Bbc-micro Archive

List Statistics

  • Total Threads: 360
  • Total Posts: 2182
  #1  
11-12-2011 03:04 PM
Bbc-micro member admin is online now
User
 


I've just acquired a 32016 Second Processor, and I've
noticed that this one has a 32082 MMU on board.

Anyone know in what contexts this provides any benefits ?
Because no pictures on the net seem to show one (it's just
an empty space, not even a socket). So I suppose it is
(optionally) useful when running some version of Panos
(I have both 1.1 and 1.4).


John Kortink

--

Email :
Homepage : http://www.inter.nl.net/users/J.Kortink


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #2  
11-12-2011 09:08 PM
Bbc-micro member admin is online now
User
 

> I've just acquired a 32016 Second Processor, and I've
> noticed that this one has a 32082 MMU on board.
>
> Anyone know in what contexts this provides any benefits ?
> Because no pictures on the net seem to show one (it's just
> an empty space, not even a socket). So I suppose it is
> (optionally) useful when running some version of Panos
> (I have both 1.1 and 1.4).
>
Is this a 1 Meg or a 4 Meg board? I'd be interested to see a picture, since
I have the chip, (oddly, it came in with some BBC scrap), and a 1 Meg 32016
board I re-cased, (it came from an earlier Cambridge workstation that
someone had broken for spares)..

I take it that the MMU is fitted as IC2. Is IC16 fitted too? and are there
more than 2 links fitted on the stack to the left of where IC 16 would go?

Mine is fitted with 32 x 256 bit RAM I've never seen a circuit diagram, and
would be interested to know if it was possible to upgrade without the much
bigger board, (there seem to be quite a few RAM links). I have saved quite a
bit of DRAM from that era for repairs and rebuilds, though my HP 9823 repair
has eaten quite a lot of the 16 K RAMs.

Hope it all works; very few 32016's seem to have survived.

Mike



_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #3  
11-12-2011 09:50 PM
Bbc-micro member admin is online now
User
 


On Sun, 11 Dec 2011 21:08:58 -0000, Mike Pepper <>
wrote:

>> I've just acquired a 32016 Second Processor, and I've
>> noticed that this one has a 32082 MMU on board.
>>
>> Anyone know in what contexts this provides any benefits ?
>> Because no pictures on the net seem to show one (it's just
>> an empty space, not even a socket). So I suppose it is
>> (optionally) useful when running some version of Panos
>> (I have both 1.1 and 1.4).
>>
>Is this a 1 Meg or a 4 Meg board?

It's the actual cheese wedge, 1 MB.

>I'd be interested to see a picture

I've just cleaned it up, and I'll be making pictures very
soon (do remind me if it takes too long).

>, since I have the chip, (oddly, it came in with some BBC
>scrap), and a 1 Meg 32016 board I re-cased, (it came from
>an earlier Cambridge workstation that someone had broken
>for spares)..
>
>I take it that the MMU is fitted as IC2.

Yes.

>Is IC16 fitted too?

No.

>and are there more than 2 links fitted on the stack to
>the left of where IC 16 would go?

I presume you mean the area marked LK5 'north' of IC16.
Then no. 2 links at 'H' and 'G', nothing at 'A' through
'F'.

>Mine is fitted with 32 x 256 bit RAM I've never seen a
>circuit diagram

I haven't either. It's clearly different from the ACW
board (the schematic of which actually is available
on the net).

>, and would be interested to know if it was possible to
>upgrade without the much bigger board, (there seem to be
>quite a few RAM links). I have saved quite a bit of DRAM
>from that era for repairs and rebuilds, though my HP 9823
>repair has eaten quite a lot of the 16 K RAMs.

It's an interesting thought. Mine has 32 x 256 (x 1) too.

>Hope it all works; very few 32016's seem to have survived.

It seems to work perfectly. And it's in perfect condition.
Full set of manuals and original discs too. I don't think
it has been used much.


John Kortink

--

Email :
Homepage : http://www.inter.nl.net/users/J.Kortink


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #4  
11-12-2011 11:12 PM
Bbc-micro member admin is online now
User
 

John Kortink wrote:
> I've just acquired a 32016 Second Processor, and I've
> noticed that this one has a 32082 MMU on board.

Hmm, a very rare beast indeed. My memory's telling me that the socket (or
solder pads) next to the CPU were intended for a 32081 FPU, not an MMU (and
indeed a have a disk claiming to be for FPU soak tests). I just had a quick
trawl of what scanned docs I have here and unfortunately can't see anything
that says one way or the other, but I'll try and take a better look when I
get a chance.

Is this a 256K/1MB board (the "small" 32016) or a 1MB/4MB board (the
"large" 32016)? (I've only ever seen solder pads on the small boards,
whereas the large boards typically have sockets fitted - but not populated
with a chip; that's from a sample of maybe 20 small boards and 10 large ones)

There did exist an Acorn/Logica project to get Xenix running on the ACW; it
wouldn't surprise me if that wouldn't require use of a 32082 MMU (but I
don't know how far that project progressed - maybe it was axed long before
they really did any serious playing with the hardware!)

I've got a service manual for the ACW, currently overseas - it'll go on the
scanning pile once it's here with me, but if I remember right that has
schematics for the large board only.

cheers

Jules

_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #5  
11-12-2011 11:39 PM
Bbc-micro member admin is online now
User
 

John Kortink wrote:
> Full set of manuals and original discs too. I don't think
> it has been used much.

I've got PanOS disks for 0.something, 1.0 and 1.2. You mentioned
you've got 1.1 and 1.4. I'd be interested in a copy of the disks
for archiving/playing with.

Jules Richardson wrote:
> Hmm, a very rare beast indeed. My memory's telling me that the socket
> (or
> solder pads) next to the CPU were intended for a 32081 FPU, not an
> MMU (and

The schematic at http://mdfs.net/Info/Comp/BBC/Circuits/32016copro.gif
[1]
shows a 32082 MMU and a 32081 FPU (as well as the 32016 CPU).

[1]At some point I'll scan in a higher resolution version, this one is
very difficult to read.

My cheese-wedge 32016s http://pics.mdfs.net/2007/10/071006.htm
says IC2 next to the empty solder pads, which the schematic
says is the 32081 FPU, and also has an 32082 socketed in IC20.

The 4B big 32016 at
http://mdfs.net/Software/Tube/32016/pics/4MCoPro.jpg
also appears to have a (or two) MMUs and an empty 40-pin space
for a FPU.

My feeling is that having an MMU is the normal configuration, it's
having a FPU that is unusual.

> I've got a service manual for the ACW, currently overseas - it'll go
> on the

http://mdfs.net/Docs/Books/32016CoPro/

JGH


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #6  
12-12-2011 05:37 AM
Bbc-micro member admin is online now
User
 

> John Kortink wrote:
> > Full set of manuals and original discs too. I don't think
> > it has been used much.
>
> I've got PanOS disks for 0.something, 1.0 and 1.2. You mentioned
> you've got 1.1 and 1.4. I'd be interested in a copy of the disks
> for archiving/playing with.
>
> Jules Richardson wrote:
> > Hmm, a very rare beast indeed. My memory's telling me that the socket
> > (or
> > solder pads) next to the CPU were intended for a 32081 FPU, not an
> > MMU (and
>
> The schematic at http://mdfs.net/Info/Comp/BBC/Circuits/32016copro.gif
> [1]
> shows a 32082 MMU and a 32081 FPU (as well as the 32016 CPU).
>
> [1]At some point I'll scan in a higher resolution version, this one is
> very difficult to read.
>
> My cheese-wedge 32016s http://pics.mdfs.net/2007/10/071006.htm
> says IC2 next to the empty solder pads, which the schematic
> says is the 32081 FPU, and also has an 32082 socketed in IC20.
>
> The 4B big 32016 at
> http://mdfs.net/Software/Tube/32016/pics/4MCoPro.jpg
> also appears to have a (or two) MMUs and an empty 40-pin space
> for a FPU.
>
> My feeling is that having an MMU is the normal configuration, it's
> having a FPU that is unusual.
>
I think it's the other way round. Mine has the FPU, (I just looked), but no
MMU. The 32081 is a 24 pin device, the MMU is the 40 pin unit.

http://www.cpu-world.com/info/Pinouts/32081.html

http://www.secondvalleysoftware.com/hardware/ns32000/index.html

If anyone kas an Athens account, here is a paper here - but the download
price is robbery!

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=4089599

@JGH: I would really like to have a high res scan of that diagram, as you
say, the present scan is quite hard to read. From what I can read, IC16 may
be another address demultiplexer, which might be the key to an upgrade from
1 Meg.

According to the ACW service manual mentioned, the two LKS links installed
were to indicate FPU fitted and MMU fitted. The rest are marked 'reserved'
and not soldered in.

With the MMU fitted, ML1 and ML links should be open.

Sedtion 5.6.2 of the service manual,states "MMU not yet available", which,
given the stories around the 032 range, seems that National had not then
released it for production. .
I remember being in my tech scrap dealer about 15 years ago, and seeing a
stack of trays of the bigger 32032 processor about to go off for gold scrap.
They were already sold, otherwise I'd have rescued a couple.

My saddest find related to the 32016 was made when peeling a label off a
recycled 5.25" floppy to find another label saying somehting like "Scheme
for 32016 Co Processor, Copyright The Rand Corporation". Sadly the disk
contained nothing but a few old games.

I beleive that ther was a BCPL too, as well as SPICE. Never seen any copies
though, and the Cambridge LISP always seems to crash on mine, though that
could be because I'm trying it out rather blindly.

Mike



_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #7  
12-12-2011 11:38 AM
Bbc-micro member admin is online now
User
 


On Sun, 11 Dec 2011 17:12:00 -0600, Jules Richardson
<> wrote:

>John Kortink wrote:
>> I've just acquired a 32016 Second Processor, and I've
>> noticed that this one has a 32082 MMU on board.
>
>Hmm, a very rare beast indeed. My memory's telling me that the socket (or
>solder pads) next to the CPU were intended for a 32081 FPU, not an MMU (and
>indeed a have a disk claiming to be for FPU soak tests).

As far as I have been able to determine, an FPU is
standard. And it's not the one next to the CPU. The
FPU has 24 pins, and is 'southwest' of the CPU.

>I just had a quick
>trawl of what scanned docs I have here and unfortunately can't see anything
>that says one way or the other, but I'll try and take a better look when I
>get a chance.
>
>Is this a 256K/1MB board (the "small" 32016) or a 1MB/4MB board (the
>"large" 32016)?

Small. It's the actual cheese wedge.

>(I've only ever seen solder pads on the small boards,
>whereas the large boards typically have sockets fitted - but not populated
>with a chip; that's from a sample of maybe 20 small boards and 10 large ones)

The socket on this one is definitely (visibly) an
upgrade. So it stands to reason that the 32082 was,
too. I'm not convinced the MMU does anything right
now. At least it stays a lot cooler than the other
chips.

There's also a link just east of it that says 'LK4
MMU'. I suppose that may be an enable, but it's not
set, which would support the theory that the MMU is
not actually enabled.

>There did exist an Acorn/Logica project to get Xenix running on the ACW; it
>wouldn't surprise me if that wouldn't require use of a 32082 MMU (but I
>don't know how far that project progressed - maybe it was axed long before
>they really did any serious playing with the hardware!)
>
>I've got a service manual for the ACW, currently overseas - it'll go on the
>scanning pile once it's here with me, but if I remember right that has
>schematics for the large board only.


John Kortink

--

Email :
Homepage : http://www.inter.nl.net/users/J.Kortink


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #8  
12-12-2011 11:43 AM
Bbc-micro member admin is online now
User
 


On Sun, 11 Dec 2011 23:39:26 +0000, rote:

>John Kortink wrote:
>> Full set of manuals and original discs too. I don't think
>> it has been used much.
>
>I've got PanOS disks for 0.something, 1.0 and 1.2. You mentioned
>you've got 1.1 and 1.4. I'd be interested in a copy of the disks
>for archiving/playing with.

I've already imaged them for archiving. Just tell me
where to plonk them. Or else I can put them up for
temporary download on my web server.

>[...]
>
>My cheese-wedge 32016s http://pics.mdfs.net/2007/10/071006.htm
>says IC2 next to the empty solder pads, which the schematic
>says is the 32081 FPU, and also has an 32082 socketed in IC20.

Yes, but the schematic is for the ACW board. The
cheese wedge board seems significantly different.

>[...]


John Kortink

--

Email :
Homepage : http://www.inter.nl.net/users/J.Kortink


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #9  
12-12-2011 01:51 PM
Bbc-micro member admin is online now
User
 

John Kortink wrote:
> On Sun, 11 Dec 2011 17:12:00 -0600, Jules Richardson
> <> wrote:
>
>> John Kortink wrote:
>>> I've just acquired a 32016 Second Processor, and I've
>>> noticed that this one has a 32082 MMU on board.
>> Hmm, a very rare beast indeed. My memory's telling me that the socket (or
>> solder pads) next to the CPU were intended for a 32081 FPU, not an MMU (and
>> indeed a have a disk claiming to be for FPU soak tests).
>
> As far as I have been able to determine, an FPU is
> standard. And it's not the one next to the CPU. The
> FPU has 24 pins, and is 'southwest' of the CPU.

Argh, yes, you're right. I had it in my head that Acorn quoted the FPU as
being an optional part, and so assumed that its logical spot was IC2 - but
it does seem that it's IC20 on the 'small' copro, and as far as I can
remember, every single one of those that I've seen has had a chip fitted in
that location.

>> I just had a quick
>> trawl of what scanned docs I have here and unfortunately can't see anything
>> that says one way or the other, but I'll try and take a better look when I
>> get a chance.
>>
>> Is this a 256K/1MB board (the "small" 32016) or a 1MB/4MB board (the
>> "large" 32016)?
>
> Small. It's the actual cheese wedge.

Aha, I have seen them fitted to a couple of ACWs too, although the vast
majority of them got the 'large' board (and every single ACW I've seen has
been different in some way; there's no concept of a standard machine!).

Purely for curiosity's sake, what speed is the CPU on your board? I've seen
both 6MHz parts and 8MHz parts fitted in the cheese wedge cases. The
'large' boards in the ACW seem to mostly have been 10MHz (and I can't
recall off the top of my head what the 1MB ACW that I kept has -
unfortunately I can't quite read the top of the chip on my photos)

>> (I've only ever seen solder pads on the small boards,
>> whereas the large boards typically have sockets fitted - but not populated
>> with a chip; that's from a sample of maybe 20 small boards and 10 large ones)
>
> The socket on this one is definitely (visibly) an
> upgrade. So it stands to reason that the 32082 was,
> too. I'm not convinced the MMU does anything right
> now. At least it stays a lot cooler than the other
> chips.

Yes, I've never seen a socket fitted in IC2's spot on the small boards,
just solder pads.

> There's also a link just east of it that says 'LK4
> MMU'. I suppose that may be an enable, but it's not
> set, which would support the theory that the MMU is
> not actually enabled.

No, that's right I think... I've got photos here of five different small
boards and they all have no socket, no MMU, and LK4 is set - which suggests
that it needs to be removed to enable it.

Given that the copro runs fine without an MMU, I expect it needs extra
software or firmware to make an MMU-equipped one "do something", i.e
there's no code in standard Pandora or PANOS to support it. Are you able to
image the ROMs? All the small boards that I've been able to check ROMs on
have exactly the same copy of Pandora, so it might be worth checking just
to see if yours has something unusual.

I take it there's nothing potentially-useful that "came with" the copro?
Floppies, hard disk, tape etc. that might contain something relevant?

cheers

Jules

_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)

  #10  
12-12-2011 02:41 PM
Bbc-micro member admin is online now
User
 


On Mon, 12 Dec 2011 07:51:32 -0600, Jules Richardson
<> wrote:

>[...]
>
>Purely for curiosity's sake, what speed is the CPU on your board?

All the chipset chips are -6, but the MMU is -10.

>I've seen both 6MHz parts and 8MHz parts fitted in the
>cheese wedge cases. The 'large' boards in the ACW seem
>to mostly have been 10MHz.

Seeing as the ACW runs at 10 (and the cheese wedge at 6)
perhaps not surprising. ;-)

>[...]
>
>> There's also a link just east of it that says 'LK4
>> MMU'. I suppose that may be an enable, but it's not
>> set, which would support the theory that the MMU is
>> not actually enabled.
>
>No, that's right I think... I've got photos here of five different
>small boards and they all have no socket, no MMU, and LK4 is set -
>which suggests that it needs to be removed to enable it.

Interesting. Then it may be doing something after all
(just not very vigourously ...). I may have to try
setting the link on my next session.

>Given that the copro runs fine without an MMU, I expect it
>needs extra software or firmware to make an MMU-equipped one
>"do something", i.e there's no code in standard Pandora or
>PANOS to support it.

Or perhaps the MMU can be detected and used only if it's
there. Or maybe only the later versions of Panos support
or require an MMU.

>Are you able to image the ROMs?

Sure. But they're marked 'Pandora 2.00', so they're probably
the same as the already known ones.

>All the small boards that I've been able to check ROMs on
>have exactly the same copy of Pandora, so it might be worth
>checking just to see if yours has something unusual.

Doesn't seem that way. Panos 1.4 however, may be.

>I take it there's nothing potentially-useful that "came with"
>the copro? Floppies, hard disk, tape etc. that might contain
>something relevant?

No. Cheese wedge, two sets of floppies (Panos 1.1 original
and a 1.4 copy), and the manuals.


John Kortink

--

Email :
Homepage : http://www.inter.nl.net/users/J.Kortink


_______________________________________________
bbc-micro mailing list
bbc-
http://lists.cloud9.co.uk/mailman/listinfo/bbc-micro
)





NewsArc Lists  |  Culture Pages   |  Computing Archive  |  Media-Pages
Link to this page on your blog or website by copying the HTML code below and pasting it into your site: