Hi all,
my Name is Alex, me and a friend of mine have been actively
participating in a computer collection, mostly repairing
stuff and bringing it back to life. We have numerous old apples, vaxes,
decstations, suns, next etc and some
real weird old stuff ;)
Lately we discovered a DG nova standing in a dark corner of a storeroom.
We have absolutely no peripherals, just
the bare machine. It has 2 CPU boards, a general IO card, 2 Memory
boards and some unknown third party controller.
The Machine as it sits now has 2-3 problems (at least):
1: Somehow the "desposit next" switch does not work as expected by not
incrementing the adress sometimes. Doing a deposit-examine next-deposit
works, so thats a minor problem.
2: Simple programs run fine in single step mode. Press RUN and the
machine seems to hang. You can not even STOP anymore, you have
to reset the machine. Annonying, but to test the simple stuff single
stepping also works....
3: The biggest problem is: The ACs have numerous bits stuck on 1. It is
the same bit pattern for all ACs. We can store and
deposit the non-stuck bits, but all others remain a 1. With this
behaviour its really really hard to write any meaningful asm to
test the machine ;)
Please see the attached Link for pictures of the machine and the
bit-pattern of the AC. Since the AC Deposit works as expected
and the error shows only when examining we are positive that the error
is not in the AC Registers itself but somewhere where
the switch location is actually shifted into the AC....or something like
that. We are also struggling a bit to get all the documentation
in a meaningful relation. So if you have any pointers on what to do or
check, please chime in.
If the machine gets back to life i have plans on emulating a storage
device to run one of the available os-es. Should be doable with
a microcontroller and some logic glue.....but it looks like a long way ;)
See actual pictures of the machine here
http://bigalpha.ath.cx/nova/pics/ori/index.html
where you can also see the bit Pattern of the AC.
I kinda wonder why its
1 11X 111 X11 1X1 11X
I guess this should point us nearer to the point of failure.
best regards
Alex