Sejin SKM-1040 (white)

[marinedalek]

Managed to get this on eBay for the princely sum of 99p! (eBay photo included because I'm lazy... also my desk is a mess!)

I noticed there's some discussion on the wiki about whether the keyswitches are linear or not. Well, on mine at least they are (very) clicky Futaba switches with inverse cross-mount keycaps. Below is an audio clip, which features:
"The quick brown fox jumps over the lazy dog"
Backslash x3
Left Shift x3
K x3 (worn spring)
K x3 Softly (actuation/release noise
Backslash x3 Softly (actuation/release noise)
https://dl.dropboxusercontent.com/u/69960/SKM1040.mp3

As you can hear, K is the odd one out with a very quiet activation point. I don't know if the previous owner hammered the K key a lot or if it's a manufacturing defect, but the listing described previous use as "light".

Another oddity is that inside, roughly half of the electronics are tied to the GND pin on the AT connector, while others (like the clock inhibit for the keyboard scanner!) are tied to the shield of the connector. As none of the adaptors I have pass through any ground connection on the shield, this initially led to no keypresses being reported when tested. I replaced the scanner IC (A 74LS166) and this initially worked, but after roughly 5 minutes of no or light keypresses, it would stop working again as charge built up on the floating shield connector and eventually reached the threshold needed to disable keyboard scanning. Once identified, I fixed this by placing a resistor between GND and the shield connector on the PCB, although in all likelihood a wire link would have worked just as well - the resistor was simply a more tentative approach.

I've been using it for about a week now and it is LOUD. I don't have a Model M to compare it against, but from memory I'd guess it's probably on par.

Has anyone ever come across a keyboard with an odd grounding arrangement like mine apparently has? Clearly Sejin expected the AT shield to be tied to ground on the PC motherboard, but was this part of the standard? Nothing I've found so far suggests so, though I've not dug too deeply.

Also, does anyone know of a source for the (admittedly esoteric) inverse-cross keycaps? For vanity's sake it'd be nice to have dark modifiers, as on the Model M et al.

[lemur]

That is a great looking board... I am not much help as it relates to the grounding concern..

but I will say, unrelated.....as a left handed person (who never played video games) I've always thought this kind of board should have been the norm for right handed people once a mouse was required equipment for a computer. For the work I do, where using a number pad and mouse/nav cluster happens often, it just feels far more natural to have the numpad on the right..

I have an external numpad that i use sometimes, and I've tried putting it where what people would commonly think of as a 'left handed' board would have one, and it just feels weird, both my arms reaching so far to my left..

i say this is a true 'right handed keyboard'

[marinedalek]

I agree! I'm right-handed and decided to try it out for just the reasons you mention. Now the alphanumeric section is central on my desk, with numpad and nav to the left, mouse to the right.

[seebart]

I agree! I'm right-handed and decided to try it out for just the reasons you mention. Now the alphanumeric section is central on my desk, with numpad and nav to the left, mouse to the right.

Me too, I'd love to have the num-block out of the way of my mouse off to the left, looks cool too. These do not show up often though. Good audio clip!

[Mr.Nobody]

Left-handed folk might love this

[balotz]

Futabas are great. The problem you described with the 'K' key is a common one. Someone figured out that it was due to the rubber mat inside the switch warping over time. It's not due to overuse - I have a brand new board which has the same issue.

There is a method which permanently restores the original tactility, without opening the switch. I described it in this thread:

keyboards-f2/futaba-clicky-switch-thoughts-t14941.html

Essentially you remove the keycap and press the slider down with a lot of force, and hold it for a little while.

[terrycherry]

Good to know many people want to buy it. have the SKM-1040 and SKM-2040 lef-hand keyboard with very good condiction if someone want it, I can sell one SKM-2040.

[marinedalek]

Thanks for the suggestion about the K key - I gave it a try and it doesn't seem to make a whole lot of difference, so it may not be the rubber mat that's messed up the click on mine.

However, I've come across another issue. The £1 PS/2 to USB adaptor that I was using worked OK, but had a few quirks such as backslash not being reported and held keys auto-releasing if another key was pressed and released. Because of that I got a Lindy converter which I've seen recommended here and other places as a reliable straightforward converter.

Unfortunately when connected to that, key messages seem to get "delayed". A release message for a tapped key might not get sent for around two or three seconds, unless another key is subsequently pressed which seems to flush the buffer. These stuck keys occur for around 80% of keypresses. I tried a different PS/2 keyboard on the converter and it doesn't seem to happen on that unless you really mash the keyboard, and even then I only got it to happen a few times. Is this a problem anyone else has had? I'm probably going to mod the keyboard to fit a USB cable at some point using a Teensy++, but it would have been nice to have it work reliably until then. A prebuilt Soarer's converter cable wasn't an option due to the extra cost of getting it shipped to the UK.

[balotz]

Thanks for the suggestion about the K key - I gave it a try and it doesn't seem to make a whole lot of difference, so it may not be the rubber mat that's messed up the click on mine.

However, I've come across another issue. The £1 PS/2 to USB adaptor that I was using worked OK, but had a few quirks such as backslash not being reported and held keys auto-releasing if another key was pressed and released. Because of that I got a Lindy converter which I've seen recommended here and other places as a reliable straightforward converter.

Unfortunately when connected to that, key messages seem to get "delayed". A release message for a tapped key might not get sent for around two or three seconds, unless another key is subsequently pressed which seems to flush the buffer. These stuck keys occur for around 80% of keypresses. I tried a different PS/2 keyboard on the converter and it doesn't seem to happen on that unless you really mash the keyboard, and even then I only got it to happen a few times. Is this a problem anyone else has had? I'm probably going to mod the keyboard to fit a USB cable at some point using a Teensy++, but it would have been nice to have it work reliably until then. A prebuilt Soarer's converter cable wasn't an option due to the extra cost of getting it shipped to the UK.

Have you tried the 'blue cube' adapter? It works well for my SKM-1030, a board not dissimilar to yours.

The click will almost certainly return if you press the slider down hard enough. I've done this for every switch on my board and the resulting click and tactility is very sharp.

[marinedalek]

Thanks for the suggestion! I'd forgotten about the blue cube. I ordered one and it arrived today; all issues are now resolved! Now I can start gaming in an obscenely loud way on this deafening keyboard