Date: Fri, 15 Feb 2013 16:25:32 +0100
From: RolexTM at gmx.de
To: cctalk at
classiccmp.org
Subject: DG Nova repair etc.
The original Nova has a 4 bit ALU
What do you mean? I think i did not get your answer.
> When I got my Nicolet, I had a similar
sounding problem.It turned
out to be a missing clock.Of course, yours may be
soemthing different
but you aregoing to need some schematics and a scope to track down.( I
did mine without a schematic but I was lucky ).Dwight
We already sighted most of the pdfs and schematics at bitsavers and
simh. We have a scope and logic analyzer
and some experience in component level repairs of electronics. This
machine is just a bit harder to diagnose
than for say our old VT100...the documentation is overwhelming and my
brain has not adapted to it well atm ;)
so i just hope to get somebody that has intimate knowledge of the
machine....
The problem of the machine only running on single step and not on "run"
may well be related to a missing clock, although
i wonder if even single instruction stepping would work without clock
since it also needs some form of clock to single step
through one asm instruction...although i might be on the wrong train
since there may well be more than one clock inside the machine. Single stepping is
clocked by the switches on the front panel. When you hitthe run switch it runs at full
speed by the processor clock. There are severalclocks used that all need to happen.On mine
it was the write clock to the core memory. Core uses distructivereads and needs the data
to be written back after each read. The missingclock caused the failed write to memory.All
the other front panel operation worked fine.I doubt your problem is the same. I was just
pointing out how differentfront panel operation was than full speed. The front panel runs
thingartifically at mechanical switch speed.Do you have schematics for your machine?Dwight