geekhack
geekhack Community => Keyboards => Topic started by: ObsessionAddict on Wed, 20 February 2013, 13:41:11
-
I'm terribly sorry for this to be my first post on this forum. Trust me, I wish it would be a nice review of my new keyboard or some cunning response to a joke...
After long discussion with myself, I finally decided to spend over 100 euro's on a keyboard... Decided on a KBT pure, brown keys orange backlight. There was only one available in my area, at amazon.fr. (I live in the Netherlands) It was a refurbished, but, I figured, what could possibly go wrong...
I was allready planning to make a review of it with some text about how I first thought it would be stupid to buy such an expensive board, then started reading among other sites, this forum, and becoming more and more obsessed...
That will have to wait... The Q- key is not working, :( the rest is fine, just the Q. (and off course, the SW1, since it is the same key)
I would hate sending it back again, since it was the last one that was available, and then I would probably have to settle with green backlight, or something else completely. So being as careful as I could, I opened her up. The soldering seems to be redone on the Q switch and it seems to make contact, so the problem is probably somewhere else in the board. (even if i shortcut the two pins of the switch, no Q appears on my screen, while this trick will work for any other key)
Any suggestions, does someone know enough of the insides of this board to suggest any other wire-shotcut I could add? Or any other possible source of the problem? Help!
-
The problem doesn't seem to be the switch itself since you can't jump it, so I'd recommend following the lines on the PCB and looking for breaks.
Other than that I'd just recommend not buying newer model keyboards used. Sure a model M from the 80's is indestructible, but I wouldn't have bought a refurbished Pure.
You may be able to get a new one from kbtpure.com or mechanicalkeyboards.com if you send it back.
-
The problem doesn't seem to be the switch itself since you can't jump it, so I'd recommend following the lines on the PCB and looking for breaks.
Other than that I'd just recommend not buying newer model keyboards used. Sure a model M from the 80's is indestructible, but I wouldn't have bought a refurbished Pure.
You may be able to get a new one from kbtpure.com or mechanicalkeyboards.com if you send it back.
^^ this is stupid... time-consuming, and not a sure fix.
solder wires directly from the switch to the endpoints of the traces.
-
It really looks like a multi layer PCD, so tracing the traces seems impossible. I tried jumping to the (possible) endpoints, (over he switch off course) but that didn't seem to work. I was hoping someone had allready done such a thing so I could be sure which pins to connect. A bit scared I will blow the board by accident if I just start connecting pins on good luck.
Guess I will just send it back then. unless someone here comes up with a superplan...
-
If you do decide to return it, as mentioned above mechanicalkeyboards.com does have them in stock, but only with black switches - http://mechanicalkeyboards.com/shop/index.php?l=product_detail&p=108
I just did a conversion to have one shipped to where you are, and it'd be ~109.67 Euro shipped :)
-
Haha, never give up!!
I tried one more jump: from the top pin of Q to the top pin of W, and that worked!! https://www.dropbox.com/s/9zi0cgc7dkzlfsa/2013-02-20%2021.57.49.jpg
I jump soldered a tiny wire between those two pins and everything works like it used to. Look me: qqqQQQQ !! :D :D :D :D
-
Haha, never give up!!
I tried one more jump: from the top pin of Q to the top pin of W, and that worked!! https://www.dropbox.com/s/9zi0cgc7dkzlfsa/2013-02-20%2021.57.49.jpg
I jump soldered a tiny wire between those two pins and everything works like it used to. Look me: qqqQQQQ !! :D :D :D :D
Congratulations!!! :)
-
Haha, never give up!!
I tried one more jump: from the top pin of Q to the top pin of W, and that worked!! https://www.dropbox.com/s/9zi0cgc7dkzlfsa/2013-02-20%2021.57.49.jpg
I jump soldered a tiny wire between those two pins and everything works like it used to. Look me: qqqQQQQ !! :D :D :D :D
I am not sure if bypassing the Diode for Q is a good idea. Holding QWA may make S appear pressed. IIRC, D1 is Q's diode. So you should probably test the continuity between the switch and the diode, and diode and the row.
-
Haha, never give up!!
I tried one more jump: from the top pin of Q to the top pin of W, and that worked!! https://www.dropbox.com/s/9zi0cgc7dkzlfsa/2013-02-20%2021.57.49.jpg
I jump soldered a tiny wire between those two pins and everything works like it used to. Look me: qqqQQQQ !! :D :D :D :D
I am not sure if bypassing the Diode for Q is a good idea. Holding QWA may make S appear pressed. IIRC, D1 is Q's diode. So you should probably test the continuity between the switch and the diode, and diode and the row.
Ahw, you're right... :( aqsw <- this is me, pressing aqw...
Thanks for pointing that out to me! I'm not a gamer, so I should be allright for now, but this is something I want to fix.
It seems like you know what your talking about, so I have a few more questions:
* Should the switch be on the kathode or the anode side of the diode?
* If the connection between the diode and the row is the one that is broken, what would be the best point to connect to the row: is there a point somewhere that is directly connected to the row?
If you can't answer this, is there a "keyboard circuit design 101" or something you could link me to? I have a degree in electrical engineering, so in theory, I should be able to understand the basics of it.
-
Switch connects to the cathode side. (which you can see if you look at D2).
If the connection if broken between the row and diode, theoretically, you can just jump the anode side of Q's diode to the anode side of W.
-
Thank you laffinude! ;D
I guess I'll take out my iron again this evening.
-
Had a similar issue with the 3 on my Pure. The solution was apparently jamming on it countless times with the power of a thousand suns. Works fine now.
-
Had a similar issue with the 3 on my Pure. The solution was apparently jamming on it countless times with the power of a thousand suns. Works fine now.
superman is that you?
-
Ok, I'm lost now. I opened up my Pure, removed the jumpwire I added yesterday, for some reason tried jumping just the Q switch, and it worked! :eek: And I'm still 100% sure that it didn't work yesterday.
In the end I just re-heated the two connections on the switch to re-do them and now I'm the proud owner of a fully working kbt pure...
qaw <-- look even this works...
:confused: Since I don't know what was wrong, I just hope it doesn't happen again. At least I now know what to do.
-
And another update....
An hour after this message, it stopped working again, I made the bridge from d1-k to the q-switch, and now things are working again. But for how long...
-
And another update....
An hour after this message, it stopped working again, I made the bridge from d1-k to the q-switch, and now things are working again. But for how long...
it might be the controller chip as a pin that's "not" soldered down properly.. this happened to someone else's board on here..
-
Doubtful it is a pin. A whole row or column would have problem otherwise. It is probably a bad via.
-
It broke again today at work, but that's probably my bridge. I didn't have any soldering tin handy, so I had to use the tin already on the bord.. Bad idea.
In the meanwhile I stole some soldering tin from the lab at my work today ( and a proper jumpwire), so this weekend I'm re-doing it.
(I'm typing on it all day today, still like the board. Only once I needed a q (twice now), but I still have the old DELL OEM keyboard connected for that.)
-
The last news:
It seems like the D1 diode is broken. I remade my last bridge and it didn't workK, tried to jump to the row, from the anode side of the diode, also didn't help.
But when I jumped just the diode, I got the same behavior as with my first fix.
I'm not sure I can get my hand on such a diode, but I think I can make the same jump as the initial fix, (see dropbox picture, but with a simple diode instead of a wire.
@Laffindude (or someone else who understands the inner workings): you think this is correct? Anything I should make sure first?
-
I'm not sure I can get my hand on such a diode, but I think I can make the same jump as the initial fix, (see dropbox picture, but with a simple diode instead of a wire.
@Laffindude (or someone else who understands the inner workings): you think this is correct? Anything I should make sure first?
I believe that will work fine. The QWAS issue was when it scanned the 2WSX column, and the current went through W switch, the jump, the closed Q to the column, and finally the closed A to the row. A current on that row normally means a closed S switch. Diode should stop the current flowing the wrong way through the jump.