On 11/27/25 12:05, Anders Magnusson wrote:
Summary: IT RUNS!
The problem was in the CIB board, a failed 74S74 flip flop in
interrupt handling.
We'll post pics and such soon.
Thanks everyone for the guidance. We'll try booting VMS on it
later tonight, temporarily from the VAX-11/750's system disk. If
all goes well there, we'll set up permanent system disks for it for
VMS, NetBSD, and possibly Ultrix-32.
Incidentally, this is the same '780 that Gunther Schadow had
quite a saga with on this very mailing list, back in 2002. To my
knowledge he never did get it running. We fixed a number of
hardware problems in this machine, so it's no wonder he'd had such a
tough time with it. It was many months of work, off and on, over
several years for us here at LSSM.
Cool! Real hardware fix!
How did you manage to find the failing chip? I.e. the error
isolation process? :-)
Well, the big break was information on this list, that the first
character of the line where it stalled was the first thing printed
with interrupts. That quickly led us to the console TX interrupt flip
flop on the CIB board. Three of us looked at the schematics that
night, separately at home, and all reached the same conclusion.
Booting VMS on the machine told us nothing, but the NetBSD kernel is
far more verbose.
The microdiagnostics specifically flagged the M8230 (CEH) as being
bad, but it was the CIB. Now the M8230 tests pass too.