deskthority

OK, what's the deal with this?

According to the wiki, "The initial resistance is relatively low, around 30 cN and rising rapidly, bottoming out at around 120 cN."

Cherry MY is not on Plotly yet, but MX Black has a preload of 40 cN and a peak/terminal force of 90 cN:

https://plot.ly/~haata/266

In front of me right now, I have a Type 1 MY actuator from seebart, a Type 3 MY actuator from RetroClinic (from a BBC Master), and a modern MX Black switch.

All three feel largely the same. Type 1 MY is the smoothest. The feel of the MX switch does seem to correspond with the heavier preload.

However, in isolation (removed from the keyboard) the MY actuator really does just feel like a typical linear switch, at the lighter end of the weight range. (SMK J-M0404 on the other hand is pretty stiff.)

I measured a three-layer membrane assembly from a modern (USB) Chicony keyboard as having something like 13 to 19.5 g actuation force, and the total travel will be around 0.09 mm, the thickness of the centre membrane.

With that said, the MY actuator appears to be designed to provide mid-travel actuation, so this will have some effect on the travel. If the feel really is vastly different from an MX switch, then I'm guessing that the deviation in the force curve comes from where the membrane is engaged.

Even then, I don't see why Type 1 and 3 would be so different — maybe there was a significant change in the membrane material or thickness, since MY is meant to use vinyl membranes, while normally it's suggested to be Mylar (which is what Datanetics were using in the 70s).

Well, it's two springs rather than one, right? Surely that will impact keyfeel. Besides, the coil spring in MY switches is much smaller and shorter than MX springs. As for why type 1 and 3 are so different, maybe there's an analogy between them and modern and vintage MX switches.

I am truly ignorant in this department, however, I once bought an MY keyboard that had been modded to remove 1 set of its springs.

The keyboard was interesting for a few minutes because it was so incredibly light (the claim was something on the order of 15-20 but I never bothered to test it) that it was frustrating for me to use. On the other hand, I suppose that one could become accustomed to it and eventually find it rather pleasant. Particularly a person with some sort of physical limitation.

Daniel Beardsmore wrote: According to the wiki, "The initial resistance is relatively low, around 30 cN and rising rapidly, bottoming out at around 120 cN."

I believe I got those figures from a force graph posted on Geekhack long ago, which had been linked to from the 2011 Geekhack Wiki. Unfortunately, the original post and image are gone; I can't even find them on web.archive.org.

Chyros wrote: Well, it's two springs rather than one, right? Surely that will impact keyfeel. Besides, the coil spring in MY switches is much smaller and shorter than MX springs. As for why type 1 and 3 are so different, maybe there's an analogy between them and modern and vintage MX switches.

The coil spring size is irrelevant (and it's not length that matters but material, wire gauge and turn density). Longer springs are only "heavier" because they're fitted into the same space as a shorter spring; for y=mx+c, you're simply increasing c, while the alleged figures for MY had a lower c (preload) and greater m (weight).

I didn't remove either the coil spring or the leaf spring for the test: that would be pointless. What I wasn't testing was blocking the motion of the leaf spring. The feet of the switch are in fact the exact distance apart as the tail of my caliper, so I can use that as a makeshift membrane assembly.

The leaf spring contains two folds. The first fold is simply to create the retaining tab. The second fold creates the point where pressure is applied to the membrane: this fold delimits the short horizontal portion of the leaf. If you rest the actuator on its feet, the leaf spring is flexed entirely about the first fold, which with the helical spring gives you a linear switch with reasonable (Type 3) to excellent (Type 1) feel and the same weight as MX Black.

However, if that second fold is blocked in, then the horizontal portion of the spring can't move (in reality it will move less than a tenth of a millimetre), and thus the long tail of the leaf spring is bowed by the slider. This is when the switch gains its "dead octopus" feel, and indeed it does feel bad.
A head-to-head press with MX Black shows MX Black to be heavier at first, and then lighter, which is consistent with MY having a lower preload and heavier weight.

As for Type 1 vs 3, I would say that Type 1 is heavier and more spongy feeling: Type 3 has a lighter and cleaner feel, although it still feels terrible. This is curious as the consensus of opinion holds Type 1 to be superior. Type 1 is certainly smoother, and if I hold them to my ear and press them slowly, both MY types are virtually silent, with MX Black being distinctly louder, and MX Linear Grey being much louder with a clear scraping sound.

Findecanor wrote: I believe I got those figures from a force graph posted on Geekhack long ago, which had been linked to from the 2011 Geekhack Wiki. Unfortunately, the original post and image are gone; I can't even find them on web.archive.org.

The sad thing is that people are still adding comments to the wiki with no references and no photos, and therefore no way to determine if what they said is true or not.

However, my empirical tests appear to broadly consistent with those claims. I don't recall Jacob mentioning that he has any G81 keyboards or MY typewriters, but hopefully MY will get graphed eventually and we'll have our evidence.