Author Topic: Hypothesis: SA is wrong when used on MX switches  (Read 39999 times)

0 Members and 1 Guest are viewing this topic.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Hypothesis: SA is wrong when used on MX switches
« on: Mon, 22 September 2014, 20:30:25 »
Or phrased in a slightly less inflammatory way, someone on Deskthority speculated (and I agree that this is highly plausible) that SA profile might have originally been intended for some type of keyswitch with a tilted stem.

As evidence, here’s a diagram superimposing SS Family profile on top of SA profile with each SA cap rotated 11°:


Check out this nice photo of a beam spring board from Halvar on Deskthority. Note how the keycaps look basically like SA, but are mounted at an angle to the beam spring switches, giving them a much nicer profile:


Edit: I wonder if SA stands for “spherical angled”, with SS standing for “spherical straight” (or maybe “spherical scupltured”).

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Update:

Daniel Beardsmore directly asked SP about this, and got the following reply from them:

Quote from: Melissa Petersen
The first keycap family produced by Signature Plastics’ previous company, Comptec Inc., was the SA family. These keys had a Spherical touch area and the same profile for All rows.

A few years later the company began producing the SS family which also had a Spherical touch area but with a Sculptured profile i.e. each row had a different keycap angle giving the keyboard a curved look. Because of the limited number of shapes that were tooled, this tooling has been retired and is no longer available for production.

In the mid 80’s an attempt was made to standardize keycaps to a ‘DIN Standard’. DIN stands for “Deutsches Institut für Normung”, meaning "German institute for standardization". This resulted in a new high profile family being produced, the DSS family, which was a DIN standard, Spherical touch, Sculptured key family. This family profile was never very popular and was quickly retired. A short time later the SA family was re-tooled to produce a sculptured look. The keycap family name didn’t change, but it was simply referred to as sculptured SA.

The fourth family tooled was the low profile DSA family. These keys met the DIN standard, had a Spherical touch area, and the same low profile look for All rows.

The DCS family followed shortly. These keys conformed to the DIN standard, had a Cylindrical touch surface, and a Sculptured profile.

The latest keycap family, introduced by Signature Plastics in 2015, is the G20 family. These keys were designed with the gaming community in mind. They have a flat touch area that is wider than standard keycaps, resulting is a smaller gap and easier transition between adjacent keys. These keycaps have the same angle for all rows, similar to the DCS R2 profile.


In short:
SA = Spherical All-rows
SS = Spherical Sculptured
DSA = DIN-compliant Spherical All-rows
DSS = DIN-compliant Spherical Sculptured
DCS = DIN-compliant Cylindrical Sculptured
« Last Edit: Tue, 10 February 2015, 14:03:26 by jacobolus »

Offline jdcarpe

  • * Curator
  • Posts: 8854
  • Location: Odessa, TX
  • Live long, and prosper.
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #1 on: Mon, 22 September 2014, 20:31:39 »
Nope. :)
KMAC :: LZ-GH :: WASD CODE :: WASD v2 :: GH60 :: Alps64 :: JD45 :: IBM Model M :: IBM 4704 "Pingmaster"

http://jd40.info :: http://jd45.info


in memoriam

"When I was a kid, I used to take things apart and never put them back together."

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #2 on: Mon, 22 September 2014, 20:32:06 »
Erm, whoops, I meant to post this in the “keycaps” subforum. Some mod willing to move it?

Offline Puddsy

  • nice
  • * Elated Elder
  • Posts: 11719
  • Location: E NLTSR
  • "Do you shovel to survive, or survive to shovel?"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #3 on: Mon, 22 September 2014, 20:32:35 »
considering it was probably intended for hall effect (the molds exist, and are older than the MX ones), you're probably correct
QFR | MJ2 TKL | "Bulgogiboard" (Keycon 104) | TGR Alice | Revo One | z | Keycult No. 1

if you're actually reading other people's posts on geekhack in 2020 you're doing something wrong

"Everything is worse, but in a barely perceptible and indefinable way" -dollartacos | "Is Linkshine our Nixon?" -NAV (fk u) | "Puddsy is the puddsy of keebs" -ns90

Offline jdcarpe

  • * Curator
  • Posts: 8854
  • Location: Odessa, TX
  • Live long, and prosper.
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #4 on: Mon, 22 September 2014, 20:33:18 »
Hypothesis still wrong no matter where thread located.
KMAC :: LZ-GH :: WASD CODE :: WASD v2 :: GH60 :: Alps64 :: JD45 :: IBM Model M :: IBM 4704 "Pingmaster"

http://jd40.info :: http://jd45.info


in memoriam

"When I was a kid, I used to take things apart and never put them back together."

Offline tjcaustin

  • King Klaxon
  • * Maker
  • Posts: 3556
  • Location: Dallas-ish
  • King of All Klaxon Sciences and Cable Makery
    • Buy stuff
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #5 on: Mon, 22 September 2014, 20:35:44 »
Yeah, because SA is just wrong no matter the switch.

#hatetrain

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #6 on: Mon, 22 September 2014, 20:36:57 »
considering it was probably intended for hall effect (the molds exist, and are older than the MX ones), you're probably correct
Aha! Perfect, that explains it.

So basically, SS is the alternative designed for straight-stemmed switches, but somehow a bunch of geekhackers got stuck using the wrong profile. How did that historical accident come about?

(Also, when can we get some group buys going for SS profile keycaps?)
« Last Edit: Mon, 22 September 2014, 21:00:49 by jacobolus »

Offline esoomenona

  • Gnillort?
  • Posts: 5323
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #7 on: Mon, 22 September 2014, 20:38:03 »
SAs, Bs, and Cs, represent yo' letter.

Offline nubbinator

  • Dabbler Supreme
  • * Maker
  • Posts: 8658
  • Location: Orange County, CA
  • Model M "connoisseur"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #8 on: Mon, 22 September 2014, 20:52:31 »
That could explain why SA feels awful to me.  Could, doesn't mean it does, just it could.

Yeah, because SA is just wrong no matter the switch.

#hatetrain

CHOO CHOO

« Last Edit: Mon, 22 September 2014, 20:54:05 by nubbinator »

Offline HoffmanMyster

  • HOFF, smol MAN OF MYSTERY
  • * Senior Moderator
  • Posts: 10889
  • Location: WI
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #9 on: Mon, 22 September 2014, 20:52:57 »
Interesting theory.  Certainly seems plausible, despite JD's compelling argument.

Check out this nice photo of a beam spring board from Halvar on Deskthority. Note how the keycaps look basically like SA, but are mounted at an angle to the beam spring switches, giving them a much nicer profile:
Show Image


^^ I like this a lot.  I'd love to try typing on such a keyboard.


Also, moved to key caps.   :thumb:

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #10 on: Mon, 22 September 2014, 21:08:38 »
Or here with the black dots is DCS profile:


jdcarpe: Sorry for the flame-bait title. I thought it would get a better response than a boring title. You can feel free to put whatever keycaps you want on whatever keyboard you want, and even if it sucks to type on it won’t affect me personally, so I don’t care. :-)  [But seriously, if you like typing on SA caps on straight-stemmed switches, go for it. I’m sure the tall profile, spherical tops, etc. are nice even when there isn’t much height difference between rows.]

I just thought people might find it an interesting idea, since it helps explain why I (and plenty of other people) don’t like SA or DSA profiles, and maybe it’ll make some folks want to try out SS profile instead, or experiment a bit with other keycap profiles.
« Last Edit: Tue, 23 September 2014, 03:28:35 by jacobolus »

Offline esoomenona

  • Gnillort?
  • Posts: 5323
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #11 on: Mon, 22 September 2014, 22:10:20 »
is the best profile.

Offline jdcarpe

  • * Curator
  • Posts: 8854
  • Location: Odessa, TX
  • Live long, and prosper.
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #12 on: Mon, 22 September 2014, 22:13:26 »
I will make a counter hypothesis:

The very reason that SA feels better is due to the extra angle. Yes, it takes some time to adjust, because years of typing on mediocrity has trained our fingers to feel that a profile like DCS is the more "natural" feeling.

Any chance of getting buckling spring with its curved backplate laid over that image for comparison as well?
KMAC :: LZ-GH :: WASD CODE :: WASD v2 :: GH60 :: Alps64 :: JD45 :: IBM Model M :: IBM 4704 "Pingmaster"

http://jd40.info :: http://jd45.info


in memoriam

"When I was a kid, I used to take things apart and never put them back together."

Offline Elrick

  • Hype Master
  • Posts: 4895
  • Location: CrapTown, Convict Settlement
  • Keyboard Orgasmist
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #13 on: Mon, 22 September 2014, 22:31:07 »
The very reason that SA feels better is due to the extra angle. Yes, it takes some time to adjust, because years of typing on mediocrity has trained our fingers to feel that a profile like DCS is the more "natural" feeling.

The Professor is right.  The only real guiding light in a sea or dark bullsh1t masquerading as fact, this premise of DCS being superior.  Every time I read statements like this makes me hold my SA's with that little more comfort because they deserve to be idolized for being the very BEST Signature Plastics could ever make.

Offline dorkvader

  • Posts: 6284
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #14 on: Mon, 22 September 2014, 22:49:35 »
considering it was probably intended for hall effect (the molds exist, and are older than the MX ones), you're probably correct

I did not know the SP HE moulds are older than their MX ones.

I will do a comparison of HE to SA profile.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #15 on: Mon, 22 September 2014, 23:19:33 »
The very reason that SA feels better is due to the extra angle. Yes, it takes some time to adjust, because years of typing on mediocrity has trained our fingers to feel that a profile like DCS is the more "natural" feeling.
Wait, I think we’re talking about different things. I totally agree that spherical keycaps are great!

Quote
Any chance of getting buckling spring with its curved backplate laid over that image for comparison as well?
Do you have a totally side-on picture? (Or I can try to make one at some point later.)

Personally I think the curved backplate is a nice try, and a great way to get a sculpted profile while using uniform keycaps (making it nice for folks who want to switch to Dvorak or whatever), but ultimately not quite as effective.

At some point in the future I’d like to try to CNC cut some sculpted spherical keycap tops for IBM buckling spring keyboards, to see what effect it has.
« Last Edit: Mon, 22 September 2014, 23:24:35 by jacobolus »

Offline Oobly

  • * Esteemed Elder
  • Posts: 3929
  • Location: Finland
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #16 on: Tue, 23 September 2014, 03:11:55 »
I will make a counter hypothesis:

The very reason that SA feels better is due to the extra angle. Yes, it takes some time to adjust, because years of typing on mediocrity has trained our fingers to feel that a profile like DCS is the more "natural" feeling.

Any chance of getting buckling spring with its curved backplate laid over that image for comparison as well?

^^This. I like SA profile.
Buying more keycaps,
it really hacks my wallet,
but I must have them.

Offline Halvar

  • Posts: 403
  • Location: Germany
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #17 on: Tue, 23 September 2014, 03:30:06 »
I think both jacobulus and jdcarpe have a point.

The SA keycaps on Cherry MX form a totally different, much flatter profile then the designers of the SA profile meant it to have. I think that speculation is very plausible

If you look at DCS, Cherry, OEM and buckling spring profiles, they're basically all stepped profiles, like SA would be on angled stems. But nowadays, we don't tilt our keyboards any more towards us like our ancestors did, who were used to that from typewriters. Instead, we lay them flat, parallel to the table, so a less stepped profile might make sense. Ask the Apple chicklet (or DSA) fans, they even prefer everything to be totally flat. SA on MX is curved and the keys are nicely spherical, but the profile is not stepped, and I think it's no wonder that many people like that.

« Last Edit: Tue, 23 September 2014, 03:58:19 by Halvar »

Offline henz

  • * Exquisite Elder
  • Posts: 1284
  • What?
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #18 on: Tue, 23 September 2014, 06:08:36 »
i love all profiles :P Maybe i like SA a little more though :) Also heard that GMK/cherry doublehsots should be the bomb. Have to get one of those.

Offline SpAmRaY

  • NOT a Moderator
  • * Certified Spammer
  • Posts: 14668
  • Location: ¯\(°_o)/¯
  • because reasons.......
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #19 on: Tue, 23 September 2014, 06:48:38 »
Has anyone asked SP about any of this? I've always had good success with the odd ball random questions being answered.

Offline Halvar

  • Posts: 403
  • Location: Germany
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #20 on: Tue, 23 September 2014, 07:02:21 »
Good idea.

And: does anyone know or even have a Hall effect keyboard that uses Signature Plastics SA?

Offline dorkvader

  • Posts: 6284
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #21 on: Tue, 23 September 2014, 15:42:52 »
Good idea.

And: does anyone know or even have a Hall effect keyboard that uses Signature Plastics SA?

no, but I have a few SA caps from SP that will fit on hall effect.

all the HE boards I have have microswitch caps, but not all were made like that. I think the nokia mikromikko 1 used a third party for the keycaps on it.

edit: appears to be the same profile. Uploading pictures now.

https://www.flickr.com/photos/dork_vader/sets/72157647537881027/


the keycaps on the left show some of the "known" hall effect stems.

grey is the traditional stem that appears on all the oldest kbs
red is the "new" stem that appeared after the mid 1970s
off-white is the stem on the relegendable keycaps on the Wang KB they came from. Not sure if made by honeywell or not.

The 2 on the right are from an SP grab bag

edit: more info to come after I help make dinner. They are impeding kb research!


blue keycap from modifier row. Part of the solarized set. Grey keycap from lower row of wang kb


Both the center row keycaps are square. grey from wang, yellow from solarized. All the alpha keys are the same proflie from the solarized set. both of these look like they have 4 way symmetry


Upper row on wang (grey cap) is the same as the number row of solarized (yellow)


Highest row on both. Grey = wang, yellow = solarized


keytops


all at once


comparing profile of sets.

ok hope this helps
« Last Edit: Tue, 23 September 2014, 16:26:04 by dorkvader »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #22 on: Tue, 23 September 2014, 18:12:49 »
Thanks for the pics! Looks like the SP version is basically the same shape but slightly taller.

Do you have any hall effect keyboards with angled stems, so we can see what these look like in their element?

Offline dorkvader

  • Posts: 6284
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #23 on: Tue, 23 September 2014, 21:56:47 »
Thanks for the pics! Looks like the SP version is basically the same shape but slightly taller.

Do you have any hall effect keyboards with angled stems, so we can see what these look like in their element?

I do and will take pictures of it later.

The ones I have now are on my flickr:
https://www.flickr.com/photos/dork_vader/sets/72157638463953404/

I'll take some better shots for getting the profile, but the above KB has angled stems and a curved profile for keycaps. note: only the alphas are like that. all the function keycaps around them are all flat / non angled.
https://www.flickr.com/photos/dork_vader/sets/72157639678586953/
So does that one.

I will measure it with a level, but I'm pretty sure the SP is not taller, it might look like it from the angle (or lack of angle) though.

Offline dgreekstallion

  • Posts: 215
  • Location: Texas
  • B.S. Fiend
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #24 on: Wed, 24 September 2014, 08:32:37 »
That HE board looks sweet. Does it work?
Recent keyboard fanatic.

Model F-122 convert.

Offline Halvar

  • Posts: 403
  • Location: Germany
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #25 on: Wed, 24 September 2014, 08:38:43 »
Wow, what beauties!  :eek:

TIL I want a Hall effect board.

Offline dgreekstallion

  • Posts: 215
  • Location: Texas
  • B.S. Fiend
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #26 on: Wed, 24 September 2014, 08:50:38 »
Wow, what beauties!  :eek:

TIL I want a Hall effect board.

Same here my friend. I also want to bring my Model F-122 to work but that in itself is a whole 'nother story.
Recent keyboard fanatic.

Model F-122 convert.

Offline MJ45

  • HHKB Pro
  • Posts: 530
  • Location: Chicagoland
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #27 on: Wed, 24 September 2014, 09:14:43 »
Although SA profile are beautiful and nostalgic looking keycaps on MX switches they tend to accentuate the wobbly feel of the stems because of the taller profile. I could not get used to height and the spherical tops on them compaired to DCS profile. For me OG Cherry or GMK is what works best with MX switches.   

Offline Halvar

  • Posts: 403
  • Location: Germany
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #28 on: Wed, 24 September 2014, 09:33:36 »
IBM beam spring boards are much more wobbly than SA on MX, and it's still a better feel.  :)

I don't have a lot of experience with SA on MX though. I have a set of Calm Depth, but its special profile with all-row-3 alphas is too flat for my taste. I think/hope that I will like Round 5 better.





Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #29 on: Sun, 28 September 2014, 02:24:27 »
Here’s what buckling spring caps look like from the side:



(Would be better to take this picture from further away with a telephoto lens, but I don’t have too much light indoors so this’ll have to do for now.)

Comparing to the beam spring profile, the number row is not as tall, and the ZXCV row is not as steeply tilted. (I think the beam spring is better in both cases, and also has a better switch orientation, but could benefit by having even taller number and QWER rows.)
« Last Edit: Sun, 28 September 2014, 02:30:54 by jacobolus »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #30 on: Sun, 28 September 2014, 02:41:56 »
Direct comparison (may not be perfectly to scale):

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #31 on: Sun, 28 September 2014, 02:47:03 »
Although SA profile are beautiful and nostalgic looking keycaps on MX switches they tend to accentuate the wobbly feel of the stems because of the taller profile. I could not get used to height and the spherical tops on them compaired to DCS profile. For me OG Cherry or GMK is what works best with MX switches.   
This is why I’m sorry that DSS profile is not available (it’s been “retired” by SP). I think the wobble is okay though, overall. That is, I think it’s still worth trying SS profile.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #32 on: Sun, 28 September 2014, 19:43:53 »
Bit better picture with a telephoto lens:


And a drawing:


Reasonably similar to DCS:


Noticeably different from beam spring:
« Last Edit: Sun, 28 September 2014, 21:37:00 by jacobolus »

Offline dorkvader

  • Posts: 6284
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #33 on: Mon, 29 September 2014, 14:10:01 »
oh I took a shot side on of that hall effect board.

Oddly enough the wang I had pictures of earlier is all flat.

On these angled stem boards they often (almost always I think) have flat profile for the function keys and sculpted / angled for the alphas. That's why the side keycaps are removed.

Also I took a few pictures at different angles so you can have an "edge on" view of each of the rows.

The trim piece in the background is level. Now you can see the angle this KB was at.


Aligned on bottom row. Sadly the thing mounts to the top piece so it's really hard to get a shot showing the full keycap.



aligned on upper middle row



top row


Let me know if you need more / different.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #34 on: Mon, 29 September 2014, 14:41:56 »
Just out of curiosity, what do the signature plastics SA caps look like when put on that one?

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #35 on: Mon, 29 September 2014, 20:11:32 »
Seems quite close to beam spring profile:


Anyone know what the history is like of such keycap profiles? E.g. were there any mechanical typewriters with this kind of keycap profile, or did it correspond to the switch to electric typewriters?

Offline Findecanor

  • Posts: 4784
  • Location: Koriko
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #36 on: Tue, 30 September 2014, 01:32:55 »
If you look at DCS, Cherry, OEM and buckling spring profiles, they're basically all stepped profiles, like SA would be on angled stems.
I am instead leaning towards the hypothesis that cylindrical profile is actually a simplification of spherical profile at the same stepping. The front edge remains, but the back edge is flattened.
I have posted this image before:
Man must shape his tools lest they shape him
-- Arthur Miller

Offline Halvar

  • Posts: 403
  • Location: Germany
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #37 on: Tue, 30 September 2014, 02:06:30 »
I think that theory just got busted in this thread. If you look at the real beam spring and hall effect profiles, they turn out to be very similar to the more recent cylindrical profiles, at least for the three main alpha rows.

SA on Cherry MX with its non-stepped profile is the real outlier here, much different from all the other profiles, both spherical or cylindrical, because they are all stepped while SA isn't. Turns out that SA on Cherry MX is actually nearer to chiclet (or other flat profiles like DSA) than to the 1970's spherical profiles.
« Last Edit: Tue, 30 September 2014, 02:09:00 by Halvar »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #38 on: Tue, 30 September 2014, 04:36:07 »
There’s supposedly an IBM technical document about this topic, but I can’t find a copy online. Might have to get some help from a librarian or pester someone at IBM:

R. H. Harris, "Curved Keyboard," IBM Tech. Disclosure Bulletin, vol. 20, No. 7, pp. 2784-2785, 12/77.

That document is cited in a patent. Also cited is something called “Honeywell’s Keyboard Handbook”, Code No. 84-20164-0 1072, Section II.

I don’t have any idea where to track down that Honeywell document. Someone want to give it a go?

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #39 on: Tue, 30 September 2014, 04:37:18 »
SA on Cherry MX with its non-stepped profile is the real outlier here, much different from all the other profiles, both spherical or cylindrical, because they are all stepped while SA isn't. Turns out that SA on Cherry MX is actually nearer to chiclet (or other flat profiles like DSA) than to the 1970's spherical profiles.

I am pretty convinced that “SA” stands for “spherical angled” (i.e. intended to be used on switches with angled stems) and “SS” stands for “spherical straight”. I wonder if someone at SP could confirm.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #40 on: Tue, 30 September 2014, 05:31:11 »
Here’s a relevant patent from 1945:
http://www.google.com/patents/US2369807


I love the thumb shift and reasonable size spacebar!
« Last Edit: Tue, 30 September 2014, 05:34:10 by jacobolus »

Offline engicoder

  • Posts: 721
  • Location: North Carolina
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #41 on: Tue, 30 September 2014, 09:29:18 »
Anyone know what the history is like of such keycap profiles? E.g. were there any mechanical typewriters with this kind of keycap profile, or did it correspond to the switch to electric typewriters?

Most mechanical and many electric typewriters are arranged with the top surface of the key cap parallel to the desktop, and each row steps from the one below it, like bleachers. Just an observation, but it looks like angled stems on the honeywells and others may have been an attempt to provide a familiar key arrangement with a flat angled PCB.
   

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #42 on: Tue, 30 September 2014, 13:32:44 »
Most mechanical and many electric typewriters are arranged with the top surface of the key cap parallel to the desktop, and each row steps from the one below it, like bleachers. Just an observation, but it looks like angled stems on the honeywells and others may have been an attempt to provide a familiar key arrangement with a flat angled PCB.
Early electric typewriters and computer keyboards did the same. Then at some point in the 70s they started modifying their “bleacher” type arrangement with a bit less jumping to the bottom row, and with the key tops angled. I suspect someone inside Honeywell, or IBM, (or elsewhere) figured out that the fingers reach for different keys from different angles, so it gets a bit easier to find the top of the key when they’re height-staggered and angled a bit. Then everyone else slowly copied that basic idea, and we had at least slightly sculpted keytops right up until laptops took over and then needed to get thinner and thinner.
« Last Edit: Tue, 30 September 2014, 13:34:26 by jacobolus »

Offline engicoder

  • Posts: 721
  • Location: North Carolina
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #43 on: Mon, 27 October 2014, 12:06:27 »
Out of curiosity, I 3D printed a few adapters to tilt caps at an 11 degree angle and see what things looked like. I don't have any SA caps handy, but DSA is the same concept, so I used those. Here are the results:
80910-0
80912-1

It adds some height to the overall profile, but they don't seem any more wobbly than MX already is. I am thinking about doing a few at other angles with the idea to create some sculpting by adjusting the angles of each row.

   

Offline dorkvader

  • Posts: 6284
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #44 on: Mon, 27 October 2014, 12:20:38 »
Just out of curiosity, what do the signature plastics SA caps look like when put on that one?
I don't have any SA caps for hall effect stems.
Out of curiosity, I 3D printed a few adapters to tilt caps at an 11 degree angle and see what things looked like. I don't have any SA caps handy, but DSA is the same concept, so I used those. Here are the results:
(Attachment Link)
(Attachment Link)

It adds some height to the overall profile, but they don't seem any more wobbly than MX already is. I am thinking about doing a few at other angles with the idea to create some sculpting by adjusting the angles of each row.



that's awesome!
« Last Edit: Mon, 27 October 2014, 12:22:24 by dorkvader »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #45 on: Mon, 27 October 2014, 12:45:13 »
Out of curiosity, I 3D printed a few adapters to tilt caps at an 11 degree angle and see what things looked like. I don't have any SA caps handy, but DSA is the same concept, so I used those. Here are the results:
Very cute!

This doesn’t get the advantage of the slight height increases for the further rows (as you would get on SA), but it does give you one advantage: now there’s a bigger height step between the back of one keycap and the front of the next keycap. As a result, your finger can be straighter while pressing a number row key, and yet you won’t run into the keycaps on the row below.

Offline NeedAFix

  • Posts: 67
  • Location: Aertherwyer
  • Ten Seconds until Midnight
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #46 on: Mon, 27 October 2014, 13:18:42 »
We must revive the curved keyboard industry, our fingers demand it!

-tired of hitting keys below or around other keys-
The Bird of the Hermes is my name,
eating my wings to make me tame.

Offline engicoder

  • Posts: 721
  • Location: North Carolina
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #47 on: Mon, 27 October 2014, 15:57:59 »
This doesn’t get the advantage of the slight height increases for the further rows (as you would get on SA), but it does give you one advantage: now there’s a bigger height step between the back of one keycap and the front of the next keycap. As a result, your finger can be straighter while pressing a number row key, and yet you won’t run into the keycaps on the row below.

What if the adapters for each row were different and the further rows, they were slightly taller?
   

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #48 on: Fri, 31 October 2014, 19:53:45 »
From IBM’s technical article “Curved Keyboard”, 1977,
http://priorart.ip.com/IPCOM/000089774

Quote
Curved Keyboard

It is well known in keyboard designs that the profile or geometry of the arrangement of keytops relative to an operators hand should have an effective dish or curved shape to enhance the operator's ease of access and feel for the keytops. A flat, slanted keyboard having key buttons of identical shape and identical keytop orientation may cause operator discomfort because of the lack of any curved or dishing effect produced along the various planes of the keytops.

This problem has been alleviated in prior designs by using individually shaped key buttons or keystems which vary as to the slant and degree of slope molded into the top of the key buttons by the row in which they are used in the assembled keyboard. This approach creates the desirable dished or curved effect for the operator. It also creates an assembly problem, since individual key buttons or keystems must be designed for each row where they are intended to be used. This fact, combined with the wide variety of font and key top designations which are required, multiplies the inventory of marked key button tops appreciably and complicates the assembly process by requiring careful selection among several keytops with the same nomenclature on them depending on the row in the keyboard in which they are to be used.

The figure illustrates another solution to the problem in which the desirable dished or curved profile is created, not by altering the keytop design, but by bending the entire keyboard itself to create the curved keytop orientation. By this approach, the key buttons may all be of the same cross section and slant, and only the nomenclature need be changed. Special keys are not required depending upon the row in which they are used.

In the figure, the substrate or circuit board 1 is illustrated in dashed lines and in solid lines. The circuit board 1, as illustrated in dashed lines, representes the flat surface board utilized previously. The keyboard 1 shown in solid lines illustrates the same substrate or circuit board, but with the circuit board bent or deformed into an arc of radius R(2). This has the effect of creating a curved or dished profile across the radius R(1) for teh keytops, and permits the design of the keytops to be identical regardless of the row in which they are employed.

Keytops 2A, 2B, 2C, 2D and 2E illustrate, in dashed lines, the varying curvatures and keytop designs which were utilized in the past to create the dished effect for noncurved circuit boards or substrates. The solid lines for the keys 2 show that the keytops may be all of the same profile but that the effective curvature R(1) is still achieved with the curved substrate or circuit board configuration. An additional advantage is that, with capacitive coupling technology, the capacitive pads 3, which are normally employed on the flat circuit board 1, are of increased area, as illustrated by the exaggerated key pads 4, since the arc of curvature in the curved configuration for the curved substrate 1 has greater expanse than the flat configuration for the capacitive pads 3 in the flat circuit board embodiment.

Images are on page 2 if someone wants to spend $40 to see them.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothesis: SA is wrong when used on MX switches
« Reply #49 on: Fri, 07 November 2014, 23:39:11 »


This is all assuming that we’re dealing with a high desk relative to the chair, so we have a steeply angled keyboard; this is not especially ergonomically friendly, but it’s what basically every typewriter and keyboard did and still does, because most people have really bad workspaces set up. For other desk heights, just tilt everything the appropriate amount.

The second to last picture here is roughly what the Maltron / Kinesis Advantage does (though it’s not super accurate... I just sketched it quickly, rather than basing it on direct measurements). The last picture is what I think a keyboard should be like, if manufacturing costs are no object. Note, I think spherical tops are fine too; the positions/orientations of the switches/keycap tops are what I’m focusing on here.
« Last Edit: Sat, 08 November 2014, 02:07:04 by jacobolus »