Feel free to use my photos for any purpose. I happily release them under Creative Commons CC-BY-3.0 license.
I’ll try to deconstruct green, yellow, and cream switches and take photographs/measurements of the spring sometime soon. I haven’t desoldered the green switch to look inside yet, but it’s clear that the yellow and cream switches have different springs: the cream switch has fewer winds [making it stiffer]. My guess is that the spring is the only difference (or at least the primary difference) between these three types of switches.
* * *
Completely off that subject, I’ve been experimenting with mixing/matching/modding old alps switches. The keyboards I have with blue and white clicky switches both are pretty inconsistent from switch to switch [I think a combination of wear, dirt, age, etc.] Anyway, the blue keyboard actually feels pretty nice to type on despite the inconsistencies, so I’ve been playing with some of the white switches.
I found that if I take the housing, spring, and slider from a dampened tactile cream switch [from AEK II], and the click leaf from a white switch, bend that click leaf slightly so that its "back" is straight [making it clickier; I suspect these started out straight and have deformed with age], and cut a few winds off the spring, I get a very nice feeling clicky switch, about the weight of the blue switches, with a louder click, and no bottom-out sound. I have only done this to a couple switches so far, trying out a few possibilities. The next thing I want to try is using one of the springs from a linear cream switch, since those are a bit stiffer, but then cutting some winds off it. This will get me a switch that requires lighter force at the start, but then gets heavier toward the bottom of a keypress. I suspect that I will prefer this to the current springs.
I hope that HaaTa’s force curve measurement device gets up and working soon. My current theory is that my criteria for keyswitches include: (1) low amount of work to reach actuation point [i.e. the integrated force to press the switch until it actuates], (2) sharp drop-off in force at tactile point, so that (3) actuation precisely at the tactile point, (4) at least a slight audible click at actuation, (5) increasing force near the bottom, so that landing is cushioned after actuation. It would be nice to try to tweak a switch based on these criteria, measure the results empirically, and then test the feel to compare.