11 Feb
2006
11 Feb
'06
3:24 a.m.
I discovered tonight that the right-most ISA slot (the one closest to
the power supply) in my newly-unearthed 5160 won't take the Silicon
Valley Computer ADP50L IDE controller I'm playing with. It will take
other (very short!) cards, but when I put the IDE controller into it,
either the option ROM is invisible (ie. 55AA not found) or won't
initialize for some reason, because I don't get functionality out of it
(least of all, the boot "banner" produced by the card is missing).
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
I remember that some of the very first PCI Pentium boards (circa
1994-1995) would only take PCI video cards in one particular PCI slot,
so maybe that's where I'm pulling this from...
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
11 Feb
11 Feb
9:51 a.m.
From the XT Technical Reference:
"Timing requirements on slot J8 are much stricter than those on slots J1
through J7. Slot J8 also requires the card to provide a signal
designating when the card is selected."
Traditionally that slot was populated by the asynch card, which probably
meets the requirements. It's possible that your IDE card is either off
on the timing or doesn't have the card select that the XT needs.
12:46 p.m.
Jim Leonard wrote:
I discovered tonight that the right-most ISA slot (the
one closest to
the power supply) in my newly-unearthed 5160 won't take the Silicon
Valley Computer ADP50L IDE controller I'm playing with. It will take
other (very short!) cards, but when I put the IDE controller into it,
either the option ROM is invisible (ie. 55AA not found) or won't
initialize for some reason, because I don't get functionality out of it
(least of all, the boot "banner" produced by the card is missing).
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
I think that there is. I have a Western Digital XT-150 IDE-XT
adapter, and WD's docs specifically state that "for IBM XT owners", the
card cannot be placed in the slot nearest the PSU. This also applies to
the WD XT-140.
I have no idea why that restriction exists; I just remembered seeing
it in the XT-150 docs.
Doc
1:30 p.m.
The old rule of thumb (or wifes tale) was on the original PC and XT's it
was best to put all the high bandwidth adapters closer towards the power
supply like the Memory, then the MFM, RLL, and SCSI controllers,
followed by networking and video, then parallel and serial cards...
Do you have another system to test your card in, perhaps it is the card
or it could be a conflicting mem i/o setting?
Curt
Doc Shipley wrote:
Jim Leonard wrote:
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.375 / Virus Database: 267.15.6/257 - Release Date: 2/10/2006
I discovered tonight that the right-most ISA slot
(the one closest to
the power supply) in my newly-unearthed 5160 won't take the Silicon
Valley Computer ADP50L IDE controller I'm playing with. It will take
other (very short!) cards, but when I put the IDE controller into it,
either the option ROM is invisible (ie. 55AA not found) or won't
initialize for some reason, because I don't get functionality out of
it (least of all, the boot "banner" produced by the card is missing).
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
I think that there is. I have a Western Digital XT-150 IDE-XT
adapter, and WD's docs specifically state that "for IBM XT owners",
the card cannot be placed in the slot nearest the PSU. This also
applies to the WD XT-140.
I have no idea why that restriction exists; I just remembered seeing
it in the XT-150 docs.
Doc
1:48 p.m.
On 2/11/2006 at 2:24 AM Jim Leonard wrote:
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
Yes. The 8th card slot on the XT is "different" in its timings. Mostly
meant for certain specific cards; I've got a bunch of old documentation for
add-in cards that say they won't work in that slot. Offhand, I can't
recall which cards work there, but a guess might be a memory expansion
card.
Cheers,
Chuck
1:51 p.m.
the timing was different for use with the expansion chassis adapter
card. Probably to account for the associated delays times with extra
buffers that would have been required for the cable connect between
the two boxes.
best regards, Steve Thatcher
At 01:48 PM 2/11/2006, you wrote:
On 2/11/2006 at 2:24 AM Jim Leonard wrote:
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
Yes. The 8th card slot on the XT is "different" in its timings. Mostly
meant for certain specific cards; I've got a bunch of old documentation for
add-in cards that say they won't work in that slot. Offhand, I can't
recall which cards work there, but a guess might be a memory expansion
card.
Cheers,
Chuck 
1:55 p.m.
In article <7.0.1.0.2.20060211134858.02342ae0 at earthlink.net>,
Steve Thatcher <melamy at earthlink.net> writes:
the timing was different for use with the expansion
chassis adapter
card.
I have one of these cards if anyone wants it!
--
"The Direct3D Graphics Pipeline"-- code samples, sample chapter, FAQ:
<http://www.xmission.com/~legalize/book/>
Pilgrimage: Utah's annual demoparty
<http://pilgrimage.scene.org>
3:56 p.m.
I discovered tonight that the right-most ISA slot (the one closest to
the power supply) in my newly-unearthed 5160 won't take the Silicon
Valley Computer ADP50L IDE controller I'm playing with. It will take
other (very short!) cards, but when I put the IDE controller into it,
either the option ROM is invisible (ie. 55AA not found) or won't
initialize for some reason, because I don't get functionality out of it
(least of all, the boot "banner" produced by the card is missing).
I can only conclude that:
1. The IDE board is slightly awry
2. The right-most ISA slot is slightly awry
3. There is some technical limitation I'm not aware of
#1 and #2 are easy to accept, but not #3. Is there such a limitation?
Yes, it's more (3) than (2). There is no fault with your particular board.
For some reason, the data lines on slot 8 are wired to the other side of
a bus buffer chip to the data lines on the other 7 slots. A card in slot
8, therefore, has to assert a special signal on the ISA connector to
enable this buffer -- my memory says it's that it must pull pin B8 low
during a read access to any device on that card.
The IBM Async card could do this (fit an unmarked jumper on the board).
Some Microsoft mouse interface cards did this too. I homebuilt one card
that did it (I needed it to go in slot 8 of an XT). AFAIK no other cards do.
-tony
4:26 p.m.
Tony Duell wrote:
For some reason, the data lines on slot 8 are wired to
the other side of
a bus buffer chip to the data lines on the other 7 slots. A card in slot
8, therefore, has to assert a special signal on the ISA connector to
enable this buffer -- my memory says it's that it must pull pin B8 low
during a read access to any device on that card.
Thanks to everyone for the information! I was *not* expecting that slot
8 really was special, but empirical evidence was telling me otherwise.
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
5:23 p.m.
On Sat, 11 Feb 2006, Tony Duell wrote:
For some reason, the data lines on slot 8 are wired to
the other side of
a bus buffer chip to the data lines on the other 7 slots. A card in slot
8, therefore, has to assert a special signal on the ISA connector to
enable this buffer -- my memory says it's that it must pull pin B8 low
during a read access to any device on that card.
The IBM Async card could do this (fit an unmarked jumper on the board).
Some Microsoft mouse interface cards did this too. I homebuilt one card
that did it (I needed it to go in slot 8 of an XT). AFAIK no other cards do.
We don't know what IBM ORIGINALLY intended to do with that slot.
Whatever it was, might not have ever happened.
When the XT came out, IBM did not want people to plug other boards in
there. They also had a horrendous surplus of some boards, such as the
Asynch (RS232 & 20mA), due to the popularity of "multifunction" boards.
So the XT was sold with a "FREE" Asynch board already installed in slot 8.
--
Grumpy Ol' Fred cisin at xenosoft.com
5:45 p.m.
On 2/11/2006 at 2:23 PM Fred Cisin wrote:
When the XT came out, IBM did not want people to plug
other boards in
there. They also had a horrendous surplus of some boards, such as the
Asynch (RS232 & 20mA), due to the popularity of "multifunction" boards.
So the XT was sold with a "FREE" Asynch board already installed in slot 8.
Under "puzzles for the ages", one might also wonder what the IBM was
thinking when they added the 20ma current loop circuitry to the board. Did
they really think that folks were going to use them to drive an ASR 33? Or
that RS232-to-current loop adapters were hard to obtain?
Cheers,
Chuck
6:22 p.m.
Under "puzzles for the ages", one might also
wonder what the IBM was
thinking when they added the 20ma current loop circuitry to the board. Did
they really think that folks were going to use them to drive an ASR 33? Or
that RS232-to-current loop adapters were hard to obtain?
A few random guesses (as I haev said before, you can find out _what_ was
done (e.g. by looking at schemaitcs or listings), but not in general _why_)
1) The IBM PC had similar design ideas to the Apple ][ (open machine,
basic/cassette port, etc). One of the early Apple ][ serial boards had a
current loop intenrface (and the manual for that _did_ describe how to
connect it to an ASR33)
2) IIRC IBM made some current loop printers. A friend of mine has a
System 23 (Datamaster), I seem to remember the printer on that has a
current loop interface
3) There were other current loop devices around, maybe the extra
circuitry was added for people who wanted to use thoase Why have to
bother with an ecternal converter unit?
-tony
6:46 p.m.
New subject: Current loop interface (was: 5160 motherboard slot restrictions?)
In article <m1F844e-000IyFC at p850ug1>,
ard at p850ug1.demon.co.uk (Tony Duell) writes:
Speaking of current loop interfaces, just exactly how is data
transmitted? Is it a current pulse train?
--
"The Direct3D Graphics Pipeline"-- code samples, sample chapter, FAQ:
<http://www.xmission.com/~legalize/book/>
Pilgrimage: Utah's annual demoparty
<http://pilgrimage.scene.org>
Under
"puzzles for the ages", one might also wonder what the IBM was
thinking when they added the 20ma current loop circuitry to the board. Did
they really think that folks were going to use them to drive an ASR 33? Or
that RS232-to-current loop adapters were hard to obtain?
A few random guesses [...]
7:05 p.m.
New subject: Current loop interface (was: 5160 motherboard slot restrictions?)
On 2/11/2006 at 4:46 PM Richard wrote:
Speaking of current loop interfaces, just exactly how
is data
transmitted? Is it a current pulse train?
It's exactly what it sounds like. Instead of a voltage driven signalling
as RS-232C, the current loop interface is driven with a current source
(20ma or the older 60ma) for space and the signal is interrupted (no
current) for mark. The big advantage is that so long as the specified
current is maintained, the distance between the sender and receiver can be
substantial. In the old days, the selector magnet on a teletype would be
part of that circuit. IIRC, this is also where the name for the "break"
key came from--it interrupts the circuit.
Cheers,
Chuck
11:06 p.m.
New subject: Current loop interface (was: 5160 motherboard slot restrictions?)
On Saturday 11 February 2006 07:05 pm, Chuck Guzis wrote:
On 2/11/2006 at 4:46 PM Richard wrote:
The first computer that I actually got hands-on time with (I actually built
the thing) was a Heath H11 system, back around the beginning of 1978 or so.
The "terminal" that we started out with at the time was an LA36, and the
interface used was current-loop. I don't know why that was chosen, at the
time...
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
Speaking of current loop interfaces, just exactly
how is data
transmitted? Is it a current pulse train?
It's exactly what it sounds like. Instead of a voltage driven signalling
as RS-232C, the current loop interface is driven with a current source
(20ma or the older 60ma) for space and the signal is interrupted (no
current) for mark. The big advantage is that so long as the specified
current is maintained, the distance between the sender and receiver can be
substantial. In the old days, the selector magnet on a teletype would be
part of that circuit. IIRC, this is also where the name for the "break"
key came from--it interrupts the circuit.
Cheers,
Chuck
11:51 p.m.
New subject: Current loop interface (was: 5160 motherboard slot restrictions?)
On 2/11/2006 at 11:06 PM Roy J. Tellason wrote:
The "terminal" that we started out with at
the time was an LA36, and the
interface used was current-loop. I don't know why that was chosen, at
the time...
Probably not a factor, but one nice thing about current loop is that you
don't need voltage sources for mark and space levels.
12 Feb
12 Feb
5:22 a.m.
New subject: Current loop interface (was: 5160 motherboard slot restrictions?)
Speaking of current loop interfaces, just exactly how
is data
transmitted? Is it a current pulse train?
Basically, yes. For the digital current loops we're talking about here,
one state is represented by a given current (normally either 20mA or
60mA) flowing round the loop. The other state is no current. So to send
serial data, you turn the current on and off apropriately.
There is also an analogue current loop used for industrial sensors. It
typically uses currents between 4mA and 20mA.
-tony
11 Feb
11 Feb
6:49 p.m.
On 2/11/2006 at 11:22 PM ard at p850ug1.demon.co.uk wrote:
A few random guesses (as I haev said before, you can
find out _what_ was
done (e.g. by looking at schemaitcs or listings), but not in general
_why_)
When the CL interface was dropped by IBM, IIRC, the silence was deafening.
I suspect that customers who really needed the CL interface added an
external converter anyway because it wasn't widely known that the card
included such an interface.
My guess is that some IBM engineer was tasked with coming up with a serial
card and probably couldn't stand the desolation of the half-empty card.
Cheers,
Chuck
7:07 p.m.
On Saturday 11 February 2006 18:22, you wrote:
The writeup that IBM published at the time, said they handled slot 8
differently to support the expansion interface. The bus driver was normally
off unless a device the the expansion chassis was addressed.
I assume it would be for RFI reduction, but that's a guess on my part.
I've never seen docs on the expansion interface, so can't verify it.
joe lang
Under
"puzzles for the ages", one might also wonder what the IBM was
thinking when they added the 20ma current loop circuitry to the board.
Did they really think that folks were going to use them to drive an ASR
33? Or that RS232-to-current loop adapters were hard to obtain?
A few random guesses (as I haev said before, you can find out _what_ was
done (e.g. by looking at schemaitcs or listings), but not in general _why_)
1) The IBM PC had similar design ideas to the Apple ][ (open machine,
basic/cassette port, etc). One of the early Apple ][ serial boards had a
current loop intenrface (and the manual for that _did_ describe how to
connect it to an ASR33)
2) IIRC IBM made some current loop printers. A friend of mine has a
System 23 (Datamaster), I seem to remember the printer on that has a
current loop interface
3) There were other current loop devices around, maybe the extra
circuitry was added for people who wanted to use thoase Why have to
bother with an ecternal converter unit?
-tony
13 Feb
13 Feb
10:14 a.m.
The writeup that IBM published at the time, said they
handled slot 8
differently to support the expansion interface. The bus driver was normally
off unless a device the the expansion chassis was addressed.
I assume it would be for RFI reduction, but that's a guess on my part.
I will ahve to check the techref to see if the expansion card (I assume
you mean the card that links to the 5161 expansion unit) does anything
with pin B8.
I do know that the expansion card does not _have_ to be in slot 8. It
will work in any slot. It's supported on the 5155, where the card will
not physically go in slot 8. Anf I've seen working systems with an Async
card in slot 8 and the expansion card in some other slot.
-tony
16 Feb
16 Feb
7:01 p.m.
Tony Duell wrote:
Store Loop is basically 38400 baud current loop, using a protocol kind
of similar to Broken Ring. It's bloody horrible, and the last few shops
using it for their EPOS systems have now gone over to nice new systems
that use Ethernet, just like the rest of the world.
I must try and grab a couple of Store Loop cards before I finish working
at Big Blue.
Gordon.
Under
"puzzles for the ages", one might also wonder what the IBM was
thinking when they added the 20ma current loop circuitry to the board. Did
they really think that folks were going to use them to drive an ASR 33? Or
that RS232-to-current loop adapters were hard to obtain?
A few random guesses (as I haev said before, you can find out _what_ was
done (e.g. by looking at schemaitcs or listings), but not in general _why_)
1) The IBM PC had similar design ideas to the Apple ][ (open machine,
basic/cassette port, etc). One of the early Apple ][ serial boards had a
current loop intenrface (and the manual for that _did_ describe how to
connect it to an ASR33)
2) IIRC IBM made some current loop printers. A friend of mine has a
System 23 (Datamaster), I seem to remember the printer on that has a
current loop interface
3) There were other current loop devices around, maybe the extra
circuitry was added for people who wanted to use thoase Why have to
bother with an ecternal converter unit?
-tony
6735
days inactive
6741
days old
20 comments
12 participants
participants (12)
-
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
cisin＠xenosoft.com -
curt＠atarimuseum.com -
doc＠mdrconsult.com -
gordonjcp＠gjcp.net -
jclang＠notms.net -
legalize＠xmission.com -
mbbrutman-cctalk＠brutman.com -
melamy＠earthlink.net -
rtellason＠blazenet.net -
trixter＠oldskool.org