geekhack
geekhack Community => Keyboards => Topic started by: jpc on Thu, 14 October 2010, 22:29:52
-
So I swapped a 71G4644 controller into a buckling spring SSK. Pics on this thread. (http://geekhack.org/showthread.php?t=12119)
It works. Except that very often, I will capitalize one more character than intended -- "IT" instead of "It" for example.
What's going on?
Possibility #1 -- The shift key is actually sticky
Possibility #2 -- I'm making typos
Possibility #3 -- Rubber dome and buckling spring controllers differ
When you touch type, you're likely depressing the "t" in "It" while releasing the Shift key.
On a rubber dome board, you've already released the Shift key before making contact on the "t" key, because the make/break point is down on the floor of the mariana trench, where the keys bottom out.
On a buckling spring board, with exactly the same motions, you could make contact on the "t" key before breaking contact on the Shift key, given the high make-contact and break-contact points.
Maybe the buckling spring controller helps you to avoid capitalizing that second letter? (And the rubber dome controller does not.) That's my guess. How would extra-capitalization-avoidance work? Maybe it delays reporting the keypress event? Gamers wouldn't like that...
Does anyone game on buckling spring? Probably a few die hards. "Oh dudes remember Doom? I've got Shotgun, Chaingun, ratatatatatatatata ... oh that was the MOdel F..."
(there! it just happened again! Seeee??)
Um and how the heck would you test this quantitatively...
-
OR could this be an NKRO limitation in the 71G4644 controller?
^^^ there it happened again. I typed "Or", promise.
-
Scratch the NKRO theory, there /should/ be no difference in the rollover combinations that work on a rubber dome and buckling spring M, given that they have the same membrane patterns.
I'm going to try two experiments:
* DItch the PS/2-USB adapter, and use a real PS/2 port, see if that helps
^^ zomg Ditch not DItch
* Else put a buckling spring controller back into this board, see if that helps
-
Hi, I'm from MAssachusetts
The real PS/2 port does not make a difference.
-
I think I'll chime in, as dumb as it may sound... disconnect the controller and re connect it.
I did a controller swap many months ago and did not firmly attach the membrane into the controller board and some keys would not register and the caps lock would not work ect.
give that a shot!
-
Beautiful foliage now, yes. There's one tree in my yard that always turns bright red and it's brilliant today.
This was bugging me so I measured. Set up the computer to beep at each keypress. Then took an audio recording of the buckling sound and the beep. I measured the time from the buckle to the end of the beep to maybe +/- 5ms precision.
This did not show any difference between the rubber dome controller and buckling spring controller. With or without the shift key depressed. So I don't think the BS controller delays reporting keypress events. Gamers can relax.
Remaining possibilities...
* I can't type anymore; early onset senility
* It's a sticky shift key
* The 71G4644 controller is crappy like every other aspect of the 71G4644. It takes occasional breaks from scanning to have a hit of crack. Returns 50ms later, rescans, and reports a set of changes that is possibly out of order. I am completely making this up if you can't tell.
-
Uh.... you did not get 5ms precision.
I'm claiming milliseconds not microseconds. It was 44Khz recording, the beep was 400Hz, the buckle and the end of the beep were pretty well defined on the waveform
-
The controllers should be identical. Even if the switch mechanism is different, the contact mechanism is identical.
-
Do the fullsize controllers definitely work with the SSKs?
-
try the other shift button?
-
THis is something I haven't noticed with my SSK with a 1391401 controller. BUt I will keep an eye on it for you...
-
DAmn I was looking for a good excuse for always capitalizing my second letter alone with the first.
I do this on every keyboard I have (I think) =(
I guess I am defective, send me back and get the refund.
-
Okay I am officially on my mini for the rest of the day to verify this. Tried 20 times to get it to happen and it's behaving correctly.
-
Correct, I am using a 101 Model M BS controller. I just wanted to make sure it's not an issue using a 101key controller in a SSK.
-
MissileMike: I don't think the problem happens with a 1391401 controller. I replaced the 71G4644 controller with a 1391401 controller, and now the board is behaving itself.
I can type caps without typing double caps, watch this
Jenny Jenny You're The Girl For Me
You Don't Know Me But You Make Me So Happy
Whee!
The wonky controller is from a 71G4644 rubber dome board built 1/26/96 and it is tan, whereas every other green Model M controller I've seen is green. Pics of the wonky one. (http://geekhack.org/showpost.php?p=233493&postcount=19)
-
Okay, well that's not so bad then... A little bit of a bummer... Funny how the 1391401 board doesn't seem like it will fit due to being too small, but when the case closes and the top goes on, it's tight as a drum.
-
Like Kishy says it could be psychological. Kinda like Welly's Filco Zero tests.
Yeah agreed.
Maybe I'll hook an oscilloscope into the membrane connector and see if the scan rate can be measured. It would be better to have a quantifiable finding.
On the other hand I feel really lazy right now. Is there a Friday night beer thread here somewhere?