7 Jul
2005
7 Jul
'05
8:14 a.m.
On Wed, 6 Jul 2005 William Donzelli <aw288 at osfn.org> wrote:
To: "General Discussion: On-Topic and Off-Topic
Posts"
<cctalk at classiccmp.org>
Message-ID: <Pine.SUN.4.20.0507062219110.19631-100000 at osfn.org>
Content-Type: TEXT/PLAIN; charset=US-ASCII
I don't see how anything could be more documented than having the complete
engineering drawings of the whole machine. And that's what you normally
got on old DEC machines.
That's what I have of the PDP-11/70. Full drawings of every curcuit in the
machine. And then I have all the technical manuals for all subsystems that
document things in a more text-like manner as well.
But as usual: when in doubt, the drawings are the definitive authority.
Johnny
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at update.uu.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Even where IBM cards of that era used
'commodity' components, they were
usually completely house-marked and impossible for a mere mortal to
figure out. This goes all the way down to resistor networks and
resistors/capacitors (except those few that have color-code bands.)
Unless you get the cross reference...
For the pre-1978 (or so) machines, you can pretty much debug down to the
chip level. With the right Blue Binders, pretty much every last part is
detailed to a silly extent. WAY more detailed than DEC docs. 
7 Jul
7 Jul
8:53 a.m.
I don't see how anything could be more documented
than having the complete
engineering drawings of the whole machine. And that's what you normally
got on old DEC machines.
That's what I have of the PDP-11/70. Full drawings of every curcuit in the
machine. And then I have all the technical manuals for all subsystems that
document things in a more text-like manner as well.
With IBM stuff, you get multiple drawings of every circuit of the machine.
The basics are the ALDs (Automated Logic Drawings) - these are the printer
generated, hard to read things that are the most detailed (all gates, all
connections, all pins labelled). Then, to make thing easy, IBM had more
traditional drawings, using nice art - still IBM specific, but much easier
to read. Then there are (many) flow charts, timing diagrams, sometimes
scope shots, and of course, warnings on where not to stick your tie.
By volume, I would say IBM docs are two to three times larger than DECs.
William Donzelli
aw288 at osfn.org
3:38 p.m.
For the
pre-1978 (or so) machines, you can pretty much debug down to the
chip level. With the right Blue Binders, pretty much every last part is
detailed to a silly extent. WAY more detailed than DEC docs.
I don't see how anything could be more documented than having the complete
engineering drawings of the whole machine. And that's what you normally got on old DEC machines.
That's what I have of the PDP-11/70. Full drawings of every curcuit in the
machine. And then I have all the technical manuals for all subsystems that
document things in a more text-like manner as well.
You may be clever enough to be able to work everything out from the
schematics alone, but I find the technical manuals useful too. If you
want an interesting puzzle, try figuring out the floating point processor
(at least the 11/45 one, I guess the 11/70 is similar) from the ROM dumps
and schematics alone. It will take you quite a while I suspect.
I think that's what was meant by 'more detailed'. The DEC docs are in one
sense complete, but you may have to think quite hard to udnerstand them.
IBM docs, from what I've heard, explain things really well, but they are
non-trivial to find.
But as usual: when in doubt, the drawings are the definitive authority.
Err, the _machine_ is the ultimate authority. There may be ECOs and FCOs
done to your PCB that are not in the docs.
-tony
6932
days inactive
6932
days old
2 comments
3 participants
participants (3)
-
ard＠p850ug1.demon.co.uk -
aw288＠osfn.org -
bqt＠Update.UU.SE