Switch ideas

aisle9 · October 1, 2023

I'm about to swap out the Gateron brown switches on my 96% Newmen GM1000 keyboard. I love Gateron yellows to the point that they've ruined browns for me, but I'm a little tired of yellows and looking for some variety. Seems like EPOMAKER has a lot of linear switches out there, and the Graywood V3 and Flamingo really have my attention. Akko matcha greens have been on my list for a long time. There are 105 keys on the board, and I'm looking to stay below $45 all-in, shopping on Amazon (possibly Newegg). Basically, I'm looking for something similar but different. Anything out there that's maybe a little more off the beaten path, but would still work?

GarlicDeliverySystem · October 1, 2023

If you are not looking for linears, I would recommend something along the lines of kangaroo or boba U4T switches. Those are rather expensive (especially with a full-size board), but I think Akko Jelly purple should be quite similar (don't have them, just looking at reviews/force curve diagrams) and don't cost as much.

For linears,maybe Akko Jelly Black?

The thing with linear switches is, at least for me, that there is very little quantifiable difference apart from actuation force and 'slope' of the spring. What I mean by the latter is how much the force required to push increases with increase in travel distance.

The rest is much harder to put in numbers or describe via text/video, things like how much wobble you can feel or what the materials make it feel or sound like.

I had a look at the raw data theremingoat has published, and it seems that two extreme cases would be for example Kailh Pro Burgundy (~8.5 gf/mm) and the Akko CS Jelly Black (~2.6 gf/mm), with G Pro 2.0 yellows (KS-9) at ~6.3 gf/mm and G Pro 3 at ~4.5 gf/mm. Oil Kings at ~4.2 gf/mm.

BIG CAVEAT: these fits were taken from the data provided here:

https://github.com/ThereminGoat/force-curves/tree/main

While I trust the reviewer to do their best, I do not know really (nor can I judge) the technical details. Which means I don't know how accurate the values are to the 'truth' (systematic errors), nor how repeatable this is and how much variation there is (precision).

Here is a quick picture, I am talking about the slope of those lines: