At 21:53 18-06-98 +0000, Joe <rigdonj_at_intellistar.net> wrote:
>At 02:50 AM 6/18/98 -0500, you wrote:
>>On Wed, 17 Jun 1998, Doug wrote:
>>
>>
>>Fixing my stuck "execute" key looks like it'll be a bastard too. It's not
>>stuck, the "snap" mechanism appears to be simply gone (not easily
>>explained since the keyboard appears to be a tight sandwich). I hope that
>>key isn't important in the big scheme of things, but I have a nagging
>>suspicion that it's needed to program.
>
> Yes, it's important. You press it to execute an expression immediately.
>You press ENTER (or STORE? I haven't used it in a while) to enter the
>expression into a stored program.
>
> Joe
>
At work we had a 9825S. Now that's another letter variant you folks had't
mentioned I think which is actually a full-option "A" model same as the "T"
is a full-option "B" model.
It originally had the calculator-style kbd before I later replaced it with
a full-travel kbd. Incidentally, that newer kbd was built by Cherry for HP.
Anyway, one of the keys became 'stuck'. Probably the Execute key IIRC (this
was arounfd '82-'84). Being myself just like Tony Duell, I removed the
upper front portion of the case and found it was able to be completely
dismantled. Well, we were absolutely dead in the water anyway as this was a
critical production machine that ran a laser interferometer encoder
calibration machine (we make linear encoders for mostly the machine tool
industry), so I grabbed my small Philips screwdriver and had at it.
I found that each of the key contacts were actually a strip of metal about,
IIRC, 3mm or 4mm wide x 12mm long. The strip had been formed into an arch
which would stand about 1mm above the surface it was setting upon. On the
keyboard assembly's PC board, each keysite had one of these arch-like
strips spotwelded to its respective PC board pads. When assembled, each key
cap had a small boss on its bottom that when pushed down against its
respective arch-like strip would cause the center of that strip to snap
down onto a contact pad on the kbd PC board. The strip acted like the
bottom of an old-fashioned oil can when the user wanted to squirt some oil
into his machinery or whatever as it would pop down and pop back with a
definite 'snap'-action. The strip "oil-canned" as we older engineers and
mechanics called it.
On that kaput keysite, one end's weld failed and had popped off -most
likely from thousands of operations over 6 to 8 years of use. Well, to get
the thing back into production I simply grabbed my small Weller soldering
pencil, jigged up the strip so it would reform its arch, held down the
broken end with a needle-like tool and soldered the danged thing down. I
reassembled the kbd (IIRC there were a LOT of small screws and keycaps,
etc.), put it back into the machine, fired it up and lo! we have Execute!
The key action was not as crisp as original as the solder fillet had crept
about one mm under the strip. But it worked.
Later, I checked with HP for a replacement kbd as I wasn't sure how long
the fix would last nor if there would be any other key collapses. The
replacement was the the Cherry-type kbd retrofit kit (for around $450 IIRC)
which was the normal issue kbd for the later "B/T" machines. This may
explain why some folks report a newer style 'full travel' keyboard on their
"A" and "S" model 9825's. The front-top of the 9825's case had to be
replaced as the kbd openings were different. I'm not sure, I'll have to
check for sure, but I *think* the name of the machine ("9825B") is moulded
into that case part. If that's the case, that *may* be why Joe's and Doug's
machines seem to be an "upgraded" A model (Quote from Doug's previous msg:
"AFAICT, it started life as a 9825A in 1979 and then got upgraded to a B/T
sometime in 1982 or later.") Yep. Kinda sounds like mine.
We bought the replacement and I kept the old kbd. I'm a certified Packrat!
I'm not sure if I still have it but I may run into it when I move a *heap*
of remaining stuff out of our old house and cram it into the new house in
the next weeks. I am almost sure it still worked.
I have both the 9825S which is the above subject and a later 9825T now in
my collection. I have to check which, but one of them had it's power
supply fail in such a way that the +5V rail had the full raw DC slammed
onto it. Smoke City. The HP service manager in the HP/Paramus facility
grumbled that he'd seen this several times before. He groused that the
designers never put in a crowbar protection circuit to prevent this from
happening. You see, the power supply is a DC/switching style. There was a
transformer, rectifier diodes and filter capacitor that provided raw DC
(about +16 volts? Gotta check my docs.) The regulation was through a
PWM-switched series-pass transistor to provide +5 volts. The pass
transistor failed in a short circuit mode and you know what happened next
:-( Every board shot, tape drive shot, display shot, printer shot.
Dang. About $2000 or $3000 worth of replacement PC boards. That is IF they
were even available. It was an old machine barely supported by HP at that
time. Not worth it for sure. Time to move to a 'modern' platform -a PC.
Mixed feelings there.
Don't be too surprised if any of you 9825 folks experience this. I darn
well hope not though! I *might* whip up some sort of crowbar for my good
9825 if it won't cobb-up things too much. Being a little gunshy from past
bad experience kinda makes me want to do that.
I can dig out those machines and my documentation and notes to answer
questions if any others here don't have the answer.
Doug, if you want me to give your "stuck" key a shot at repair let me know.
That is, if it's the older style kbd. I don't think you were clear as to
which style it was, IIRC.
It will be some time before I get my workshop rebuilt into the new house
(maybe toward fall or even winter), but I'm willing to try it for you as
I'd done before at work.
Joe, I printed out your msg giving your 9825 configurations. I'll check it
with my two 9825's to see how things compare.
Doug, in your
http://www.yowza.com/classiccmp/hp/ page you have refrences
to articles listed in the June 76 HP Journal. I'm interested in that info
as you'd imagine. Is it online at HP or anywhere or do you have the issue?
If you have the issue could you make xerographic copies of those articles
for me (if otherwise not on a webpage)? Lemme know on both the kbd fix and
HPJ articles.
Thanks,
--Chris
-- --
Christian Fandt, Electronic/Electrical Historian
Jamestown, NY USA
Member of Antique Wireless Association
URL:
http://www.ggw.org/freenet/a/awa/
Received on Fri Jun 19 1998 - 10:48:07 BST