Author Topic: Keyboard Matrix HELP  (Read 6559 times)

0 Members and 1 Guest are viewing this topic.

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« on: Thu, 22 April 2010, 00:46:50 »
Recently my G15 Keyboard Ver.1 has had some unresponsive Keys, I took it apart readjusted everything and it worked fine. I took it apart a second time to do some cleaning and caused a few keys to become unresponsive again. I was unable to fix it so simply this time so I spent some time inspecting it. Now half the board wont work.

http://i278.photobucket.com/albums/kk92/diablo8000/Picture19.jpg

http://i278.photobucket.com/albums/kk92/diablo8000/Picture20.jpg

I used a Multi-Meter to test continuity in the matrix where the keys are unresponsive. Yes there is continuity, and what appears to be proper traces. I am able to trace all the supposed "bad keys" all the way back to here.

http://i278.photobucket.com/albums/kk92/diablo8000/Picture33.jpg

This TAB connects to the Keyboard... board, to ensure that it wasnt the board connectors I used my second matrix sheet from my other G15 and confirmed the connection in the board is fine and works with thos keys on the second matrix sheet I have. I am at a loss as to further instructions of repair.

P.S. dont tell me to go "buy" a new keyboard, thats not what I'm interested in, I am interested in understanding.
Soopa, Doopa, Fooba, Home of the Fooper

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #1 on: Thu, 22 April 2010, 06:18:21 »
Yes, they are in the correct order as you mentioned, in fact they are stuck together in that order out of the factory. I have removed the bottom sheet for easier use with the multimeter and I can definitely confirm that I line them up properly before each test, yet still I cant figure out why this one sheet has failed but has good traces / continuity in the not working keys.
Soopa, Doopa, Fooba, Home of the Fooper

Offline kishy

  • Posts: 1576
  • Location: Windsor, ON Canada
  • Eye Bee M
    • http://kishy.ca/
Keyboard Matrix HELP
« Reply #2 on: Thu, 22 April 2010, 08:07:55 »
You could try jumping (with a wire or something) the appropriate connectors on the controller PCB and see what happens...if it doesn't send the keystroke, the membrane isn't the problem.
Enthusiast of springs which buckle noisily: my keyboards
Want to learn about the Kishsaver?
kishy.ca

Offline trievalot

  • Posts: 246
Keyboard Matrix HELP
« Reply #3 on: Thu, 22 April 2010, 08:12:48 »
the Model M is somwhat similar in sheet construction, i needed a silver pen to redraw a couple "burnt out" tracks
[SIGPIC]

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #4 on: Thu, 22 April 2010, 09:20:46 »
Quote from: kishy;174395
You could try jumping (with a wire or something) the appropriate connectors on the controller PCB and see what happens...if it doesn't send the keystroke, the membrane isn't the problem.

go on....... I am slightly uniformed in how to do this, I have a Silver conductive pen as well, and I used it on the sheet with the broken NUM pad connectors, I re-established a connection between them and the connector TAB, but it only gave the NUMpad 5 connection, and when I hit 5 it hits = sign, lol. What am I doing wrong here? I don't want this thread to get confusing tho. The sheet I am concerned about is the most non responsive sheet I have, where most of the keys on it don't work, yet there is a ohm trace back to the connector tab as showed in pic 33.
« Last Edit: Thu, 22 April 2010, 09:32:39 by Pandawhat »
Soopa, Doopa, Fooba, Home of the Fooper

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #5 on: Thu, 22 April 2010, 09:38:18 »
20 seconds prior to this post I re-confirmed that the "bad keys" have a proper trace all the way back to the connector TAB as showed in pic 33. I used warm water to clean the sheets in a rinse style hand wash just before they started to malfunction and loose responsiveness, the other day.
Soopa, Doopa, Fooba, Home of the Fooper

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #6 on: Thu, 22 April 2010, 09:46:48 »
How do I do this exactly? Short out the two fingers on the controller? I am not sure if I fully understood. I apologize, everything I learnt up to now is basically self taught. Attempting to learn as fast as I can.
Soopa, Doopa, Fooba, Home of the Fooper

Offline nanu

  • Posts: 290
    • http://T-T.be/portal
Keyboard Matrix HELP
« Reply #7 on: Thu, 22 April 2010, 10:38:19 »
The membrane likely slips into some header on the controller.  The header connector in turn has one metal contact point per membrane trace.  You should be able to connect all possible pairs of these contact points to generate all keystrokes, since that's what the membrane is supposed to do: connect two wires/traces/leads together to actuate a keystroke.

Suppose there are 20 contact points/pins/leads/traces/solder blobs at the connector, and it's soldered onto the controller PCB. I'm grossly abusing terms here as you can see.

While AquaKeyTest is running, do the following:

Take one end of a wire, place it on pin 1.
Then "sweep" the other end of the wire, contacting it on each pin from 2 to 20.

Repeat this for each successive pin:
Put one end of that wire on pin 2, and sweep across 3 to 20.
You don't have to sweep 19 per pin because a previous sweep would've tested that pair...

On a normal keyboard, you should be able to fully "light up" Aqua's Key Test.
Beware however, if your keyboard supports extra keys such as those nifty ones which will either shutdown or put a Windows machine in standby...
Take note of these pin combinations and avoid them until you complete the sweeps.

Offline Rajagra

  • Posts: 1930
Keyboard Matrix HELP
« Reply #8 on: Thu, 22 April 2010, 11:42:39 »
Here is the problem: membranes are crap - at least ones like the G15 has. If you test again with your meter, you'll find that the tracks don't conduct too well even when they are intact, they have quite a high resistance. Several K Ohms if I recall correctly. And the longer a track is, the higher the resistance. Yours seems to have the same problem mine had, some of the tracks look darkened. They seem to oxidise, especially if something has spilled into them. This further compromises their ability to conduct a signal.

I don't know at what point the resistance becomes too high for reliable operation, but I know mine is unreliable despite all tracks having some connectivity. Oddly, all keys work on mine, but two keys also wrongly switch the macros to set 1 or set 3. (Sadly, negates the main usefulness of this board.) There are no shorts, I tested it thoroughly.

I think the only fix is to replace the membrane, maybe from a faulty board off ebay (different fault, obviously!)

Don't buy a G11 as a donor, the membrane layout is different. I fell for that trap.

Edit> I think cleaning these membranes is a risky process in itself, especially if you use alcohol or other solvents. The conductivity of the tracks is so bad to start with that removing the smallest amount of material renders the keyboard useless.
« Last Edit: Thu, 22 April 2010, 11:50:23 by Rajagra »

Offline kishy

  • Posts: 1576
  • Location: Windsor, ON Canada
  • Eye Bee M
    • http://kishy.ca/
Keyboard Matrix HELP
« Reply #9 on: Thu, 22 April 2010, 15:34:55 »
Whoa, yeah, sorry I guess I missed that part.

And since I don't like the keyboard in question, my attention span is quite narrow...
Enthusiast of springs which buckle noisily: my keyboards
Want to learn about the Kishsaver?
kishy.ca

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #10 on: Fri, 23 April 2010, 14:28:54 »
I haven't had time to attempt anything yet, but possibly today or tomorrow I'm going to start with a proper cleaning solution for the membrane. If that fails to work I will then try the AquaKey test. If none of thos work I guess ill be searching for parts. I will keep this updated. Thanks again for everyones help, I very much appreciate everyones input.

P.S. since I have two matching membranes, one where the main board wont work, and the second one with just the key pad not working. Could  I use a pair of scissors, a silver pen and somehow connect the two? lol, or has this membrane driv'in me off the deep end?
« Last Edit: Fri, 23 April 2010, 14:39:39 by Pandawhat »
Soopa, Doopa, Fooba, Home of the Fooper

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #11 on: Mon, 26 April 2010, 12:03:00 »
The 99% ISOPROPANOL worked on one keyboard, it returned full function of the failed keys. The second keyboard with the least working keys is next, I gave it a rub down but I did not go over it with a Q-tip and clean up the dark traces, I have a feeling that this could work, if not AquaKey test is next.

Thank you so much to everyone who contributed their thoughts.
Soopa, Doopa, Fooba, Home of the Fooper

Offline itlnstln

  • Posts: 7048
Keyboard Matrix HELP
« Reply #12 on: Mon, 26 April 2010, 12:04:51 »
I would have just used Everclear.  That way, if it didn't work, you could just drown your sorrows.


Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #13 on: Mon, 26 April 2010, 17:00:18 »
The membrane that works and the one that doesn't were both tested on the same keyboard "board", this is the type of connectors from the matrix to the board, so I wouldnt think that the fingers are the reason behind the other membrane not working, but then again I'm not certain about anything.

Soopa, Doopa, Fooba, Home of the Fooper

Offline Rajagra

  • Posts: 1930
Keyboard Matrix HELP
« Reply #14 on: Mon, 26 April 2010, 17:08:44 »
It's quite scary pulling out and reinserting those connectors. They go in fairly deep, and it can take some force. And although there are rigid parts of the ribbon to help it go in, that also creates a risk of folding and cracking. And the G15 has so many of them. The connector for the display is especially "fun" to persuade into place, despite having a ZIF socket.

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #15 on: Wed, 28 April 2010, 21:10:27 »
Hmm, had the whole thing working changed the LEDs in the G15 everything was working ship shape, put it together, few key failures, re establish the connection..... ok every things working put it back together.... FAIL.....FAIL... loosing more keys as i go... AHHH shoot me

I need a solution to re write most of the matrix, yes i have a silver conductive pen, but its not nearly precise enough for massive trace lines, so close together. hmmmmmmmm
Soopa, Doopa, Fooba, Home of the Fooper

Offline Rajagra

  • Posts: 1930
Keyboard Matrix HELP
« Reply #16 on: Wed, 28 April 2010, 21:18:15 »
Yup, I had to repair a couple of breaks. The only way to do it was to paint as carefully as I could - but tolerate where tracks got shorted - and when the paint had half dried, use a pin to scrape off paint between the tracks. It's a trial and error method though. And even when I thought I had it working I later discovered the macro buttons were self-switching to a different macro set. I gave up. Eventually I'll build a new keyboard using the G15 controller and mechanical switches.

Offline Pandawhat

  • Thread Starter
  • Posts: 18
Keyboard Matrix HELP
« Reply #17 on: Wed, 28 April 2010, 22:14:06 »
Praise Raj!
Soopa, Doopa, Fooba, Home of the Fooper

Offline wellington1869

  • Posts: 2885
Keyboard Matrix HELP
« Reply #18 on: Wed, 28 April 2010, 22:16:29 »
Quote from: Pandawhat;177154
Praise Raj!


Maharaj.


"Blah blah blah grade school blah blah blah IBM PS/2s blah blah blah I like Model Ms." -- Kishy

using: ms 7000/Das 3