Date   : Sun, 28 Sep 2003 17:02:57 +0100
From   : Andrew Hancock <Andrew.Hancock@...>
Subject: Re: BBC Micro lazy delayed boot! - Interesting develo

 
TO TEST only with Acorn 0.9 DFS slightly different problem is observed. This
is different to DNFS 1.2, which doesn't display on anything on screen. 

If I substitute the DNFS 1.2 ROM with an original Acorn 0.9DFS eprom, in any
socket, with or the 8271 chip installed, the Beeb powers-up immediately with


With 8271 installed

BBC Computer 32K
Acorn DFS
_ flashing cursor and long Beep................................

15 seconds (faster if warm!)

BASIC appears
>

same on pressing BREAK and CTRL-BREAK (with Beep.............)


Without 8271 installed

BBC Computer 32K
Acorn DFS
_ flashing cursor and long Beep................................

same on pressing BREAK and CTRL-BREAK (with Beep.............)

Machine never reports to Basic and sticks.


I always through the DFS checked for the prescence of 8271 before mapping
DFS service rom in memory.


If I install both DNFS 1.2 (low priority) and DFS 0.9 (high priority), I get
both problems.

1. Flashing cursor only for 15 seconds or faster if Beeb is warm.
2. BBC Computer 32K, Acorn DFS, Flashing cursor 15 seconds or faster if Beeb
is warm, long tone when Basic prompt.

So total Boot time is 30 seconds. I'm surprised at this because I would have
thought as 0.9DFS is at a higher priority it would not map/execute DNFS 1.2.

*HELP reveals both DFS 0.90 and DFS 1.2 installed.

If DNFS 1.2 is installed in a right most (high priority) socket, only the
following is observed as usual.


1. Flashing cursor only for 15 seconds or faster if Beeb is warm.

*HELP reveals only DFS 1.2 installed. I would expect this.

I'm not too sure if this sheds any more light on the subject. But I'm
surprised at the different between DFS versions.

Regards

Andrew


-----Original Message-----
From: Sprow [mailto:info@...] 
Sent: 28 September 2003 12:40
To: bbc-micro@...
Subject: Re: [BBC-Micro] BBC Micro lazy delayed boot!

In article <bq7VupBwKsd$EwsY@...>,
   Mike Tomlinson <mike@...> wrote:
> In article <6A4F555CEA3CD5118DBD0002E30AC0912C082B@...
> consultants.co.uk>, Andrew Hancock <Andrew.Hancock@...
> consultants.co.uk> writes

> >And I've just discovered this morning, the colder the machine the longer 
> >the
> >wait, because yesterday evening as the Beeb was left on the warmer it got
> >the quicker it booted.
>
> Check the power supply voltages with a meter.  This sounds like the
> classic C9 problem in the PSU.

But the fact that it makes some initial attempt to startup (albeit the first
and second beep are far apart) would imply there is juice coming out of the
PSU.

As I think another suggested prodding the nNMI and nIRQ lines on the CPU
with a logic probe would be a good first guess,particularly with the
relationship between removing and inserting the 8271.See if the slow startup
correlates with one of the two lines transitioning.

Also (my current favourite guess) would be to look at the output of IC16 and
IC33,giving the huge ceramic disc capacitor a wiggle,
Sprow.