How to fix IBM PC/AT keyboard?
Hartmut Krafft
this may be slightly OT (if so, tell me, I'm new to this group :-)
I have an old IBM PC/AT keyboard here (w/o separate keypad, resembling
the XT layout) which worked well until it got drenched with juice :-(
So I took it apart, cleaned everything, fumbled all the parts together
again (which was quite a challenge). Now, unfortunately, some keys
behave as if they were pressed all the time, although they aren't.
There seems to be some sort of isolation problem between the contact
board and the contact plates that make up the actual key switch.
Any hint will be greatly appreciated, as this keyboard always was
great to use.
On the bottom cover, it carries ID NO. 515X-55-03756L, mfg. date
January 86.
TIA
Hartmut
--
(-: PS/2 9577-0NG - OS/2 Warp - Linux 2.2.13 :-)
Remove all numbers from email address to reply directly.
Stephan Goll
first you should determine what makes the contact. The keys can be
opened in most cases, but dont ask me how. If there are real switches
inside the keys, mounted and soldered on a pcb, you have to clean and
may be adjust them. If the contact is made using thin and partly
metallized foils, separated by a thin isolation foil, you have to
disassemble the layers and clean them. This is mostly harder, because
the keyboard I repaired had a solid piece of steel below the key
assembly, fixed with melded plastics, to made this keyboard rock solid.
For assembling this after cleaning I used a drill and small metric
screws because the plastic pins can only be used to hold the parts in
place.
Greetings,
Stephan
Hartmut Krafft schrieb:
Riccardo Rubini
"Hartmut Krafft" <123har...@mail.ru> ha scritto nel messaggio
news:dumNy3q19HmE-p...@CTRK5.intern...
> Hello,
>
> this may be slightly OT (if so, tell me, I'm new to this group :-)
>
> I have an old IBM PC/AT keyboard here (w/o separate keypad, resembling
> the XT layout) which worked well until it got drenched with juice :-(
>
> So I took it apart, cleaned everything, fumbled all the parts together
> again (which was quite a challenge). Now, unfortunately, some keys
> behave as if they were pressed all the time, although they aren't.
> There seems to be some sort of isolation problem between the contact
> board and the contact plates that make up the actual key switch.
>
> Any hint will be greatly appreciated, as this keyboard always was
> great to use.
> On the bottom cover, it carries ID NO. 515X-55-03756L, mfg. date
> January 86.
> TIA
>
> Hartmut
>
> --
> (-: PS/2 9577-0NG - OS/2 Warp - Linux 2.2.13 :-)
> Remove all numbers from email address to reply directly.
Hi Hartmut,
I got the same problem and I tried to fix two IBM keyboards... I was nearly
to get crazy! Does your keyboard uses springs or is it soft-touch? This is
really important for me to better understanding the problem. Let me know.
Riccardo
Al Savage
wrote:
> I have an old IBM PC/AT keyboard here (w/o separate keypad, resembling
> the XT layout) which worked well until it got drenched with juice :-(
>
> So I took it apart, cleaned everything, fumbled all the parts together
> again (which was quite a challenge). Now, unfortunately, some keys
> behave as if they were pressed all the time, although they aren't.
> There seems to be some sort of isolation problem between the contact
> board and the contact plates that make up the actual key switch.
If it's an original AT keyboard, then it uses the "buckling spring"
system. Generally, it just requires great care in reassembly,
especially as you slide sideways the metal "lower half" onto the the top
half's hooks. I used to have a bag full of spares for them, but I
haven't opened a real IBM keyboard up in over a decade. They were very
expensive and I used to have to repair them. They are not as easy to
repair as most other keyboards of the era. I used to repair a lot of
terminal keyboards (soldering in new keyswitches, removing staples &
paperclips, etc.).
Regards,
Al S.
Hartmut Krafft
> On Wed, 22 Nov 2000 10:28:05, 123har...@mail.ru (Hartmut Krafft)
> wrote:
>
> > I have an old IBM PC/AT keyboard here (w/o separate keypad, resembling
> > the XT layout) which worked well until it got drenched with juice :-(
> >
> > So I took it apart, cleaned everything, fumbled all the parts together
> > again (which was quite a challenge). Now, unfortunately, some keys
> > behave as if they were pressed all the time, although they aren't.
> > There seems to be some sort of isolation problem between the contact
> > board and the contact plates that make up the actual key switch.
>
> If it's an original AT keyboard, then it uses the "buckling spring"
> system. Generally, it just requires great care in reassembly,
> especially as you slide sideways the metal "lower half" onto the the top
> half's hooks.
Yes, I noticed that :-) The problem seems to be: The key switches
aren't 'real' switches consisting of contacts, but rather some kind of
sensors. If you press a key, a conductive plate tilts and gets pressed
against a PCB with electrodes entirely covered by an isolating
varnish, seemingly causing a change in capacity (I _guess_ :). The
keyboard's electronics detect this change and determine which keycode
to issue. When I reassemble the keyboard the electronics just go
berserk and detect a variety of pressed keys all the time, although
none actually are pressed. This is not a matter of contact between
switching pads and PCB (I tried to put a thin plastic film, the one
used to wrap up sandwiches, between pads and PCB, it didn't help). Do
the electronics have to be adjusted after the assembly? I'm stunned by
this problem, I admit: everything about this keyboard seems perfectly
logical if you look at it, but it just won't behave that way - I fully
share the feelings of Riccardo (in this thread) who was nearly driven
crazy by this. Isn't there anyone to give us enlightenment ? Anyone
who worked on this stuff, maybe ?
Regards
Hartmut
--
Charles Lasitter
Krafft) wrote:
>Isn't there anyone to give us enlightenment ? Anyone who worked
>on this stuff, maybe ?
I learned some of the same difficult mysteries of the 'buckling
spring' keyboard but with less enlightenment. I basically came
to the conclusion that you can't fritz with it beyond a certain
point, and that if water ever gets in between those membranes,
it is likely 'hosed' :)
CL.
+-------------------------------------------------------------+
| Charles Lasitter | Mailing Address | Shipping Address |
| 919/286-0100 | PO Box 2478 | 1304 Broad St |
| cl+at+ncdm+dot+com | Durham, NC 27715 | Durham, NC 27705 |
+-------------------------------------------------------------+
Hartmut Krafft
(Charles Lasitter) wrote:
> On Thu, 23 Nov 2000 13:16:21, 123har...@mail.ru (Hartmut
> Krafft) wrote:
>
> >Isn't there anyone to give us enlightenment ? Anyone who worked
> >on this stuff, maybe ?
>
> I learned some of the same difficult mysteries of the 'buckling
> spring' keyboard but with less enlightenment. I basically came
> to the conclusion that you can't fritz with it beyond a certain
> point, and that if water ever gets in between those membranes,
> it is likely 'hosed' :)
>
Thanks to you all!
Finally, I fixed the problem. It took some time though ;-)
First, I caught myself in a kind of mystification :)
All the little black devices that get pressed against the base PCB
carry numbers, and most of them are different. So I thought that there
maybe had to be a mysterious plan according to which they had to be
reassembled. No coincidence w/ scan codes, key numbers, though.
Of course, I couldn't find out the scheme. There isn't any :-)
After some guessing about the working principle of the keyboard, I
finally found the trap I'd caught myself in: I had assumed that, after
putting together the steel top and bottom plate and reassembling the
keys, I could test the keyboard function. But the capactive sensors
inside are seemingly so very sensible that this is not possible.
That's were the grey coanting on th inside of the plastic case comes
in: it is an electrical shield. So, only after reassembling the
keyboard completely and assuring a good conductive connection between
case and chassis, it now works flawlessly again. And I am happy :)
One question remains, though: this is an PC/AT keyboard with the
standard 5-pin DIN plug (non-PS/2). It works on all my AT type (klone)
PC's (from 386DX40 to Pentium III) with or without A PS/2 adaptor. It
won't work, though, on my 9577-0NG: there's no error when IPLing, the
lights blink, but then, no key reacts when pressed. Another mystery.
Regards
Hartmut
--
Remove all numbers from email address to reply.
Gfretwell
fooling with one. They were always FRUs at IBM.
Internal parts were never available. Occasionally I would pop a failing keytop
out to blow the thing out but beyond that it was in the recycle box.
You have a lot more patience than I do.
Hats off!
Rich
When push comes to shove, and you really need to get that keyboard
working that has a stretched bent spring hanging out where the
whole key plunger has been removed...
Well it is possible to shave down a pencil, and use it to unscrew
a good spring off a junk keyboard, and screw it in as an "FRU" part.
We did quite a few like that. With a little practice, it works.
Otherwise the dumpster would be full <g>.
Cheers Rich