Date   : Mon, 02 Oct 2006 19:26:08 +0100
From   : "Andrew Chesterton" <andrew.chesterton@...>
Subject: Re: One more BBC B to fix

 
> In message <Y2AVZvBLIAIFFwuo@...> you wrote:
> 
> > In article <1ff6e36e4e%mjfoot@...>, Michael Foot 
> > <mjfoot@...> writes
> > 
> > >Do you think the fault is caused by a video circuit 
> problem? Do you 
> > >think it's
> > >more likely to be IC related, connection related or some 
> other component?
> > 
> > It's hard to tell, really.  The symptoms you describe seem 
> to be video 
> > or memory related, but you say the machine will boot 
> properly on the 
> > odd occasion before locking up.  Intermittent faults like 
> this can be 
> > a bear.
> 
> It's confusing that's for sure. I guess no one has fixed this 
> type of thing before.
> 
> > Have you scoped or metered the 5v power connections?  Can you try 
> > another power supply?
> 
> Yes, the PSU reads 5.25V on all 3 leads. the -5V is fine too. 
> I've not put another PSU on this machine but I have powered 
> up a spare Issue 7 board I have and it booted fine. Maybe I 
> should try a good PSU just to cover all bases.
> 
> > Leave the machine on for a while, then feel each of the RAM 
> chips with 
> > the back of a finger to see if one or more is noticeably 
> warmer than 
> > the others.
> 
> The RAM all feels the same temp. The 6502 and VIDPROC are 
> warm too. Upon switching it on today, I got a flashing cursor 
> in the top left corner in what looked like a mode 4 screen. 
> Crtl-Break did nothing so I turned it off and back on again 
> and it went back to the white screen with random LEDs.

Somebody correct me if I'm wrong, but this random cursor behaviour could
signify an addressing or Language issue (ie BASIC not working). If you are
getting a cursor, then the OS has started and is looking for something to
load, no cursor but random machine behaviour could be the machine has
crashed in the process of loading the OS and is now accessing the wrong
locations in the memory map. I haven't got a schematic to hand, but I would
check the main address decoders and any address buffers/latches.

Just a thought, but might be worth a look.
> 
> > Try putting the user 6522 in the system 6522 socket, 
> leaving the user 
> > one empty.
> >
> > Try swapping the 6502 from a known working machine.
> 
> Done all that. Everything that is socketed has been swapped. 
> The keyboard's been swapped too.
> 
> > On pre-issue 7 boards, it's very important to make sure all 6 5V/0V 
> > power connectors are attached and are clean and tight.
> 
> I'll try another PSU but I have done measurements around the 
> board and 5V was read where I tested it.
> 
> The fact that the machine does what it does only when jumper 
> 26 is south is stange. The other jumper settings don't allow 
> Break to do a reset.
> 
> Cheers,
> 
> Mike.
> 
> -- 
> Michael Foot <mjfoot(at)paradise(dot)net(dot)nz> * 
> http://homepages.paradise.net.nz/mjfoot/
> * Replace letters between () with @ and . when replying
>