4 Apr
2005
4 Apr
'05
3:58 a.m.
On Sun, 2005-04-03 at 18:43 +0100, Tony Duell wrote:
Mention of interrupts is interesting...
Actually, the machine has two SCC chips (each having two ports) for a
total of four serial ports. One daisy-chains the interrupt line from the
other.
What I thought was the bad SCC (which runs the console port #1 and
serial port #2) is actually at the end of the chain - I suppose it's
possible that the other SCC (running ports #3 and #4) is causing
problems.
I'll see where else that interrupt line goes to too...
Problem is there are a lot of unknowns on the board - at least 30 PAL
chips, plus a handful of Sun chips having around 200 pins each for which
there are likely no specs in the public domain :-(
Of course, tracing as much as I can to do with the serial circuitry
might be an option, and may show up some useful things to test.
cheers
Jules
Right. It's *not* the 8530 SCC chip that's faulty, but it's *not* the
data bus either - there's plenty of bus activity so bit 6 isn't
permanently jammed high.
After checking the bus, I replaced the SCC chip with a socket and tried
my spare SCC chip from home (remarkably easy job as it turned out) -
gave exactly the same results on the console.
First thought : There's some other device -- maybe soemthing like an
interrupt vector source buffer -- that's being enabled along with the
Z8530. Of course this might be inside that LSI chip you mention. I can't do
much about 1 and 2 without knowing the pinouts of the LSI
chip or having full schematics of the board :-(
Well, there's a winter-evening project for you -- draw out the
schematics.