30 Nov
2007
30 Nov
'07
4:05 p.m.
Yep - I was poking around at the Acorn ACW PSU recently [1] and it struck me
that it should be hard to modify pretty much any 'simple' SMPSU to run off US
'mains', even if it was notionally 220V (-ish) only. You're right about
The reverse is, alas, not true. It's easy to design an SMPSU which runs
at about 170V DC, produced by rectifying 115V mains. There's no easy way
to convert those to run off 230V mains (other than a step-down
transformer...)
'auxiliary' stuff run direct from the AC input
of course.
[1] It was a model made by VMS - *not* the same as the one in your hand-drawn
ACW schematics. I actually found two ACWs with the same supply, suggesting I
didn't just happen to have someone's post-production replacement. I ended up
This does not suprise me. When I got my ACW, the PSU was 'held' in by
heatsink compound sticking it to the monitor chassis. The 2 fixing screws
were the wrong trhead for the holes in the PSU bracket. And the output
cable is a bit of a kludge. But it fits, and it gives the right output
voltages...
I suspect Acorn used several different supplies. And maybe the supply in
mine was replaced later.
drawing out the schematics, which I should probably
mail a copy of to you
sometime just for completeness' sake...
It would certainly be interesting to see , although the chances of me
ever having to work on another ACW are slim...
Talking of rare UK workstations, did you see that optimist on E-bay who's
trying to sell a non-working Whitechapel MG1 [1] with a defective
monitor? Starting bid \poudns 3000. Hmmm...
[1] A 32016-based workstaition, the 'MG' standing for 'monochrome
graphics' or 'Milliard Gargantubrain' (!) It has the odd feature of a
bit-serial graphics processor thats' slower than doing the graphics
operation on the 32106, and was thus left out of later machines. One
option for it is a 3-slot ISA backplane (!). Oh, and it has a
battery-operated power switch, the batteries routinely go flat, the cure
being to connect a 9V battery to a header inside provided for the purpose
(according to the techincal manual).
never
prodded a 6809 myself (just 6502 and Z80), but I guess I could stand to
learn too :-)
You'll love it. It's waht the 6502 should ahve been. 16 bit index
registers, 16 bit stack pointer -- no make that 2 16 bit stack pointers,
progam counter relative addressign modes and long branches for
position-independant code, 'zero page' in any 256 byte page of memory, 2
accumulators that can be combindd to make a 16 bit one, the most
orthogonal instruction set of any 8-bitter, etc, etc, etc. I always wish the Beeb had used that as its CPU.
Acorn clearly knew about
it (they'd made a 6809-bsed System CPU board), but for some odd reason
they didn't use it in the Beeb.
You know, one day when I get time (yeah, right), I'd quite like to build one
of the 6809 System boards. I've got the board manual/schematic, and I believe
I've got a copy of the firmware ROM somewhere. I've also got at least one
Control Universal System-bus 6809 board and want to
compare one to the Acorn
schematic sometime - as Acorn licensed the System technology to CU, it'll be
interesting to see if the CU board is essentially the same as the Acorn one.
I'm not sure [1] when the 6809 CPU came out? Acorn's System 1 was circa
1978/79, so maybe that pre-dates the 6809 by a while - and it was just for
legacy reasons that Acorn stuck with the 6502 for the beeb after using it for
the System and Atom machines? (somewhere I've even got an Acorn System bus
expansion module for a BBC micro, so it'll take System-era boards)
The 6809 and 6502 buses are _very_ similar (read 'E' for 'Phi2',
basically ;-)). Most I/O cards would work with either CPU (hence the lack
of problems in haveing a 6809 CPU in a System). Of course the
instruction set, and thus firmware/software are very different.
Yes, I suspect the Ssytem 1 does predate the 6809 -- just. But what about
the relative times of the Beeb and the 6809-based System CPU board? I
would have thoguht the Beeb could have been designed to use the 6809.
-tony