Ferranti ULAs

[gareth evans]

Used to work at Vanwell Data Systems in Melksham, Wilts, who
provided Ferranti with the memory-behind-the-pin test eqpt.

ISTR visiting Ferranti in Gem Mill in Oldham.

[Ian]

Post by gareth evans
Used to work at Vanwell Data Systems in Melksham, Wilts, who
provided Ferranti with the memory-behind-the-pin test eqpt.
ISTR visiting Ferranti in Gem Mill in Oldham.

Now a housing estate. Worked there around 1986, then in Hollinwood,
where they did the ULAs and ASICs.

https://www.jusme.com/perm/gem/gallery/pages/page_0001.html

"There's trouble at t' mill"

[Bob Eager]

Post by Ian

Used to work at Vanwell Data Systems in Melksham, Wilts, who provided
Ferranti with the memory-behind-the-pin test eqpt.
ISTR visiting Ferranti in Gem Mill in Oldham.

Now a housing estate. Worked there around 1986, then in Hollinwood,
where they did the ULAs and ASICs.
https://www.jusme.com/perm/gem/gallery/pages/page_0001.html
"There's trouble at t' mill"

I didn't see the original post, but I bought my first computer in 1984.
Had a lot of those ULAs in it. I did a fake service call, because I knew
the first thing the technician would do was to glue heatsinks to all ofg
them - they ran VERY hot.

For those who are interested, it was an Advance 86B, see:

http://www.computinghistory.org.uk/det/7141/Ferranti-Advance-86/

The comment about compatibility is actually from me!

[gareth evans]

Post by Bob Eager

Post by Ian

Used to work at Vanwell Data Systems in Melksham, Wilts, who provided
Ferranti with the memory-behind-the-pin test eqpt.
ISTR visiting Ferranti in Gem Mill in Oldham.

Now a housing estate. Worked there around 1986, then in Hollinwood,
where they did the ULAs and ASICs.
https://www.jusme.com/perm/gem/gallery/pages/page_0001.html
"There's trouble at t' mill"

I didn't see the original post, but I bought my first computer in 1984.
Had a lot of those ULAs in it. I did a fake service call, because I knew
the first thing the technician would do was to glue heatsinks to all ofg
them - they ran VERY hot.
http://www.computinghistory.org.uk/det/7141/Ferranti-Advance-86/
The comment about compatibility is actually from me!

AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

Had to do the board support using one of these in the Herald Pentara
and KBX100 PABX exchanges circa 1981

(R.I.P. Pye TMC)

[Vir Campestris]

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

err...
<https://en.wikipedia.org/wiki/Intel_8259>

tells me what my memory dredged up from all those years ago:

"Edge and level interrupt trigger modes are supported by the 8259A."

I used it in a non-IBM compatible device.

Andy

[Scott Lurndal]

Post by Vir Campestris

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

err...
<https://en.wikipedia.org/wiki/Intel_8259>
"Edge and level interrupt trigger modes are supported by the 8259A."

The tricky part is that the edge signal asserted level must be maintained until the CPU
acknowledges the interrupt to the 8259/8259A.

[gareth evans]

Post by Vir Campestris

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

err...
<https://en.wikipedia.org/wiki/Intel_8259>
"Edge and level interrupt trigger modes are supported by the 8259A."
I used it in a non-IBM compatible device.

Accepting that I might have experienced a senior moment*****,
but that statement is about the 8259A and not the 8259.

*****( I used to be big-headed, but now I'm perfect"???? :-)

[Johnny Billquist]

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

Not sure why you would expect edge triggered any more than level. In
fact, most older hardware use level triggered.
Edge triggered is sometimes problematic/risky, and certainly run the
risk of ending up interrupts not recognized if shared interrupt lines,
or mistakes in software.

Johnny

[Scott Lurndal]

Post by Johnny Billquist

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

Not sure why you would expect edge triggered any more than level. In
fact, most older hardware use level triggered.
Edge triggered is sometimes problematic/risky, and certainly run the
risk of ending up interrupts not recognized if shared interrupt lines,
or mistakes in software.

Most modern hardware uses edge semantics for interrupt detection.

All PCI and PCI Express message signaled interrupts are edge[*].

The only level sensitive interrupts still used are generally for
legacy devices (e.g. the ARM PL011) or cases where multiple interrupt
sources share the interupt line (e.g. legacy PCI INTA, INTB, INTC
and INTD interrupts).


[*] I am aware of implementations that support message signaled
level sensitive semantics by designating a pair of MSI-X entries;
one of which asserts the level, the second which deasserts the
level. Often used when converting a legacy device to a PCI express device.

[Johnny Billquist]

Post by Scott Lurndal

Post by Johnny Billquist

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

Not sure why you would expect edge triggered any more than level. In
fact, most older hardware use level triggered.
Edge triggered is sometimes problematic/risky, and certainly run the
risk of ending up interrupts not recognized if shared interrupt lines,
or mistakes in software.

Most modern hardware uses edge semantics for interrupt detection.
All PCI and PCI Express message signaled interrupts are edge[*].
The only level sensitive interrupts still used are generally for
legacy devices (e.g. the ARM PL011) or cases where multiple interrupt
sources share the interupt line (e.g. legacy PCI INTA, INTB, INTC
and INTD interrupts).

Note how I said "older hardware" and shared interrupt lines... ;-)

Johnny

[Ahem A Rivet's Shot]

On Fri, 17 Mar 2023 16:07:54 +0100

Post by Johnny Billquist

Post by gareth evans
AAAH!!! The dreaded 8259 Interrupt Controller. Not edge triggered as
one would expect but ISTR level sensitive with edge lockout.

Not sure why you would expect edge triggered any more than level. In
fact, most older hardware use level triggered.
Edge triggered is sometimes problematic/risky, and certainly run the
risk of ending up interrupts not recognized if shared interrupt lines,
or mistakes in software.

Also prone to false triggering in high noise environments with
inadequate decoupling (DAMHIK).

[OldbieOne]

Post by Bob Eager

Post by Ian

Used to work at Vanwell Data Systems in Melksham, Wilts, who provided
Ferranti with the memory-behind-the-pin test eqpt.
ISTR visiting Ferranti in Gem Mill in Oldham.

Now a housing estate. Worked there around 1986, then in Hollinwood,
where they did the ULAs and ASICs.
https://www.jusme.com/perm/gem/gallery/pages/page_0001.html
"There's trouble at t' mill"

I didn't see the original post, but I bought my first computer in 1984.
Had a lot of those ULAs in it. I did a fake service call, because I knew
the first thing the technician would do was to glue heatsinks to all ofg
them - they ran VERY hot.
http://www.computinghistory.org.uk/det/7141/Ferranti-Advance-86/
The comment about compatibility is actually from me!

I don't remember ever seeing, or hearing about them, but I certainly remember WH
Smiths.

This is another new one for me. This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

My first Bri'ish PC experience was the then brand new Apricot XEN-i 386.
Murica had the Apple, Blighty had the Apricot. I kept waiting for the Grape PC
or possibly even a Pentium Pommegranate to make a showing, lol!

--
OldbieOne [TM]
The One Who Tells It Like It is!
Brought to you by RetroPC using
Forte Agent News Reader .99g/32

[Sn!pe]

OldbieOne <***@whitexmasradio.com> wrote:

[...]

Post by OldbieOne
This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

I had one after my IBM PC XT. It was basically a [ditto] in
a pretty case with a better display. Italian make, I think.

[Charlie Gibbs]

Post by Sn!pe
[...]

Post by OldbieOne
This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

I had one after my IBM PC XT. It was basically a [ditto] in
a pretty case with a better display. Italian make, I think.

I think we still have an Olivetti M18 around here somewhere.
This was the "luggable" version. My arms are a couple of inches
longer than they used to be because I had to park several blocks
from the customer site to which I hauled it one day. Even worse,
to fit between the slots, the handle had to be at right angles to
how a normal suitcase handle is oriented. Ouch.

The ancestor of the software I still maintain was born on it.

[OldbieOne]

Post by Charlie Gibbs

Post by Sn!pe
[...]

Post by OldbieOne
This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

I had one after my IBM PC XT. It was basically a [ditto] in
a pretty case with a better display. Italian make, I think.

I think we still have an Olivetti M18 around here somewhere.
This was the "luggable" version. My arms are a couple of inches
longer than they used to be because I had to park several blocks
from the customer site to which I hauled it one day. Even worse,
to fit between the slots, the handle had to be at right angles to
how a normal suitcase handle is oriented. Ouch.

I do remember seeing those luggables, but thankfully never had to carry one, or
any machine in the "luggable" era. My first portable experience was when we had
already moved on to the laptop era - at the time when they were still too heavy
for use on one's lap, but at least had a better form factor.

Post by Charlie Gibbs
The ancestor of the software I still maintain was born on it.

I'm sure it would fire right up, too, after a little work. The only capacitor
blight I've experienced was from late 90's through early 00's mainboards.
Machines made before, or after then, seem to at least function.

Of course, any CMOS battery would be long since dead and may have caused
some board trace destruction, but I doubt the Olivetti had RTC so probably
wouldn't have anything to worry about there.


--
OldbieOne [TM]
The One Who Tells It Like It is!
Brought to you by RetroPC using
Forte Agent News Reader .99g/32

[D.J.]

Post by OldbieOne

Post by Charlie Gibbs

Post by Sn!pe
[...]

Post by OldbieOne
This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

I had one after my IBM PC XT. It was basically a [ditto] in
a pretty case with a better display. Italian make, I think.

I think we still have an Olivetti M18 around here somewhere.
This was the "luggable" version. My arms are a couple of inches
longer than they used to be because I had to park several blocks
from the customer site to which I hauled it one day. Even worse,
to fit between the slots, the handle had to be at right angles to
how a normal suitcase handle is oriented. Ouch.

I do remember seeing those luggables, but thankfully never had to carry one, or
any machine in the "luggable" era. My first portable experience was when we had
already moved on to the laptop era - at the time when they were still too heavy
for use on one's lap, but at least had a better form factor.

Post by Charlie Gibbs
The ancestor of the software I still maintain was born on it.

I'm sure it would fire right up, too, after a little work. The only capacitor
blight I've experienced was from late 90's through early 00's mainboards.
Machines made before, or after then, seem to at least function.

The electrolyte paste info the Chinese hacked from Japan was copied
down wrong. The electrolitic capacitors in the Amiga computers have to
be replaced. How, on youtube.

[OldbieOne]

Post by Sn!pe
[...]

Post by OldbieOne
This was around the time my father brought home
an Olivetti M23 with whopping 20MB hard drive and a "full" 640K RAM.
Aother old, forgotten machine. I don't think it was actually British though.

I had one after my IBM PC XT. It was basically a [ditto] in
a pretty case with a better display. Italian make, I think.

I seem to recall it had an actual RS232 port on-board, unlike Big Blue at the
time, and a color monitor which was kinda impressive in the world of amber or
green at the time. I don't remember what resolution it would render but to my
eye, when VGA was the "new big thing" it didn't look much, if at all, different.


--
OldbieOne [TM]
The One Who Tells It Like It is!
Brought to you by RetroPC using
Forte Agent News Reader .99g/32

[Ahem A Rivet's Shot]

On Thu, 16 Mar 2023 14:27:40 GMT

Post by OldbieOne
This is another new one for me. This was around the time my father
brought home an Olivetti M23 with whopping 20MB hard drive and a "full"
640K RAM. Aother old, forgotten machine. I don't think it was actually
British though.

Olivetti was Italian, but there is a British connection. They
bought Acorn when it collapsed - leading to people in Cambridge singing
"Just Acorn-etti ...".

[OldbieOne]

Post by Ahem A Rivet's Shot
On Thu, 16 Mar 2023 14:27:40 GMT

Post by OldbieOne
This is another new one for me. This was around the time my father
brought home an Olivetti M23 with whopping 20MB hard drive and a "full"
640K RAM. Aother old, forgotten machine. I don't think it was actually
British though.

Olivetti was Italian, but there is a British connection. They
bought Acorn when it collapsed - leading to people in Cambridge singing
"Just Acorn-etti ...".

LOL!!!! Leave it to the Brits to come up with a perfectly apt snark!

--
OldbieOne [TM]
The One Who Tells It Like It is!
Brought to you by RetroPC using
Forte Agent News Reader .99g/32

[Bob Eager]

Post by OldbieOne

Post by Bob Eager

Post by Ian

Used to work at Vanwell Data Systems in Melksham, Wilts, who provided
Ferranti with the memory-behind-the-pin test eqpt.
ISTR visiting Ferranti in Gem Mill in Oldham.

Now a housing estate. Worked there around 1986, then in Hollinwood,
where they did the ULAs and ASICs.
https://www.jusme.com/perm/gem/gallery/pages/page_0001.html
"There's trouble at t' mill"

I didn't see the original post, but I bought my first computer in 1984.
Had a lot of those ULAs in it. I did a fake service call, because I knew
the first thing the technician would do was to glue heatsinks to all ofg
them - they ran VERY hot.
http://www.computinghistory.org.uk/det/7141/Ferranti-Advance-86/
The comment about compatibility is actually from me!

I don't remember ever seeing, or hearing about them, but I certainly
remember WH Smiths.

They weren't heavily marketed. And only a few selected larger WH Smith
branches had them (and didn't, I think, know much about them). I live in
East Kent, and bought mine after driving down to Croydon.

They were bulky, all black plastic; the joke was that they guy's previous
product had been digital watches - in black plastic. Mine worked well
enough, but after fitting a hard disk I had to run it with the top cover
off, or it overheated.

It was eventually replaced with a Taiwanese clone at a third of the price.