Re: Tandem Minicomputers

2019-09-30 Thread Wayne S via cctalk
SII ( system integrators inc, not System Industries Inc) produced a big system 
for Newspaper editing. Could be over 3000 seats. It used Coyote terminals. 
IIRC, i think it ran the Guardian operating system.

Sent from my iPhone

> On Sep 30, 2019, at 09:45, Chuck Guzis via cctalk  
> wrote:
> 
>> On 9/30/19 8:57 AM, Jon Elson via cctalk wrote:
>>> On 09/29/2019 11:46 PM, Jason T via cctalk wrote:
>>> Well I said no more computers I can't lift, but exotic systems keep
>>> finding me.  So today we pulled a Tandem CLX out of a basement, along
>>> with a few boxes of docs, 9-track tapes and random odd and ends:
>>> 
>>> https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8
>>> 
>>> There's such as thing as "so obscure that no one knows/cares about
>>> it".  I've had those before.  Do I have another?  It sure is heavy.
>>> 
>>> -j
>> Tandem was hot stuff back in the 1980's.  These appear to be the last
>> gasp of their technology.
>> Once you had fast networking between processors, the whole Tandem
>> concept became pretty easy to do on a few ordinary processors, without
>> special hardware.  So, their whole reason for being became moot.
> 
> Jimmy Treybig was a master marketeer.  I recall that when his offerings
> debuted, he equip a guest with a pair of diagonal cutters and instruct
> them to pick a wire, any wire and cut it.
> 
> The system would just keep running.
> 
> The old OnSale auction site that featured some really good deals back in
> the 1990s was powered by Tandem.
> 
> https://web.archive.org/web/19980212000633/http://www.onsale.com:80/
> 
> --Chuck
> 
> 


Re: Standard Engineering CAMAC crate controller with LSI-11

2019-09-30 Thread Jon Elson via cctalk

On 09/30/2019 09:17 PM, Adrian Stoness via cctalk wrote:

just a box of boards or a system missing a memoru module?

OK, if you are not familiar with it, CAMAC is a standard for 
data acquisition and control, that has 25-slot powered 
crates.  At the right edge, the last 2 slots are dedicated 
as controller slots.
This was a 3-slot set that could be used as a programmable 
controller for the CAMAC crate.

So, it has 3 CAMAC modules, one of which has the LSI11/2 in it.
But, since the LSI11/2 doesn't have memory on board, and the 
other boards don't seem to have any memory, it seems 
something is missing.  I think there must have been a 4th 
module that communicated by a ribbon cable across the 
front.  It may still be around here...


Jon


Re: HP vintage boards being sold as scrap - WON

2019-09-30 Thread Guy Dunphy via cctalk
Yesterday:
>These will go up on my site at   http://everist.org/pics/pcbs

Then promptly the web hosting server goes down, since this morning of 20190930 
Tue in Australia.
I don't yet know why, or have any estimate of when it will come back up.

Guy





Re: Standard Engineering CAMAC crate controller with LSI-11

2019-09-30 Thread Adrian Stoness via cctalk
just a box of boards or a system missing a memoru module?


On Mon, Sep 30, 2019 at 8:45 PM Jon Elson via cctalk 
wrote:

> I have a 3-board set, a Standar Engineering MIK-11/2, which
> is a controller for a CAMAC crate, with an LSI-11/2 (M7270)
> in it.  It also has a serial port in it.  I thought there
> was a memory that went with it, but I can't seem to find that.
>
> Anybody have any use for this?
>
> Jon
>


Chabin TLC cables

2019-09-30 Thread Jon Elson via cctalk
I have some bundles of Chabin TLC (transmission line cables) 
from a Memorex 3674 disk control.
This is the exact same material used in IBM 360s.  It has 18 
signal conductors and 36 grounds.
They are about 3 feet long, and I have something like 11 of 
them.


Any interest?

Jon


Tek 500 pulser modules

2019-09-30 Thread Jon Elson via cctalk
I have 3 units of the Pulse Instruments PI-451A programmable 
pulse driver.  One has a note that it is bad.  These seem to 
be general purpose pulse generators, capable of +/- 25 V 
into 50 Ohms (but that is just from reading the labels on 
the panel.)


Anybody have a use for them?  If I had a 500 bin I could 
probably use them, but I don't.


Jon


Standard Engineering CAMAC crate controller with LSI-11

2019-09-30 Thread Jon Elson via cctalk
I have a 3-board set, a Standar Engineering MIK-11/2, which 
is a controller for a CAMAC crate, with an LSI-11/2 (M7270) 
in it.  It also has a serial port in it.  I thought there 
was a memory that went with it, but I can't seem to find that.


Anybody have any use for this?

Jon


RAMTEK memory boards FTGH

2019-09-30 Thread Jon Elson via cctalk
I have 20 RAMTEK memory boards from late '80s frame 
buffers.  They each have 20 TMS4161-20NL video RAM chips on 
them.  The RAMTEK part # is ASSY 510857


Anyone have any interest?

Jon


Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread ben via cctalk

On 9/30/2019 2:43 PM, Josh Dersch via cctalk wrote:

See Chapter 3, pg. 249 in the PDF.

- Josh


I only  read THE TABLE OF CONTENTS...
I was not expecting a bunch of files all in one big PDF.
I view all my PDFs on my tablet, the screen size fits better
than the WIDE SCREEN CinemaScope* format of todays PC's.
I'll download it and view it later.
Ben.
*Well heading in that direction,



Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread Josh Dersch via cctalk
On Mon, Sep 30, 2019 at 2:29 PM Noel Chiappa via cctalk <
cctalk@classiccmp.org> wrote:

> > From: Josh Dersch
>
> > descriptions of the PDP-11/45 DELPHI system
> > ...
> > moves on to Algol and LISP
>
> I later became the 'owner' of that PDP-11/45 (our group at LCS traded an
> -11/40, which EECS wanted for their DECSystem-20, for it).
>
> That Algol and LISP were later moved to Unix V6 when the group that had
> done
> DELPHI converted to Unix. I have both - alas, the source for the Algol has
> been lost. :-(
>

Very cool.  Any idea what ultimately happened to that 11/45?  Are the Algol
and LISP available anywhere?

Thanks!
Josh



>
> Noel
>


Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread Paul Koning via cctalk



> On Sep 30, 2019, at 4:39 PM, ben via cctalk  wrote:
> 
> A quick look,  nothing is said about Algol, did you mean Assembler?
> what I see is interesting is that LISP is on a 11. I allways thought
> LISP only ran on BIGGER machines.
> Ben.

From what I heard fairly recently, some early LISP implementations were on 
quite small machines.  One was the PDP-8 (see 
http://www.softwarepreservation.org/projects/LISP/other/) and that same page 
also mentions LISP for the ZEBRA -- a Dutch machine, with 33 bit words.  

I think another was the Electrologica X1, another Dutch machine, like the ZEBRA 
from 1958, with 27 bit words and a base configuration of 4k words of core 
memory.  I may be off; if it wasn't the X1, it would be the X8, from 1968, also 
27 bits but more likely to be seen with 32 k or so of memory.

A lot of things can be done in small amounts of memory if you work at it, 
though admittedly most who have that skill probably also have white hair.  :-)  
For example, the world's first ALGOL compiler -- a full language compiler, not 
a subset -- ran on an EL-X1 with 4k words of memory.  And it was built by 2 
people, in 6 months.  Now admittedly one of those two was E.W.Dijkstra...

paul



Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread Noel Chiappa via cctalk
> From: Josh Dersch

> descriptions of the PDP-11/45 DELPHI system
> ...
> moves on to Algol and LISP

I later became the 'owner' of that PDP-11/45 (our group at LCS traded an
-11/40, which EECS wanted for their DECSystem-20, for it).

That Algol and LISP were later moved to Unix V6 when the group that had done
DELPHI converted to Unix. I have both - alas, the source for the Algol has
been lost. :-(

Noel


Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread Josh Dersch via cctalk
On Mon, Sep 30, 2019 at 1:39 PM ben via cctalk 
wrote:

> On 9/30/2019 12:50 PM, Josh Dersch via cctalk wrote:
> > Thought someone here might find this interesting; I have a binder of
> > materials describing the entire course (descriptions of the PDP-11/45
> > DELPHI system, readings, coursework, quizzes, exams (with answers)) for
> MIT
> > 6.031 "Structure and Interpretation of Computer Languages", 1974.
> >
> >
> http://yahozna.dyndns.org/scratch/scans/mit/MIT%206.031%20Structure%20And%20Interpretation%20Of%20Computer%20Languages%201974.pdf
> >
> >
> > It starts with PDP-11 assembly language, moves on to Algol and LISP and
> is
> > over a thousand pages of material.  Get studying!
> >
> > - Josh
> >
>
> A quick look,  nothing is said about Algol, did you mean Assembler?
> what I see is interesting is that LISP is on a 11. I allways thought
> LISP only ran on BIGGER machines.
> Ben.
>


See Chapter 3, pg. 249 in the PDF.

- Josh


Re: Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread ben via cctalk

On 9/30/2019 12:50 PM, Josh Dersch via cctalk wrote:

Thought someone here might find this interesting; I have a binder of
materials describing the entire course (descriptions of the PDP-11/45
DELPHI system, readings, coursework, quizzes, exams (with answers)) for MIT
6.031 "Structure and Interpretation of Computer Languages", 1974.

http://yahozna.dyndns.org/scratch/scans/mit/MIT%206.031%20Structure%20And%20Interpretation%20Of%20Computer%20Languages%201974.pdf


It starts with PDP-11 assembly language, moves on to Algol and LISP and is
over a thousand pages of material.  Get studying!

- Josh



A quick look,  nothing is said about Algol, did you mean Assembler?
what I see is interesting is that LISP is on a 11. I allways thought
LISP only ran on BIGGER machines.
Ben.



Re: Tandem Minicomputers

2019-09-30 Thread jim stephens via cctalk




On 9/30/2019 9:39 AM, Paul Berger via cctalk wrote:


On 2019-09-30 1:27 p.m., Diane Bruce via cctalk wrote:

On Mon, Sep 30, 2019 at 10:57:25AM -0500, Jon Elson via cctalk wrote:

On 09/29/2019 11:46 PM, Jason T via cctalk wrote:

Well I said no more computers I can't lift, but exotic systems keep
finding me.  So today we pulled a Tandem CLX out of a basement, along
with a few boxes of docs, 9-track tapes and random odd and ends:

https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8

There's such as thing as "so obscure that no one knows/cares about
it".  I've had those before.  Do I have another?  It sure is heavy.

-j


Tandem was hot stuff back in the 1980's.  These appear to be
the last gasp of their technology.

I remember hearing that the credit/debit card database we were talking
to was a Royal Bank Tandem in Toronto when I was first involved
with Point of Sale protocols.


All of the Canadian banks used Tandem systems to front end the 
Interact it was part of the initial plan for that system.


Paul.




Once you had fast networking between processors, the whole
Tandem concept became pretty easy to do on a few ordinary
processors, without special hardware.  So, their whole
reason for being became moot.

Indeed.


Yes I suspect anything that HP may flog as nonstop these days would be 
a high availability cluster which would probably include shared 
storage.  SAN has made shared storage trivial and even allows you to 
spread the nodes in the cluster geographically. When I first 
encountered HA cluster the shared storage was parallel SCSI with an 
initiator in each system.  The first that aloowed you to spread the 
nodes enough to have them in different building used IBM's SSA, a 
predecessor of SAN, to connect the storage.



In the early 2000s, I visited the HP site which had the developers for 
the Tandem product, VMS HP/UX.  They were using our ICE for development 
of the Itanium versions of VMS and for HP/UX.  I don't think there was 
any plan to support the Nonstop in such a fashion, but the developers 
were all there @ the same Cupertino site.


As to developing something to do a useful version of TAL and Nonstop w/o 
any specialized hardware, I don't think so.  We had a system which was 
used for a port of the Ultimate / Pick system to Tandem, which was 
eventually shutdown and abandoned, but going the other way, to run TAL 
in any useful manner without the checkpointing would be pretty much useless.


And Stratus is still there with financial industry using it heavily.

The Tandem method used a checkpointing comparison method which had to 
use TAL as the basic platform.  Periodically comparisons would be 
performed, but there was more than a little computing between comparison 
checkpoints.  And the voting out of a failed unit was done by hardware.


Stratus used a bus version. so they could have hardware pulled and their 
loss was in a few microseconds.  For the realtime trader guys Stratus 
always won over Tandem.  but where there was no need, either platform 
was excellent as pointed out for recovering and continuing after a failure.


Also, I don't know if Tandem had it, but Stratus had serial connections 
and gizmos which would allow Session protection on a terminal.  you 
could cut one or the other connection from your computer (required 2) 
and not miss a beat on a terminal.  If the terminal went black though 
there were not as many options.


thanks
Jim


Jon

Diane

Paul.






Course materials for MIT 6.031 (c. 1974) scanned

2019-09-30 Thread Josh Dersch via cctalk
Thought someone here might find this interesting; I have a binder of
materials describing the entire course (descriptions of the PDP-11/45
DELPHI system, readings, coursework, quizzes, exams (with answers)) for MIT
6.031 "Structure and Interpretation of Computer Languages", 1974.

http://yahozna.dyndns.org/scratch/scans/mit/MIT%206.031%20Structure%20And%20Interpretation%20Of%20Computer%20Languages%201974.pdf


It starts with PDP-11 assembly language, moves on to Algol and LISP and is
over a thousand pages of material.  Get studying!

- Josh


Re: Tandem Minicomputers

2019-09-30 Thread Chuck Guzis via cctalk
On 9/30/19 10:44 AM, Wayne S wrote:
> SII ( system integrators inc, not System Industries Inc) produced a big 
> system for Newspaper editing. Could be over 3000 seats. It used Coyote 
> terminals. IIRC, i think it ran the Guardian operating system.
> 

At the time, I wondered if the name had been swiped from "The Forbin
Project" film.

--Chuck



Re: Tandem Minicomputers

2019-09-30 Thread Chuck Guzis via cctalk
On 9/30/19 8:57 AM, Jon Elson via cctalk wrote:
> On 09/29/2019 11:46 PM, Jason T via cctalk wrote:
>> Well I said no more computers I can't lift, but exotic systems keep
>> finding me.  So today we pulled a Tandem CLX out of a basement, along
>> with a few boxes of docs, 9-track tapes and random odd and ends:
>>
>> https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8
>>
>> There's such as thing as "so obscure that no one knows/cares about
>> it".  I've had those before.  Do I have another?  It sure is heavy.
>>
>> -j
>>
> Tandem was hot stuff back in the 1980's.  These appear to be the last
> gasp of their technology.
> Once you had fast networking between processors, the whole Tandem
> concept became pretty easy to do on a few ordinary processors, without
> special hardware.  So, their whole reason for being became moot.

Jimmy Treybig was a master marketeer.  I recall that when his offerings
debuted, he equip a guest with a pair of diagonal cutters and instruct
them to pick a wire, any wire and cut it.

The system would just keep running.

The old OnSale auction site that featured some really good deals back in
the 1990s was powered by Tandem.

https://web.archive.org/web/19980212000633/http://www.onsale.com:80/

--Chuck




Re: Tandem Minicomputers

2019-09-30 Thread Paul Berger via cctalk



On 2019-09-30 1:27 p.m., Diane Bruce via cctalk wrote:

On Mon, Sep 30, 2019 at 10:57:25AM -0500, Jon Elson via cctalk wrote:

On 09/29/2019 11:46 PM, Jason T via cctalk wrote:

Well I said no more computers I can't lift, but exotic systems keep
finding me.  So today we pulled a Tandem CLX out of a basement, along
with a few boxes of docs, 9-track tapes and random odd and ends:

https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8

There's such as thing as "so obscure that no one knows/cares about
it".  I've had those before.  Do I have another?  It sure is heavy.

-j


Tandem was hot stuff back in the 1980's.  These appear to be
the last gasp of their technology.

I remember hearing that the credit/debit card database we were talking
to was a Royal Bank Tandem in Toronto when I was first involved
with Point of Sale protocols.


All of the Canadian banks used Tandem systems to front end the Interact 
it was part of the initial plan for that system.


Paul.




Once you had fast networking between processors, the whole
Tandem concept became pretty easy to do on a few ordinary
processors, without special hardware.  So, their whole
reason for being became moot.

Indeed.


Yes I suspect anything that HP may flog as nonstop these days would be a 
high availability cluster which would probably include shared storage.  
SAN has made shared storage trivial and even allows you to spread the 
nodes in the cluster geographically. When I first encountered HA cluster 
the shared storage was parallel SCSI with an initiator in each system.  
The first that aloowed you to spread the nodes enough to have them in 
different building used IBM's SSA, a predecessor of SAN, to connect the 
storage.




Jon

Diane

Paul.


Re: Tandem Minicomputers

2019-09-30 Thread Diane Bruce via cctalk
On Mon, Sep 30, 2019 at 10:57:25AM -0500, Jon Elson via cctalk wrote:
> On 09/29/2019 11:46 PM, Jason T via cctalk wrote:
> > Well I said no more computers I can't lift, but exotic systems keep
> > finding me.  So today we pulled a Tandem CLX out of a basement, along
> > with a few boxes of docs, 9-track tapes and random odd and ends:
> >
> > https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8
> >
> > There's such as thing as "so obscure that no one knows/cares about
> > it".  I've had those before.  Do I have another?  It sure is heavy.
> >
> > -j
> >
> Tandem was hot stuff back in the 1980's.  These appear to be 
> the last gasp of their technology.

I remember hearing that the credit/debit card database we were talking
to was a Royal Bank Tandem in Toronto when I was first involved
with Point of Sale protocols.

> Once you had fast networking between processors, the whole 
> Tandem concept became pretty easy to do on a few ordinary 
> processors, without special hardware.  So, their whole 
> reason for being became moot.

Indeed.

> 
> Jon

Diane
-- 
- d...@freebsd.org d...@db.net http://www.db.net/~db


Re: Tandem Minicomputers

2019-09-30 Thread Grant Taylor via cctalk

On 9/29/19 10:46 PM, Jason T via cctalk wrote:
There's such as thing as "so obscure that no one knows/cares about it". 
I've had those before.  Do I have another?  It sure is heavy.


I think HP(E) acquired (what as left of) Tandem.  I think that HP(E) 
still sells and supports NonStop equipment today.


I'm guessing that NonStop is about as common as IBM mainframes.  Maybe 
slightly less so.


Common? No.
So obscure that no one knows / cars about it?  Nope.



--
Grant. . . .
unix || die


Re: Tandem Minicomputers

2019-09-30 Thread Stefan Skoglund via cctalk
sön 2019-09-29 klockan 23:46 -0500 skrev Jason T via cctalk:
> Well I said no more computers I can't lift, but exotic systems keep
> finding me.  So today we pulled a Tandem CLX out of a basement, along
> with a few boxes of docs, 9-track tapes and random odd and ends:
> 
> https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8
> 
> There's such as thing as "so obscure that no one knows/cares about
> it".  I've had those before.  Do I have another?  It sure is heavy.
> 
> -j

Where in the world ?
did you have the passwords and so on so you didn't need to crack it
(which i hope is NOT easy.)



Re: Tandem Minicomputers

2019-09-30 Thread Jon Elson via cctalk

On 09/29/2019 11:46 PM, Jason T via cctalk wrote:

Well I said no more computers I can't lift, but exotic systems keep
finding me.  So today we pulled a Tandem CLX out of a basement, along
with a few boxes of docs, 9-track tapes and random odd and ends:

https://photos.app.goo.gl/m2N7RKN3JXcmVTUC8

There's such as thing as "so obscure that no one knows/cares about
it".  I've had those before.  Do I have another?  It sure is heavy.

-j

Tandem was hot stuff back in the 1980's.  These appear to be 
the last gasp of their technology.
Once you had fast networking between processors, the whole 
Tandem concept became pretty easy to do on a few ordinary 
processors, without special hardware.  So, their whole 
reason for being became moot.


Jon


Re: IBM 5110 transformer?

2019-09-30 Thread Mattis Lind via cctalk
Really good point. Thanks! I will check with my friend.

/Mattis

Den mån 30 sep. 2019 kl 16:46 skrev Paul Berger via cctalk <
cctalk@classiccmp.org>:

>
> On 2019-09-30 3:44 a.m., Mattis Lind via cctalk wrote:
> > A friend of mine is trying to repair a IBM 5110. He is convinced that the
> > transformer is bad.
> >
> > Anyone knows the spec of the transformer?
> >
> > Someone that has a spare?
> >
> > Looking into the tech documentation tell me that the machine requires +/-
> > 5V and +/- 12 V and also +8.5V
> >
> > A very rough guesstimate based on the number of wires from the PSU to the
> > backplane would give 20A 5V, 4A +12V, 4A +8.5V, 1A -5V and 1A -12V.
> >
> > Anyone with a better guess?
> >
> > /Mattis
>
> Does this transformer has a resonant capacitor on it?  They where very
> common on IBM power supplies from that time period and if the capacitor
> goes short you will not get any output from the transformer.  This is a
> lesson I learned as a green service rep 40 years ago after lugging in a
> very heavy transformer for a machine.
>
> Paul.
>
>


Re: IBM 5110 transformer?

2019-09-30 Thread Paul Berger via cctalk



On 2019-09-30 3:44 a.m., Mattis Lind via cctalk wrote:

A friend of mine is trying to repair a IBM 5110. He is convinced that the
transformer is bad.

Anyone knows the spec of the transformer?

Someone that has a spare?

Looking into the tech documentation tell me that the machine requires +/-
5V and +/- 12 V and also +8.5V

A very rough guesstimate based on the number of wires from the PSU to the
backplane would give 20A 5V, 4A +12V, 4A +8.5V, 1A -5V and 1A -12V.

Anyone with a better guess?

/Mattis


Does this transformer has a resonant capacitor on it?  They where very 
common on IBM power supplies from that time period and if the capacitor 
goes short you will not get any output from the transformer.  This is a 
lesson I learned as a green service rep 40 years ago after lugging in a 
very heavy transformer for a machine.


Paul.



Re: IBM 5110 transformer?

2019-09-30 Thread Mattis Lind via cctalk
Den mån 30 sep. 2019 kl 15:37 skrev dwight :

> Is it a switcher or an AC?
>

Good old transfomer. He could probably find something suitable if the specs
were known. Or replace it with switchers at least for testing the machine.
But still it would be great to know the specs.


> A custom transformer is going to be quite expensive. About the only way is
> to cannibalize another machine.
> Dwight
>

/Mattis


>
> --
> *From:* cctalk  on behalf of Mattis Lind
> via cctalk 
> *Sent:* Sunday, September 29, 2019 11:44 PM
> *To:* General Discussion: On-Topic and Off-Topic Posts <
> cctalk@classiccmp.org>
> *Subject:* IBM 5110 transformer?
>
> A friend of mine is trying to repair a IBM 5110. He is convinced that the
> transformer is bad.
>
> Anyone knows the spec of the transformer?
>
> Someone that has a spare?
>
> Looking into the tech documentation tell me that the machine requires +/-
> 5V and +/- 12 V and also +8.5V
>
> A very rough guesstimate based on the number of wires from the PSU to the
> backplane would give 20A 5V, 4A +12V, 4A +8.5V, 1A -5V and 1A -12V.
>
> Anyone with a better guess?
>
> /Mattis
>


Re: IBM 5110 transformer?

2019-09-30 Thread dwight via cctalk
Is it a switcher or an AC?
A custom transformer is going to be quite expensive. About the only way is to 
cannibalize another machine.
Dwight


From: cctalk  on behalf of Mattis Lind via 
cctalk 
Sent: Sunday, September 29, 2019 11:44 PM
To: General Discussion: On-Topic and Off-Topic Posts 
Subject: IBM 5110 transformer?

A friend of mine is trying to repair a IBM 5110. He is convinced that the
transformer is bad.

Anyone knows the spec of the transformer?

Someone that has a spare?

Looking into the tech documentation tell me that the machine requires +/-
5V and +/- 12 V and also +8.5V

A very rough guesstimate based on the number of wires from the PSU to the
backplane would give 20A 5V, 4A +12V, 4A +8.5V, 1A -5V and 1A -12V.

Anyone with a better guess?

/Mattis