Tony, and all,
Again thanks for the good counsel. The basic PERQ passed each phase of checkout testing
just fine. The monitor and keyboard both had issues . For the moment, I've swapped in
working items, while I look through those issues. With those in place, the system booted
right into PNX. Basic UNIX operations ran without a hitch. I've not located the right
archive box for software yet, so that will have to wait for another weekend. There's
also problem of cable insulation peeling off on the 'good' keyboard and monitor
leads. That too gets queued for attention. But all in all quite a successful outcome. I
will keep interested parties posted. If you want to see this system, it's in
TNMoC's multi-purpose/scientific computing gallery.
ard at p850ug1.demon.co.uk wrote:
All, especially Tony,
Fortunately my love of classic computers, expecially classic PERQs,
well
exceeds my dislike of certain organisations :-)
I'm getting ready to power up a PERQ 1a which
has been sitting for a
while (5+ years). I have no reason to believe it has any faults.
I've
done the obvious prep, like unparking the disk,
and removing the
locking
It's probably too late to say this, but you shouldn't remvoe the
lcoking
clip from the postioner on the SA4000 winchester drive until it's
spinning. It's too easy to move the heads as you pull the clip out, and
doing tha on a stationary disk is not the best idea. Although to be
fari,
these drives are pretty solid, and it's hard to wreck them.
screw. And I will need to build a dummy load to
retest the power
The power supply is one of those infernal Boaschert 2-stage ones.
However, it doesn't seem to need a dummy load to regulate correctly.
You
can run it with no load and it should be fine.
supply. Are there any other gotchas or quirks I
should be aware of?
I
have no previous experience with this model, and
would err on the
side
of leaving it dormant rather than proceeding
blindly.
atmittedly the PERQ 1 was the last (of 4) PERQs that I obtained, but I
also had to start PWERQ repaires with no knowledge :-).
I apologise if I mention things that are obvious. Some things seem
obvious to me, but I then find after much emailing to and fro that this
was the problem.
Tke the casing off first. I guess you've done that., though. If you
haven't, the front and rear panels come off first, then the sides. THis
frees the top cover over the floppy drive, etc.
There are interlock switches that prevent the machine pwoering up with
the front and rear panels removed. They;re those standard white ones
found in a lot of machinesm, and you can pull thje actuators out a
little
way to turn them on, so you can run the machine with the cover removed.
The PSU is under the main card cage, and is a right pain to get to.
Basically you have to remvoe the card cage. So to save a bit of time,
you
can test it without removing it :
Unplug the ribbon cables from the front of the CIO board. The top one
goes to the floppy drive (SA850 pinout), the bottom one ot the
wincheter.
Lable them if you are not sure. Then pull the boards out of the card
cage. You don't bneed to remvoe them all the way, jsut make sure they
don't connect to the backplane. There are 3 boards i na normal PERQ,
form
the left they are the CPU, then the memory (and video) and finally the
CIO board (input/output, disk controllers, etc). If you are very
fortuantel there's a 4th vaord on the right of these. That's the OIO
(Optional Input/Ouptu). It adds a 16 bit parallel interface (a bit like
a
DEC DR11-C [1]), and oen or both of ethernet (which is not standard on
PERQ 1 and 1a models) and Canon CX laser pritner interface. There are
other boards that could be inthat slot (like a QIC-02 tape interface or
a Datacopy 300 camera controller), but you are very unlikely to see
those.
Now disconenct the DC power cables from the floppy drive nad
winchester.
Unplug the AC cables to the motors if you like, in fact the Winchester
motor cable should have been unplugged when the spindle locking scrrew
was fitted.
HAve noting plugged into any of the D connectors o nthe backplane
At this point there.is nothign conencted to the DC output of the PSU.
Connect mains, with the interlocs pulled out, the breaker closed, flip
the little switch on the front./ The fans should start up. If that's
OK,
check the DC voltages on the labelled traces on the backplane.
It's rarew for them to be out of tolerance. Eitehr they are OK, or
they're all missing. If nthe latter case, you have to work o nthe PSU,
whioch I can explain when you need to do it. For the moment I will
assume
the PSU is OK.
Power down. and remvoe the mains lead. Put the logic boards back in,
connect the keyboard ot the appropriate conenctor on the backplane. On
the underside of the keyboard is a 3 digit 7 segment display. Power up
again and press and release the reset button on the back of the
keyboard.
Look at that display. if it gets to 010, then the CPU has passed
power-on
diagnositcs, and is attmpting to load microcode from disk. If not, you
haev some logic faultfinding to do.
Power down, reconenct the drive data ables to the CIO board and the
drive
power cables. Conenct the monitor to the correct conenctor, and power
up
again. Again blip the reset button. Watch the display. At 030, you
should
see memory test patterns on the CRT screen. If not, maybe there's a
monitor problem. If you are lucky, the machine will complete the boot
from the winchester, and you vncan then run whatever OS it's loaded. If
not, you might have to re-install thigns from floppy.
I've left out a lot of what to do when it doesn't work. What I epxext
is
you do the above 9which is safe, and will protect the mahcine from
amjor
damage), adn then get back to me when you ahve problems -- like the PSU
is giving no outputs, or the DDC (7 segment display) sticks at 009, or
whatever. THen I can help uou further.
-tony