Kailh X Switches

I was rummaging through my drawer recently, and found a bag of 7 Kailh X switches I have ordered some time ago to try, and never got to it because I was missing the special key caps for them. I decided that maybe it’s time to build something with them, so I ordered 35 more (for some reason they come in packs of 7) and 50 key caps, and designed this simple PCB for them:

As you can see, they are a bit tricky, because they have a rectangular part that’s sunken into the PCB. In addition, to make sure the keyboard is as thin as possible, I used 0.6mm PCB, because that’s how much that part sticks out.

The PCB arrived first, so I soldered the 7 switches I had, and immediately saw problems with them — they wouldn’t contact reliably, or the contacts would remain stuck even after the stem returned to its place. But I decided to reserve my judgement until I get the key caps, since they form an important part of the mechanism.

Then the keys and key caps arrived:

Turns out that the new switches are more reliable than the ones I ordered a while ago. You can tell them apart by looking at that rectangular part that sticks into the PCB — the old switches have sharp corners, the new ones have them a bit more rounded, which works better with how PCBs are CNC-ed.

There are holes in the keyboard from the switches I removed because I couldn’t get them to work reliably. I complained to Kailh, and they promised to send me replacements for them.

While those switches have very innovative construction and are really low-profile — the keyboard is only 6mm thick — they still have a lot of quality problems. I suspect the problem is too much tolerances in the shape of the stem, but they will not always click, and what’s worse, not always contact, depending on which part of the key you press. This varies between individual switches, and is better with the newer ones, but still not perfect — I suspect Kailh simply needs more time to improve the quality.

Finally, a thickness comparison with my daily keyboard with choc switches, to give you an idea:


So thin!
Thank you for sharing that with us :wink:
Great build, too bad that the stwitches are not very reliable…

I hope they improve the design, typical low profile designs are underwhelming since they only reduce the travel by about half a mm. Have you used kailh choc mini switches? They seem to have more travel but appear about as reliable as regular choc switches

1 Like

Yes, you can see that my daily keyboard, to which I compare them at the end, uses the Choc switches. They are great, and I think they are at the moment the best compromise between the low-profile and reliability.

The Choc switches have 3mm of key travel, while the X switches only have 2mm, but I personally like short travel in keyboards, so it’s fine. They are still nicely clicky (when they don’t fail to click, that is).

I’m not talking about choc switches, I’m talking about mini choc switches :slight_smile:
Theyre essentially choc switches with a shorter, recessed stem design. You can find them on ali by looking for kailih switches with a part number of CPG1232. If you haven’t experimented with them yet, you might find them as a good alternative to kailih X switches

Ah, yes, I tried those as well. They were even worse: https://hackaday.io/project/174982-turbot-keyboard/log/184323-friends-dont-let-friends-buy-kailh-sunken-switches

1 Like

That’s disappointing, I’d heard the opposite, but seeing how you laid it out clearly shows why they’re flawed. Great write up!

It may be just my bad luck, and they might have improved since. They do tend to be bad initially and then improve as they get their process right.

1 Like

That’s a shame. There’s little point to buying a mechanical switch if you have to bottom out for them to actuate, more so for a clicky switch :frowning:

I’m sure they will get there in time.

Making reliable, extreme low profile switches is very challenging. Several generations of Apple laptops can attest to that…

1 Like