Why is my ggoold old SSK reggistering the GG key twice?

User avatar
depletedvespene

12 Oct 2017, 13:18

Problem: the GG key registers twice, most of the time. It ocasionally registers once, and every now and then it doesn't reggister at all; these latter two happen mostly if I make an effort to type the problematic key very lightly. As you can imagine, this issue is quite annoyingg.

I fired up Switch Hitter, and this is what it shows (contrast the G key with the F and H):

Code: Select all

08:34.0410 F (0x46, BIOS 0x21) DOWN
08:34.0483 F (0x46, BIOS 0x21) UP -> 72ms
08:35.0675 G (0x47, BIOS 0x22) DOWN
08:35.0763 G (0x47, BIOS 0x22) UP -> 89ms
08:35.0771 G (0x47, BIOS 0x22) DOWN
08:35.0787 G (0x47, BIOS 0x22) UP -> 16ms
08:37.0147 H (0x48, BIOS 0x23) DOWN
08:37.0290 H (0x48, BIOS 0x23) UP -> 143ms
08:40.0731 F (0x46, BIOS 0x21) DOWN
08:41.0026 F (0x46, BIOS 0x21) UP -> 296ms
08:42.0266 G (0x47, BIOS 0x22) DOWN
08:42.0499 G (0x47, BIOS 0x22) UP -> 232ms
08:42.0514 G (0x47, BIOS 0x22) DOWN
08:42.0530 G (0x47, BIOS 0x22) UP -> 15ms
08:43.0458 H (0x48, BIOS 0x23) DOWN
08:43.0674 H (0x48, BIOS 0x23) UP -> 216ms
Indeed, the G key is registered twice, the second "fake" keypress being quite short. I'm sure that those of you that routinely repair keyboards will have a ggood idea of what the underlyingg hardware problem is, so I'll ask: what should I look for? A bent springg? A damaged flipper? Dirt on the board? Corrossion? The SDL/PS-2 cable? The bluecube I'm usingg now?

Thanks in advance, gguys and ggals.

User avatar
wobbled

12 Oct 2017, 16:24

Try a different key on there first of all, i have had keys that didnt hook into the spring very well.
Then try reseat the spring using a cocktail stick or a small screwdriver.
If your key still isnt working and the click of the key actuating sounds as it should then you are likely looking at an issue with the membrane. You can buy replacements from unicomp though so youre not entirely screwed if thats the case

User avatar
depletedvespene

12 Oct 2017, 16:47

wobbled wrote: Try a different key on there first of all, i have had keys that didnt hook into the spring very well.
Then try reseat the spring using a cocktail stick or a small screwdriver.
If your key still isnt working and the click of the key actuating sounds as it should then you are likely looking at an issue with the membrane. You can buy replacements from unicomp though so youre not entirely screwed if thats the case
Thanks!

I see they sell replacement membranes, but I need the part number for an actual SSK. I'll have to see if the membrane (should that be the problem) indicates its own part number.

andrewjoy

12 Oct 2017, 17:15

IT could just be a bit of dirt , fire a few short blasts of canned / compressed air in there

Findecanor

12 Oct 2017, 17:47

Worn-out rivets on the inside? Time for a bolt mod?

User avatar
depletedvespene

12 Oct 2017, 17:51

Findecanor wrote: Worn-out rivets on the inside? Time for a bolt mod?
... it just came back from a bolt-mod. This evening, after I get back home, I'll try out the suggestions I've received so far.

User avatar
Nuum

12 Oct 2017, 18:07

Maybe try to tighten or even loosen the bolts around G, then?

User avatar
depletedvespene

13 Oct 2017, 02:35

Everything seems to be working now!

First, I replaced the SDL-PS/2+bluecube with a SDL-USB cable. No change.

I then opened up the keyboard, took everything apart, checked the flipper under the G key (which seemed fine), and examined the rest of the parts, membrane included... everything looked fine.

I then reassembled everything, a process that took much longer. MY HANDS HURT! I also lost a flipper at some point (thank God I had bought a set of flipper replacements on a whim). Finally, tested every key with Switch Hitter. Everything is working properly, including the G key. The only anomaly (which I don't really know whether is an actual anomaly) is that the Pause key registers its down and up events in 7..11 ms, independently of how long I actually press the key.

Code: Select all

27:46.0596 Print Screen (0x2C, BIOS 0xE037) DOWN
27:46.0677 Print Screen (0x2C, BIOS 0xE037) UP -> 80ms
27:46.0789 Scroll Lock (0x91, BIOS 0x46) DOWN
27:46.0861 Scroll Lock (0x91, BIOS 0x46) UP -> 71ms
27:47.0012 Pause (0x13, BIOS 0xE11D) DOWN
27:47.0020 Pause (0x13, BIOS 0xE11D) UP -> 8ms
27:48.0004 Print Screen (0x2C, BIOS 0xE037) DOWN
27:49.0197 Print Screen (0x2C, BIOS 0xE037) UP -> 1192ms
27:49.0693 Scroll Lock (0x91, BIOS 0x46) DOWN
27:50.0861 Scroll Lock (0x91, BIOS 0x46) UP -> 1167ms
27:51.0236 Pause (0x13, BIOS 0xE11D) DOWN
27:51.0244 Pause (0x13, BIOS 0xE11D) UP -> 8ms
As long as the down-up events are in proper sequence, I can live with that.

It'd seem that Nuum got it right. :-)


There's an oddity about this keyboard that I only discovered during the disassembly: this particular SSK (1391472) has 1-piece keycaps EXCEPT the F and J keys, which are 2-piece (stem+cap). At first, I only noticed this about the F key, which made me think that perhaps someone had switched it, but then I saw that the J key (which does have a small 1 and a front-printed "End" legend) is of the same type. Curious, isn't it?

I'm much relieved now. During the reassembly, I was afraid I wouldn't be able to properly do it and I'd end up disabling or damaging the keyboard. Seems I am NOT that much of a clumsy git, after all...

User avatar
wobbled

13 Oct 2017, 11:32

Glad you got it working man, i didnt realize it had been bolt modded though! Its likely the flipper simply wasnt seated correctly during reassembly. I have had it happen to me a few times, absolute nightmare cos it requires full disassembly unless you are a god with a pair of tweasers!

User avatar
depletedvespene

13 Oct 2017, 11:48

wobbled wrote: Glad you got it working man, i didnt realize it had been bolt modded though! Its likely the flipper simply wasnt seated correctly during reassembly. I have had it happen to me a few times, absolute nightmare cos it requires full disassembly unless you are a god with a pair of tweasers!
Since I don't have a proper "seating" to mount the keyboards innards, the worst part of the reassembly was putting the flippers in place — I had to do it twice, something that was not easy to do with the thick, clumsy fingers I sport. All seems well now, so it was worth the effort. :-)

andrewjoy

13 Oct 2017, 14:03

depletedvespene wrote:
wobbled wrote: Glad you got it working man, i didnt realize it had been bolt modded though! Its likely the flipper simply wasnt seated correctly during reassembly. I have had it happen to me a few times, absolute nightmare cos it requires full disassembly unless you are a god with a pair of tweasers!
Since I don't have a proper "seating" to mount the keyboards innards, the worst part of the reassembly was putting the flippers in place — I had to do it twice, something that was not easy to do with the thick, clumsy fingers I sport. All seems well now, so it was worth the effort. :-)

Now try it with a model F 122 where you have to slide( with a hammer) the backplate on and bend some tabs over it .

User avatar
depletedvespene

13 Oct 2017, 14:19

andrewjoy wrote:
depletedvespene wrote:
wobbled wrote: Glad you got it working man, i didnt realize it had been bolt modded though! Its likely the flipper simply wasnt seated correctly during reassembly. I have had it happen to me a few times, absolute nightmare cos it requires full disassembly unless you are a god with a pair of tweasers!
Since I don't have a proper "seating" to mount the keyboards innards, the worst part of the reassembly was putting the flippers in place — I had to do it twice, something that was not easy to do with the thick, clumsy fingers I sport. All seems well now, so it was worth the effort. :-)

Now try it with a model F 122 where you have to slide( with a hammer) the backplate on and bend some tabs over it .
:shock: Nope nope nope nope nope. :shock:

In my latest buying spree, I also got a couple M122 units off e-bay (one was VERY cheap; the other one, I had to get because it's the ONLY keyboard I've seen that shares my birthday). Both seem in decent condition, but DO neeed some work (the latter one, in particular, seems to be screaming for a bolt mod). After yesterday's work, I am definitely going to hire someone to do those repairs for me.

User avatar
clickykeyboards

13 Oct 2017, 15:52

"There's an oddity about this keyboard that I only discovered during the disassembly: this particular SSK (1391472) has 1-piece keycaps EXCEPT the F and J keys, which are 2-piece (stem+cap). At first, I only noticed this about the F key, which made me think that perhaps someone had switched it, but then I saw that the J key (which does have a small 1 and a front-printed "End" legend) is of the same type. Curious, isn't it?"

I have seen 50+ keyboards with the oddity --- same one-piece key caps for 99 keys (52G9700, 1370477, 1391472, 1392934), but have both the F and J keys be two-piece assemblies.

Theory is that because these are the homing keys and have the extra bump on the front surface to indicate where you should "home" or place your fingers.. that IBM re-used the 2-part key design with key bumps from the other model M lines such as the 1391401.

The first time I saw it, I thought it was just an odd exception.. but after seeing repeated examples over the years with keyboards collected from many different IBM warehouses and non-IBM sources, there was likely a centralized production reason for this feature.

User avatar
depletedvespene

13 Oct 2017, 16:07

clickykeyboards wrote: Theory is that because these are the homing keys and have the extra bump on the front surface to indicate where you should "home" or place your fingers.. that IBM re-used the 2-part key design with key bumps from the other model M lines such as the 1391401.
It's what I thought, too. Thanks for providing extra data to reinforce my theory.

Now, I shall begin the quest for my new holy grail: 1-part F and J (and numpad 5) keys! :mrgreen:

(I'm gonna be laughing my ass off if this evening I find either of those in a "grab bag" I bought off e-bay a couple weeks ago)

Arakula

13 Oct 2017, 19:52

depletedvespene wrote: The only anomaly (which I don't really know whether is an actual anomaly) is that the Pause key registers its down and up events in 7..11 ms, independently of how long I actually press the key.
That's actually no wonder, if you look at the PS/2 protocol.
The Pause key fires one long PS/2 code sequence when it's pressed; it doesn't send anything when it's released. The "Up" for the USB equivalent is artificially created by the PS/2-to-USB-converter, whatever that may be.

Post Reply

Return to “Keyboards”