13 May
2003
13 May
'03
3:31 p.m.
Now... the "good" news is that this only
affects interrupts and DMA
requests. AFAIK, console ODT (the '@' prompt) has nothing to do wth
interrupts. You should be able to stuff the CPU, a DL-11 and little
else (probably a bus terminator at the end of the CPU backplane, and
maybe some memory) and get the prompt when you power on. First order
There is a problem with this. The standard terminator for the 11/34
contains circuitry to assert SACK if a grant ever gets to the terminator
(this should not normally occur, becuase the device asserting the request
line (to cause the aribiter to assert the grant line) shouldn't pass the
grant on). Anyway, an open grant line looks to the terminator as though a
grant is asserted,, so the terminator asserts SACK. Since nothing can be
done to remove this 'grant', SACK stays asserted, and the Unibus is
locked up.
My experience suggests that a small Unibus machine (say 1 or 2 backplanes
in the same box) will run without a terminator at the far end of the bus.
So to start with it's safe to pull the terminator if you think you have
an open grant problem. On these 11/34 machines, an open grant will cause
the 'Run' LED to be on, and you won't be able to halt the CPU from the
panel.
of business, though, is to check your voltages. The
fact that you have
Agreed!
-tony