Tony Firshman wrote:
>Do you get initial screen blanking, and then
>nothing (I wish TT did a splash screen!).
Or better: A progress indicator while starting up.
You probably miss the old QL way of a screen memory test you can *watch* :-)
>Describe exactly, because, like the QL, the various faults have very
>recognisable sequences.
Probably. You have a lot more experience in fault search than me.
>A really nice effect is when absolutely nothing happens. The Q40 screen
>memory stays there for a time, even when power is off.
I often wondered why it must be refreshed every 8 ms if it can survive a
power-off for several seconds.
>>I don't know the IO card Qbranch delivers with the Q40. I remember there
>>were printer IRQ problems with the card and SMSQ/E. So maybe you should
>>check if the PAR IRQ jumper is removed. If not, remove it.
>Cannot be that, as the Q40 would _never_boot.
I wasn't 100% sure, because I had IO cards that could run with this IRQ
enabled.
>The IRQ problem was card
>related, not printer. The problem (please correct me Peter (8-)# ) was
>that the IRQ from the card was pulsed, but Q40 required a longer
>(continuous?) IRQ. The relevant IRQ was 7.
You are absolutely correct :-)
The Q40 requires level-triggered interrupts. All interrupts from the IO
card are level-triggered, except the printer IRQ in SPP mode which is
edge-triggered. At design time I had not considered this a problem, because
the IO cards I used could be switched by software into another mode with
either level-triggered interrupts or FIFO.
My idea about Bill's problem was that it might have been a card that works
with printer IRQ enabled. Somehow the starting printer's signals on the
handshake-inputs could make the IO chip fire an intterupt at a time when
the booting SMSQ doesn't expect it. Similar happened with an ethernet card.
But you are right, it is unlikely.
Peter
