Better Mechanical Keyboards Through 3D Printing

You’re not cool unless you have a mechanical keyboard. No, you won’t be able to tell if your coworkers don’t like it, because you won’t be able to hear their complaining over the sound of your clack-clack-clacking. You can even go all-in with switch modifications, o-rings, and new springs, or you could use your 3D printer to modify the touch of your wonderful Cherry MX switches. That’s what a few researchers did, and the results are promising.

The ‘problem’ this research is attempting to solve is bottoming out on Cherry MX keyswitches. If you’re bottoming out, you’re doing it wrong, but nevertheless, you can get a publication out of solving repetitive strain injury. This was done by modeling the bottom housing of a Cherry MX switch by printing most of it in nylon on a Stratasys Objet 350 polyjet printer, with a tiny bit of of the housing printed with a polymer with a hardness of Shore 40. No, Shore A, Shore B, or Shore 00 was not specified, but hey, it’s just a conference paper.

The experimental test for this keyswitch was dropping a 150 gram weight from 125 mm onto the keyswitch, with a force sensitive resistor underneath the switch, connected to an Arduino. Data was logged, filtered, and fitted in Excel to create a plot of the force on dampened, rigid, and commercial switch housings. Results from ANOVA were p > 0.05 (p=0.12).

Despite the lack of significant results, there is something here. The Objet is one of the few printers that can do multimaterial printing with the resolution needed to replicate an injection molded part. There is a trend to the data, and printing squishy parts into a keyswitch should improve typing feel. There will be more work on this, but in the meantime we’re hopeful some other experimenters will pick up this train of research.

14 thoughts on “Better Mechanical Keyboards Through 3D Printing

  1. Measuring a keyswitch with some complex experimental protocol involving a 150 gram weight, precision sensors and an arduino (and even then getting vague results) reminds me of audiophile equipment woo, and the fact that people design machines for humans using metrics that humans most likely can’t discern using their own senses, then fool themselves into thinking they can through cognitive bias. I doubt there’d be any way to tell the difference in a double-blind test typing an actual document.

    But hey, I understand the pursuit despite that. We aren’t rational beings. Everyone do what makes you happy.

    Has anyone ever thought of putting the whole keyboard on a big, shock-absorbing rubber mat? Does it have to be built into the individual switches?

    1. A shock-absorbing rubber mat would of course help but I think the real issue as Benchoff mentioned is that if you are bottoming out when you press the switches, you’re doing it wrong.

      1. My cherry MX brown keys bottom out during, er, extreme coding stints. Stock standard keyboard and switches. Do you (and Brian) mean my typing is wrong (very likely ;) ) or that you are modding the keys badly?

        1. I used to use browns but I tried using red linear switches and have never looked back. I find that its a lot harder to bottom out on linear switches. If you ever get a chance to try red I think you will like them for extreme coding.

  2. the only mechanical keyboard i own is one i found in the dumpster some years ago. half the keycaps are missing but it still works great. ive not had much luck 3d printing keycaps. they fall off and become cat toys. ive been meaning to clean it up and bring it down to my main rig, but cleaning is not something i get excited about.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.