Date   : Sat, 25 Oct 2008 02:22:39 +0100
From   : pete@... (Pete Turnbull)
Subject: A500 development ROMs

On 25/10/2008 00:21, Jonathan Graham Harston wrote:

> The A500 machine with Tube podule. The machine refered to in Acorn
> history as "we carefully disconnected the umbilical, rebooted with
> fingers crossed, and it worked!".
>  
> Maybe the two have the same name because they are the same thing -
> an Archimedes motherboard with ARM, VIDC, MEMC and socket for Tube
> podule. The A500 machine being inside a metal case. All the photos
> I've seen suggest this.

It seems not; Chris has photos of an A500 2nd Processor which is 
different from an A500 machine's PCB.

> That's the bunny ? I'd fogotten the Tube interface was on a
> seperate podule. Any chance of a ROM dump?

Er, no, because it doesn't have one.  Now look what you've done -- made 
me go and dig out boxes that haven't seen daylight in 15 years.  I've 
found my Watford Digitiser podule with Mike's name on it, two varieties 
of HCCS Vision digitisers, a Wild Vision Chromalock podule, and several 
serial, multi-I/O, and network podules.  I always get distracted by 
re-discoveries when I do that...

Anyway, the Tube podule is as I remembered -- just a connector at each 
end, a Tube ULA, three buffer chips, a resistor and some capacitors.

> Pete Turnbull wrote:
>> Yes.  IIRC the last Springboard was sold in early 1897, about 4-5 years
>> before PCI (released 1992).  I expect Jonathan just meant "PC expansion
>> bus" and wrote "PCI" because that's what came to mind, being most common
>  
> Yes, thinko caused by not bothering to go downstairs and read the
> Springboard manual.

Goes with my typo for the year, which Sprow pointed out.  Amazing that 
we had 2-micron RISC processors only a generation after Babbage ;-)

>> At first, the A500s were used as second processors connected to a Beeb,
>> not to give a Beeb a second processor to play with, but to give the A500
>> (a complete but at that stage dumb computer) I/O capability while its
>  
> That was my understanding. Effectively a full hardware Archimedes
> with just enough on-board firmware to use a Beeb for I/O so that
> native firmware could be written.

Although it could be that the original A500 2P was a pretty minimal 
early system, little more than an ARM development system, and the later 
A500 machine was really a more advanced prototype Archimedes.  The 
boards look very similar, and they seem to have the same amount of ROM 
capacity (512KB).  Chris's A500 2P circuit diagram is dated December 
1985, while my Tube podule (without which the A500 machine would be 
useless) is date coded August 1986, and it's an Issue A.

> The Mike Harrison I knew at Watford Electronics in 1990 seemed to
> be in his late-20s-ish. I dunno, he was just generically "older
> than me" :)

He looked younger to me in 1987 :-)

-- 
Pete                                           Peter Turnbull
                                               Network Manager
                                               University of York