ethan.dicks at gmail.com
Sun Sep 17 19:27:58 CDT 2006
On 9/18/06, aliensrcooluk at yahoo.co.uk <aliensrcooluk at yahoo.co.uk> wrote:
> I got some 68000 DIP chips for �3.00 off of www.vint
> agecomputermarketplace.com the other week. I finally
> got around to checking that they are of use to me (
> by comparing them with notes I have on the internals
> of my old A600 Rev 1.1) and discovered that they ma
> y not be of use to me! :(
Not of use how? I've never owned an A600 (I have just about every
other model)... do they have PLCC (square-package) CPUs?
Were you trying to aquire these chips for spares or for an upgrade?
> Here's the low-down on them:
> 2 C91E DIP chip (rectangle with pins on long sides) x 2
Sounds like a Motorola CPU that can run as fast at 10MHz (they marked
them as high as 12MHz, but I think those can be clocked at 16MHz).
> S (large logo S) SCN68000CAN64
> 2208N19 DIP chip x1
Sounds like a Signetics or similar licensed "clone" 68000 made in late
1990 (week 35). At one point, there were several makers of the CPUs,
> Any idea's what computer they are from/for?
Um... pretty much anything that would use a 68000 at 10MHz or slower
(the earliest chips were 4MHz, but 8Mhz was very common). They'd work
in an A1000, A500, A2000, Perkin-Elmer UNIX 73xx workstation, some VAX
peripherals like a DMB32 serial interface, several Atari models, older
Macs (original 128K, 512K, Plus, Mac/SE...), the Sega Genesis game
console and plenty more places.
Were you looking for a 68010? That was a common upgrade in the
A1000/A500/A2000 days - for about $20, you used to be able to get them
new. They were a revision to the 68000 with the specific features of
99% binary code compatibility (the MOVEcc instruction became priv'ed
on purpose); "loop mode" - a one-instruction cache that allowed simple
"DBcc" loops to forego re-fetching the MOVE instruction, something
that allowed block moves to execute 50% faster, resulting in a "free"
~5% speed increase in practice on an Amiga; and "virtual memory
support" - the ability to generate a page fault and save enough state
to allow the OS to grab a page from disk, then restart the instruction
right where it left off as if nothing had happened. UNIX on the 680x0
architecture relies on instruction restart to implement virtual memory
(though the Perkin-Elmer workstation I mentioned earlier used a pair
of 68000s to perform the same task - one would be halted while the
other fetched the page from disk). Amigas don't care about
instruction restart, but the oldest Sun workstations came with 68010s
for that very reason, as did the AT&T UNIX PC.
Hope this all makes sense and is useful to you.
More information about the cctech