only two DMA
channels, you're forced to go to something like an 8203/8207 DRAM
controller. The use of both the 8089 (and its support circuitry) and
the memory controller would have offended the bean counters mightily,
I suspect. (Witness the cost-cutting done on the Peanut).
As far as PC-type 8-bit expansion cards using DMA, there were
several. I don't know how long they were on the drawing board, so
it's pretty difficult saying whether they were or weren't in the
original game plan.
* MFM hard disk controller (uses DMA 3)
* GPIB adapter
* SDLC adapter
--Chuck
11 Aug
11 Aug
3:08 p.m.
Of course they
should have used an 8089 line Apricot did, but that's
another story...
Remember that the 8237 DMAC furnished dual functionality--regular DMA
services and also memory refresh. With an 8089 having only two DMA channels, you're forced to go to something like an
8203/8207 DRAM
controller. The use of both the 8089 (and its support circuitry) and
the memory controller would have offended the bean counters mightily,
I suspect. (Witness the cost-cutting done on the Peanut).
I have noi idea of the cost of an 8089 at the time, but th esupport
circuity would have been a lot less than that needed by the 8237 (no pag
register and associated logic). Yes, you'd have needed some memory
refresh hardware, but countless other machines provided that without
needed a DMA channel to do it (and you didn't _have_ to used a
single-chip DRAM controller). Remmber that in the PC, all the DMAC did
for the RAM was the refresh operation, the timing of RAS/CAS, addres
multiplexing, etc was done in separate hardware. Adding a refresh counter
and th eassociated logic would npt have been that much exta.
-tony
4:52 p.m.
On 11 Aug 2011 at 20:08, Tony Duell wrote:
I have noi idea of the cost of an 8089 at the time,
but th esupport
circuity would have been a lot less than that needed by the 8237 (no
pag register and associated logic). Yes, you'd have needed some memory
refresh hardware, but countless other machines provided that without
needed a DMA channel to do it (and you didn't _have_ to used a
single-chip DRAM controller). Remmber that in the PC, all the DMAC did
for the RAM was the refresh operation, the timing of RAS/CAS, addres
multiplexing, etc was done in separate hardware. Adding a refresh
counter and th eassociated logic would npt have been that much exta.
Looking back at the 8089 datasheet, I'm not convinced that it
provides much for the money. In the minimum-component "local" mode,
it executes out of the 808x CPU memory space, being another contender
for that resource. At 5MHz with no contention/interleaving it can
transfer a byte every microsecond, best case--not much better than an
8237.
The bean counters would not have been happy.
--Chuck
8:08 p.m.
Looking back at the 8089 datasheet, I'm not
convinced that it
provides much for the money. ?In the minimum-component "local" mode,
it executes out of the 808x CPU memory space, being another contender
for that resource. ?At 5MHz with no contention/interleaving it can
transfer a byte every microsecond, best case--not much better than an
8237.
This is the main reason why the 8089 failed in the market. For what it
did, it was just too expensive.
--
Will
9 Aug
9 Aug
5:44 p.m.
On 08/08/2011 05:36 PM, William Donzelli wrote:
The original IBM PC design team couldn't find their own asses with
both hands, a funnel, and a roadmap. It was a success in the market
purely by accident. I don't think their original intentions really
carry much weight here; it was a miracle the thing worked at all, much
less sold.
-Dave
--
Dave McGuire
Port Charlotte, FL
ALL of those,
with the exception of the 1.2M and the 1.4M, MOST before the
5160 came out, I used on my 5150, which I ordered from IBM in 1981 with
16K of RAM.
Do you honestly think that the IBM PC design team thought that people
would hang all this different stuff off their machines?
11:03 p.m.
On Tue, 9 Aug 2011, Dave McGuire wrote:
The original IBM PC design team couldn't find
their own asses with
both hands, a funnel, and a roadmap. It was a success in the market
purely by accident. I don't think their original intentions really
carry much weight here; it was a miracle the thing worked at all, much
less sold.
I disagree. No problem with the sentiment that it was a miracle that it
worked at all, although the prevailing "wisdom" at the time seemed to be
that parity errors could be eliminated as a problem by ceasing to
implement parity.
BUT, . . .
30 years ago, we agreed that no matter how bad IBM might implement the
machine, that within less than a decade, computers would be in two
categories: IBM and imitations of IBM. We underestimated Apple's
remarkable ability to survive IBM.
The speculation before the PC was bizarre! One mag said that the new
machine would be switchable between emulating TRS80, Apple][, and C/M.
10 Aug
10 Aug
12:46 a.m.
On 9 Aug 2011 at 20:03, Fred Cisin wrote:
The speculation before the PC was bizarre! One mag
said that the new
machine would be switchable between emulating TRS80, Apple][, and C/M.
There was a lot of speculation that the new machine would be 68000
based. Shortly before the 5150 escaped, IBM came out with their
rather cool CS9000 lab computer using the 68K and that just raised
the anticipation.
It was a letdown to discover that the 5150 was the Displaywriter Jr.
When compared with the 5100, which was quite innovative, the letdown
was palpable.
--Chuck
8 Aug
8 Aug
3:38 p.m.
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
Who ever used more than 2 floppies on a PC anyway? Or more than 2 hard
disks (back then, before PC based servers)?
3:19 p.m.
replying to ard at p850ug1.demon.co.uk (Tony Duell)
Actually, I forgot about the scond disk controller, which means the thing
can actually have 6 floppy drives. However, the last 2 are not accessible
> Nice! There's been a resurgence in
appreciation of the "classics".
The 5160 I just mentioned with 4 floppy drive
How did you accomplish that?
- a controller that honors all 4 drive selects, not just 2
- primary & secondary disk controller
- Compaticard or other disk controller
at "non-standard" address & IRQ? from MS-DOs, only from special softwate.
As for the 4 normal drives, I have the original IBM controller board.
This has an edge conenctor at the front fro a nromal cable (with the 'IBM
twist') to connect it to the 2 internal 360K (full height) 5.25" drives.
On the rear bracket of that controller is a DC37 connector which takes
another cable with an 'IBM twist' to link to 2 exteranal 720K 3.5" drives.
[...]
MDA and CGA
cards,
Dual display? What programs support that? From what I rememebr, and I am not going to pull the
case now, the
boards in this machine are :
-- A clone XT multi-IO card, modified. This provides 2 of the serial
ports, a parallel port, a joystick port, and originally a floppy
controller. I replaced the single chip for that with a pin-compatible
PC/AT disk controller chip, changed the clock crystal, and modified the
address decoding to put it ant another address. This is the 'extra' disk
cotnroller that I ahve hooked up to a pair of 8" drives
-- A I2C interface (Elektor design), based o nthe PCF8584 chip
-- Hard disk controller
-- MDA card
-- CGA card
-- Standard floppy controller
-- Dual serial card
-- 8255 24 line user port (Maplin kit)
Of course I've done the normal modification ot the motherboard to put
640K of RAM on there.
For those of us USING the stuff at the time,
we were always upgrading it piece by piece.
That's why it's so hard to find one with all original parts.
I was quite lucky. I managed to find another PC/XT (IBM 5160) which was
_all_ original IBM. From what I recall, it has a CGA card, floppy
controller, Async serial card, parallel card, one 360K floppy drive
and.... the expansion unit contianing 2 hard drives and the controller
for them. Oh, and origianl IBM 5153 CGA monitor too.
A PDP8/e with 32K words of real core
Do the boards have clear plastic for you to see the cores? That's such a treat: true non-volatile memory!
Indeed... Non volatile, but reading is destructive :-)
> My high school started everyone with
programming BASIC
> on the HP 9820A:
I find that hard to believe for one good reason.
The HP9820 never ran BASIC...
You got me there.
It was kinda almost similar to BASIC.
I got a HP-28C when it was new and it's still my primary calculator.
I bought the first 28C I saw -- back when it was new, and I was an
undergrad at the time. Really cut into my budget. A few yers later, I
ordered a 48SX the day it was announced, that machine cured the major
problem I had with the 28C, namely no way to back it up (the 48SX hasa
serial port, of course).
nd then some yeras later I made the big mistake of buying a 49G. That
machine was not, IMHO 'of merchantable quality' when I bought it. There
were so many bugs in the first few ROM releases (flash upgradable,
thankfully). Heck, the version that was in the machine when I bought it
would return a dot product when you executed the cross product function.
Advertised features simply did not exist in the first few ROM versions.
What makes it worse is that I wrote a letter of complaint to HP. They
didn't bother to reply. I sent them a further letter, enclosing IRCs.
They didn't reply again. And for that reason. while I will happily
restore and use classic HP amchines, I will never buy another new product
from them. I do not deal with companies who ignore
their customers.
I really need that 4 line display to use RPN so I can
see the stack.
Yes, it is very useful, as is the arbirary size stack of any types of
object (the older machines had a 4 level stack of real numbers only). Not
many machines allow you to build up a string on the stack, and then
execute it as a program.
-tony
5:05 p.m.
On 8 Aug 2011 at 20:19, Tony Duell wrote:
Actually, I forgot about the scond disk controller,
which means the
thing can actually have 6 floppy drives. However, the last 2 are not
accessible from MS-DOs, only from special softwate.
I have an old "tower" fitted with 5 3.5" floppies and 1 360K floppy,
three controllers (including the one on the hard disk controller).
The three floppy controllers use different DMA, IRQ and port
addresses. All are accessible through DOS (drivers).
I have a bit of software that we marketed for a short time that
writes to three drives simultaneously as a copying utility. We sold
it to a few commercial customers.
--Chuck
11 Aug
11 Aug
4:31 p.m.
On 08/07/2011 07:57 PM, Jeff Jonas wrote:
My pre-PC systems (mostly Z80 CP/M)
all support 4 floppies per the original specs.
But NOOOOOOOOOOOOOO, the PC's "cable twist" ruined that,
limiting things to 2 floppies and 2 hard drives per controller.
Not 'till SATA have we been able to un-do that mistake :-(
Huh? There have been quite a few mass storage interfaces that don't
limit you in such ways. In fact, there are still quite a few right now.
(sorry for the late reply, I've been on the road)
-Dave
--
Dave McGuire
Port Charlotte, FL
4797
days inactive
4805
days old
70 comments
10 participants
participants (10)
-
ard＠p850ug1.demon.co.uk -
blstuart＠bellsouth.net -
cclist＠sydex.com -
cisin＠xenosoft.com -
coredump＠gifford.co.uk -
jeffj＠panix.com -
mcguire＠neurotica.com -
pu1bzz.listas＠gmail.com -
terry＠webweavers.co.nz -
wdonzelli＠gmail.com