Whitechapel Computer Works MG-1

[tony duell]

> > It needs a good set of NiCd cells on the power control board to start it
> > up.
> > If not, you have to do a 'jumpstart' involving connecting a 9V battery to
> > a connector
> > on that board. Some owners added an external socket wired there to make it
> > easier to so this.
> 
> Another jumpstart option was to boot one up, leave it running for a moment
> and then connect it across to the dead machine

Quite likely. All it really needed was a voltage to energise the mains swtiching relay.
Once the PSU had started up, the relay would be held in until it was turned off 
by software (this was a trick used on a number of UK unix boxes, the power-off
button tells the system to flush buffers, do an orderly shutdown and then turn off
the power relay). 

The method using a 9V battery is official, it's in the technical manual I have somewhere.


> > The 32016, 32018 (FPU), MMU chips, etc are in nice turned-pin sockets. The
> > EPROMs are in cheap sockets. Replacing the latter sockets will often get a
> > dead machine going.
> 
> I think some of them might be early enough that they were still calling
> them 16032

Quite likely. The chip was renamed for marketing reasons IIRC (it was a 16 bit chip 
with 32 bit registers, renamed as a 32 bit chip with a 16 bit bus or some such). 

[RAM expansion]

> There were also some 2M cards made if you needed a lot of memory. It was

Yes. The RAM cards were essentially DRAMs and buffers, The 512K one had
4164s, the 2M one 41256s. There were other differences, and I have never tried
to turn a 512K one into a 2M one.

> possible to stack enough 512K cards that the case would not fit back on,
> that did not do much to improve the machines already uncertain reliability.

The RAM board connectors were wire-wrap DIN41612s (64 pin AB row). They were 
soldered to the RAM boards, you could stack the boards by putting the long wire 
wrap pins of the top board into the sockets on the board below (this was official), 
There were screwed spacers to support the boards, but it was not uncommon to 
forget those. At  which point you got RAM wobble (as in the ZX81...)
[...]

> The keyboard was all kinds of awful, the mouse had a metal ball that needed

The keyboard was a Keytronics capacitive thing, and feels much like any other.
The mouse is one of those Swiss Depraz ones, nice when you get used to it. 
Its a standard quadrature 3 button mouse, in fact my machine came with
a defectrive Depraz mouse (which I repaired) and an Acorn mouse re-wired
to work. I am pretty sure the keyboard interface is documented in the manual,
it's something like 1200 baud asynchronous so you could always substitute
another keyboard if you really hate the WCW one.

-tony