Date   : Tue, 14 Mar 2006 18:37:29 +0000 (GMT)
From   : Sprow <info@...>
Subject: Re: Web site: Mainly Master 512

In article <000a01c6477d$05882390$0301a8c0@...>,
   David Harper <dl.harper@...> wrote:
> Fragula wrote:

> > The 512 was more or less where I "left off" the beeb (as there was a lot
> > of misinformation as to whether it would even run on a B, I didn't bother)
> >
> > I might be "clocking" one over the next few months.. With the kids I've
> > done the whole water cooling etc. overclocking PCs and doing a blog and
> > photos about it thing... now its time to get back to basics. >;{> and
> > what better way to start?
> >
> > I've gotten some nice fast DRAMs, looking for a tasty 80186-nn (faster
> > the better providing its a pincompatible PLCC68) at the mo.
> >
> > Gonna invest in nice quality sockets and a few
> > (24/28/32/36/40/44/48/49/50 MHz) crystals, or maybe work a variable RC
> > replacment and use a counter to measure "best stable"
> >
> > Warm up the iron and have some fun! The 80186s come up fairly frequently
> > so i'm not too bothered about desecration/blasphemy etc., and i don't
> > think its got anything to break that can't be easily fixed anyway.
> > Except maybe the Tube ULA.

Spare Tube ULAs available
 http://www.sprow.co.uk/bbc/forsale.htm
last few!

> Can't see any problem with this. The Tube is an asynchronous connection

[snip]

> Perhaps someone on the list knows the electronic requirements of the Tube 
> ULA chip and its response times.

These are covered in app note 4, the electrical specs (as in setup/hold
times) are geared to cater for a 4MHz 6502, the other times are protocol
related geared towards a base 2MHz 6502. 

I would tend to agree that a faster parasite (eg. ARM7 in my case) doesn't
cause any problems at all - you just end up spending more cycles waiting for
the host to catch up,
Sprow.