Date   : Wed, 01 Oct 2003 14:17:23 +0100
From   : Sprow <info@...>
Subject: Re: BBC Micro lazy delayed boot!

In article
<6A4F555CEA3CD5118DBD0002E30AC0912C0854@...>,
   Andrew Hancock <Andrew.Hancock@...> wrote:

Hi,

[snip]

> > >1. Outputs of IC 16
> > 
> > All normal.

Good.

[snip]

> > >WITHOUT 8271 installed.
> > >
> > >As above.
> > >
> > >Except IC 1 6502
> > >
> > >Pin 6 NMI PERMANENTLY LOW No transition observed.
> > 
> > No, this is wrong.  It should be pulled high by R81.  Check this for dry
> > joints.  If ic27 is socketed, does it go high if ic27 is taken out?
> > What is the status of link S2?

A partially yanked out disc interface would cause this because of the
inversion from IC27 (PC chips are usually active high interrupts).
This is the purpose of S9: if you remove the disc interface you need to fit
it to ensure the output of IC27 is high.

[snip]

> If DFS 0.9 and 8271 are installed, the machine still delays for 15 seconds
> at power-on, but rather than displaying a flashing cursor, it displays the
> full 
>
> BBC Computer 32k
> Acorn DFS
> _ flashing cursor here, long beep for 15 seconds then
> BASIC

With DFS 0.9 the absence of a disc when attempting to *CAT can't be Escape'd
from (it should timeout with a disc error if you wait long enough).
I might have missed it,but what's the state of nNMI and nIRQ during the 15s
long pause?

If one of the two is jammed on then almost all of the processor time will be
spent entering and leaving the interrupt handlers - you can see the affect
this has by reading the "Slower Downer" article on my webpages,
Sprow.