Author Topic: fixing the ergodox thumb section  (Read 151415 times)

0 Members and 2 Guests are viewing this topic.

Offline davkol

  •  Post Editing Timeout
  • Posts: 4994
Re: fixing the ergodox thumb section
« Reply #50 on: Fri, 19 September 2014, 06:26:10 »
Let's be specific, WHO will manufacture and distribute it? Because it took about two years to get ErgoDox from an overpriced DIY kit to at least a readily-available DIY kit with optional assembly, and that's mostly thanks to what some people in the community do in their free time. It's a problem similar to preinstalls on desktops.
Well, for example, I think if the design seems better (especially if it seems like people would prefer it given the option), then it should be possible to convince MassDrop to distribute it. I’ve talked to them about other ergonomic keyboard ideas in the past, and they seem pretty receptive to trying new things.
They don't even have the updated normal PCB design yet.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #51 on: Fri, 19 September 2014, 06:54:04 »
Also, these angles and bigger thumb keys look like made for typing with resting wrists.
I can’t parse this. But anyway, I don’t think palm rests are at all necessary, but might be helpful for certain amounts of tenting.

They don't even have the updated normal PCB design yet.
What’s that? Updated by who, when, how? Does it have some substantial advantage compared to the previous one?

That doesn’t really prove anything. They have a very small staff, not including anyone directly doing keyboard engineering, and also not including anyone compulsively reading geekhack. Have you tried reaching out to them about this updated PCB design?

In any event, this is getting fairly off-topic IMO.
« Last Edit: Fri, 19 September 2014, 07:00:16 by jacobolus »

Offline davkol

  •  Post Editing Timeout
  • Posts: 4994
Re: fixing the ergodox thumb section
« Reply #52 on: Fri, 19 September 2014, 07:37:08 »
Also, these angles and bigger thumb keys look like made for typing with resting wrists.
I can’t parse this. But anyway, I don’t think palm rests are at all necessary, but might be helpful for certain amounts of tenting.
How would you press the bigger thumb keys? With joints, or fingertips?

They don't even have the updated normal PCB design yet.
What’s that? Updated by who, when, how? Does it have some substantial advantage compared to the previous one?

That doesn’t really prove anything. They have a very small staff, not including anyone directly doing keyboard engineering, and also not including anyone compulsively reading geekhack. Have you tried reaching out to them about this updated PCB design?
Someone has forked the PCB design, fixed some technical stuff and added support for backlighting (which is something many people want, although I don't really get it). MD supposedly worked with him on pushing it to actual users... well, that was many months ago.

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: fixing the ergodox thumb section
« Reply #53 on: Fri, 19 September 2014, 08:47:26 »
Let's be specific, WHO will manufacture and distribute it? Because it took about two years to get ErgoDox from an overpriced DIY kit to at least a readily-available DIY kit with optional assembly, and that's mostly thanks to what some people in the community do in their free time. It's a problem similar to preinstalls on desktops.
Well, for example, I think if the design seems better (especially if it seems like people would prefer it given the option), then it should be possible to convince MassDrop to distribute it. I’ve talked to them about other ergonomic keyboard ideas in the past, and they seem pretty receptive to trying new things.
this could work. Call it the ergodox 2.0

I thought I'd posted already, but of the layout in the image in the OP,

I like the last one the best followed by the upper right. I'm not sure about the two wider keys for thumbs though I think they'd be harder to hit like that, but they are certainly not as "wide".

I have pretty flexible hands (not like paganini but close) so I don't have too much issue with reaching the thumbs.

They don't even have the updated normal PCB design yet.
What’s that? Updated by who, when, how? Does it have some substantial advantage compared to the previous one?

That doesn’t really prove anything. They have a very small staff, not including anyone directly doing keyboard engineering, and also not including anyone compulsively reading geekhack. Have you tried reaching out to them about this updated PCB design?
Someone has forked the PCB design, fixed some technical stuff and added support for backlighting (which is something many people want, although I don't really get it). MD supposedly worked with him on pushing it to actual users... well, that was many months ago.
even worse, they are not even using the most updated official PCB from the ergodox website! It has support for PCB mount stabs. The leeku clone PCB that MK has uses that one.
« Last Edit: Fri, 19 September 2014, 08:49:26 by dorkvader »

Offline Analogy

  • Posts: 8
Re: fixing the ergodox thumb section
« Reply #54 on: Fri, 19 September 2014, 15:59:21 »
The fundamental problem with all of these thumb clusters IMO is that they're on the same plane as the finger keys, when the thumb articulates in a plane 90 degrees to the rest of the fingers. The solution is to turn the thumb cluster vertical so that the thumb is pressing buttons by curling inward, like the rest of the fingers do. Designing the keyboard this way gives you the correct tenting angle for free, uses less space on the desk, and gives your thumb access to far more keys than any other design I've seen, with far less movement needed.


Offline Analogy

  • Posts: 8
Re: fixing the ergodox thumb section
« Reply #55 on: Fri, 19 September 2014, 16:02:04 »
Downside: More complex build, needs more PCBs and more individual case elements, and a ribbon cable to connect the thumb cluster back to the main keyboard. But it seems worthwhile to me?

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #56 on: Fri, 19 September 2014, 16:49:26 »
The fundamental problem with all of these thumb clusters IMO is that they're on the same plane as the finger keys, when the thumb articulates in a plane 90 degrees to the rest of the fingers. The solution is to turn the thumb cluster vertical so that the thumb is pressing buttons by curling inward, like the rest of the fingers do.
This is how Oobly’s DIY keyboard is set up, but personally I’m convinced it’s suboptimal design. A completely flat half is also suboptimal, in a different way. I’m not convinced that either is especially better than the other. I think the Maltron is better than either one (but very expensive to mass produce), and I think there are alternative designs that would be better still (but would again be difficult to mass produce).

To understand why requires thinking about hand and thumb anatomy. When you make a squeezing motion with your thumb in toward your palm, you mainly flex the metacarpophalangeal joint (that is, the second joint from the fingertip). The muscles/tendons that flex this joint are not nearly as strong, and the joint itself is not nearly as flexible, as the carpometacarpal joint of the thumb (for one thing it doesn’t have as long a lever so it gets less fingertip motion for the same amount of muscle contraction; for another, the muscles are physically weaker and smaller). If you do some detailed study of what positions and motions of your thumb have the best strength and agility, I think you’ll find that pressing roughly downward (like the motion you’d use on a standard keyboard spacebar) is quite a bit stronger than the squeezing motion you’d use on your proposed setup. Finally, the squeezing motion can’t benefit from some slight assist from a rotating wrist that you sometimes get with a downward-ish press.

The same is true for the finger joints: we almost exclusively flex the carpometacarpal joint to actually press finger keys, and we extend or flex the distal two joints (particularly the metacarpophalangeal joint) in between keypresses to locate various key-tops. If we were using the fingers in the same way as you suggest using the thumbs, the keys would all be pointed such that you’d press them by squeezing toward the body like a trigger, rather than pressing downward.
« Last Edit: Fri, 19 September 2014, 16:55:43 by jacobolus »

Offline davkol

  •  Post Editing Timeout
  • Posts: 4994
Re: fixing the ergodox thumb section
« Reply #57 on: Fri, 19 September 2014, 17:00:36 »
The thumb can move *forward* just like other fingers though, i.e. not lying flat on the keycap, or the "squeezing" motion can happen just like when clenching ones fist.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #58 on: Fri, 19 September 2014, 17:02:55 »
The thumb can move *forward* just like other fingers though, i.e. not lying flat on the keycap, or the "squeezing" motion can happen just like when clenching ones fist.
I don’t understand what you mean. (Can you explain either in technical anatomy terms, or draw a picture, or something?) Which direction is “forward” in this context?

Offline davkol

  •  Post Editing Timeout
  • Posts: 4994
Re: fixing the ergodox thumb section
« Reply #59 on: Fri, 19 September 2014, 17:43:56 »
Take a look at trackballs: Logitech TrackMan Marble (angled) and Kensington Slimblade (flat). Buttons can be pressed with fingertips  or whole fingers in either case. Clawing the marble is indeed a potential issue. So is pressing slimblade buttons with whole thumbs (down-up motion), while it isn't a problem on the marble (in-out motion, gripping).

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #60 on: Fri, 19 September 2014, 18:34:56 »
I still don’t understand what your point is.

(Based on my own relatively limited experience with these devices and what seems like a natural motion to me...) Logitech’s finger trackball, and Kensington’s trackball, are mostly spun side to side via whole hand movement based on rotation at the wrist; they’re spun forward back via a combination of wrist and finger movement. None of these movements are similar to keyboard presses.

Logitech’s thumb trackball is spun forward/backward using a combination of whole-hand movement and movement of all 3 thumb joints; it’s spun side to side almost exclusively by the carpometacarpal joint. Spinning the thumb trackball to the left is sort of similar to motion used for a standard keyboard spacebar press. The other thumb trackball motion is not similar to any of the keypress motions we’ve been talking about in this thread.

All these trackballs’ buttons have nearly no travel, making them substantially different from keyboard keys. Personally I think both the kensington and logitech finger trackballs have pretty bad button positioning and orientation, while the logitech thumb trackball is okay (but I’d prefer longer travel lower force switches).

Maybe you can explain your thinking in a bit more detail / provide some more contextual description about your thoughts about mouse/trackball ergonomics or your experiences? I’m having trouble figuring out how this relates to our discussion of keyboard keys.
« Last Edit: Fri, 19 September 2014, 18:40:24 by jacobolus »

Offline plainbriny

  • Posts: 192
  • Location: Taiwan
Re: fixing the ergodox thumb section
« Reply #61 on: Fri, 19 September 2014, 21:34:00 »
This whole thread is very interesting.

However, the major issue is that many proposed changes require redesign of the PCB and case.

Does anyone willing to take the job?

As for angled or flat design, I have 3 ergodoxes in use now, one with the original design, with 2x thumb keys, the second one is the one with angled thumb cluster, the "frankendox", and the third one is a 80 key variant of the original, with the 2x thumb keys replaced by 2 1x keys.

I have to say, these are all not optimal, but I can live with them. So far I prefer the angled design, but it is more complicate to make than the flat design.

I have no idea what the optimal design would be, however, at least to me, the major issue is not the thumb cluster, but the pinky finger.

======= below has nothing to do with thumb cluster, you can just ignore them =======

The problem is that with the current layout (qwerty), my left pinky is too heavily loaded. There are two approaches to solve my problem, one is to redesign the layout, but I would need some to retrain my finger and this would be a long process. I don't have time for the time being. The second means is to adjust the staggering of the pinky column, because I use the '1', 'q', 'a', 'z' keys a lot, yet to reach '1' and 'q', the left pinky finger need larger movements. Thus lower the pinky column can shorten the moving distance of the pinky. However, this require the redesign of the PCB, which is beyond my ability now.

==========================================================

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #62 on: Fri, 19 September 2014, 22:32:38 »
However, the major issue is that many proposed changes require redesign of the PCB and case.

Does anyone willing to take the job?
That’s correct. If we can come up with a design that people like, I’m happy to design a lister-style case, and I’m quite confident I can convince someone to help out with the PCB. Making a PCB takes work, but it’s relatively straightforward work, without too many design decisions. Figuring out the design to make is the harder part to get right.

Quote
I have no idea what the optimal design would be, however, at least to me, the major issue is not the thumb cluster, but the pinky finger.

The problem is that with the current layout (qwerty), my left pinky is too heavily loaded. There are two approaches to solve my problem, one is to redesign the layout, but I would need some to retrain my finger and this would be a long process. I don't have time for the time being. The second means is to adjust the staggering of the pinky column, because I use the '1', 'q', 'a', 'z' keys a lot, yet to reach '1' and 'q', the left pinky finger need larger movements. Thus lower the pinky column can shorten the moving distance of the pinky.
You’ll notice that some of my mockups have the pinky columns shifted down a bit compared to the current design. The optimal spot is probably almost a full key below the ring finger column, but I’ve heard from at least two people who didn’t like quite so much stagger, and I wanted to stay relatively close to the original Ergodox design, so I only shifted it down by a little bit.

The other very helpful change is to use keycaps for the further away rows which are taller than the keycaps for the nearer rows. For instance, that’s what I was indicating with the colors in mockups like this one:



(this is a diagram slightly more relevant to a standard keyboard, but you get the idea)

Even better would be to have all the pinky keys raised up higher than the other rows, the way e.g. the Maltron has, but that requires abandoning the single flat PCB/plate, and makes a keyboard much more difficult to manufacture. But just shifting the pinky row closer to the body by half a key and using taller keycaps for further-away rows is a big improvement.
« Last Edit: Fri, 19 September 2014, 22:52:00 by jacobolus »

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: fixing the ergodox thumb section
« Reply #63 on: Sat, 20 September 2014, 12:33:55 »
To understand why requires thinking about hand and thumb anatomy. When you make a squeezing motion with your thumb in toward your palm, you mainly flex the metacarpophalangeal joint (that is, the second joint from the fingertip).

This is my issue (with people's proposed ergo designs): so many people's hands are different, you can't make an ergo KB for them all, or even most. I think the only thing you can do is just make a KB individualized to each person.

For example, I just tried this and notice that my second joint from the tip does not move much at all. Almost the entirety of my thumb's motion comes from moving the other two joints. My left hand, which is significantly more flexible is different: there' I have enough mobility in my second joint to do that sort of motion as you describe.

Offline PieterGen

  • Posts: 135
Re: fixing the ergodox thumb section
« Reply #64 on: Sat, 20 September 2014, 14:33:39 »
...there are alternative designs that would be better still ...If you do some detailed study of what positions and motions of your thumb have the best strength and agility, I think you’ll find that pressing roughly downward (like the motion you’d use on a standard keyboard spacebar) is quite a bit stronger than the squeezing motion you’d use on your proposed setup. Finally, the squeezing motion can’t benefit from some slight assist from a rotating wrist that you sometimes get with a downward-ish press.


^ Agree ^ Good explanation. Thanks by the way for all thoughts & sketches, they really help me, as a source of inspiration, in the design of my own keyboard.

Edit: The silly USA patent system lets keyboard ideas be patented. Yep, hand separation, tenting and angling are patentable "break through inventions", apparently....  Anyway, I found a patent that was issued to a certain Redmond based patent giant, that nevertheless contains interesting figures. Of field test with respondents who could indicate their optimal degrees of tenting etc.   http://www.google.com/patents/US7338224

« Last Edit: Sat, 20 September 2014, 15:35:12 by PieterGen »

Offline PieterGen

  • Posts: 135
Re: fixing the ergodox thumb section
« Reply #65 on: Sat, 20 September 2014, 14:49:57 »
We have had some discussion on whether or not bigger (than 1x1) keys are needed for the thumbs. I'd now like to pose the same question for the pinky keys.  ;D

Fact #1: The part of the pinky that touches the keycap has a small surface. Smaller that that of other fingers
Fact #2: The bones in the pinky (called the phalanx bones: the distal phalanx, carrying the nail, the middle phalanx, and the proximal phalanx) are shorter than the bones in the other fingers. Meaning: if you flex or extend the pinky, that little finger does not move far. In other words: both the bottom row key and the top row key should be closer to the how row key. 

Smaller keycaps, more closely together (vertically) might solve the problem. We will get away with smaller keycaps, since our pinky tips are smaller; and it givces the possibility to put them closer together. A problem is though: are there smaller keycaps on the market for Cherry MX switches? And is it possible tou mount the switches closer together?

Thoughts? Or am I solving a  non-problem?

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #66 on: Sat, 20 September 2014, 16:40:11 »
Smaller keycaps, more closely together (vertically) might solve the problem. We will get away with smaller keycaps, since our pinky tips are smaller; and it givces the possibility to put them closer together.  A problem is though: are there smaller keycaps on the market for Cherry MX switches? And is it possible tou mount the switches closer together? [bold added]
Not really, no.

In general, slightly rectangular keycaps would be beneficial for all the keys on a column-staggered keyboard, for everyone with small-to-average-sized hands. But for a project aiming to be similar to the Ergodox, it’s nice to let people use standard 3/4" keycaps.

I agree with you that the pinkies don’t need 1.5u keys on the sides. I’m not sure whether it’s better to stay similar to the standard Ergodox, or better to try to make improvements to the pinky columns by altering the design a bit.

I think something like this design would work pretty well, but it might be better to just stick to the way the Ergodox does it.
« Last Edit: Sat, 20 September 2014, 17:18:13 by jacobolus »

Offline Findecanor

  • Posts: 5035
  • Location: Koriko
Re: fixing the ergodox thumb section
« Reply #67 on: Sat, 20 September 2014, 18:12:53 »
I’m happy to design a lister-style case, and I’m quite confident I can convince someone to help out with the PCB.
I don't feel ready to lead the PCB design, but I would be willing to proof-read Kicad and QCad files. (and suggest improvements, of course... )

Quote
I have no idea what the optimal design would be, however, at least to me, the major issue is not the thumb cluster, but the pinky finger.
You’ll notice that some of my mockups have the pinky columns shifted down a bit compared to the current design.
I think we should look at the Japanese M-system keyboards for how big the staggering should be. Who has one that we could measure?

Even better would be to have all the pinky keys raised up higher than the other rows
You could lift them a little bit using different keycaps, but that will give you only a very small difference.
I did that on my ErgoDox (see OP) where I got custom WASD keyboards caps in OEM profile made to match my BTC caps in DCS profile. Shifting one row would be even better.

A problem is though: are there smaller keycaps on the market for Cherry MX switches? And is it possible tou mount the switches closer together?
1: Not really. 2: You could cut away portion of the skirt on each keycap and mount the switches closer together, but then more of the switch would be visible.

Offline SonOfSonOfSpock

  • Posts: 321
  • Location: Colorado, USA
Re: fixing the ergodox thumb section
« Reply #68 on: Sat, 20 September 2014, 22:23:16 »
This thread talks a little bit about smaller key caps.
http://geekhack.org/index.php?topic=61786.0

I am glad you brought up the keys being closer for the pinky column. I like that idea and have thought that for flat boards, that would be more ergonomic. I think if this design were to bring the pinky keys closer together, we would have to switch to Alps or Matias switches which are slightly more compact than Cherry MX switches. Getting compatible Alps key caps for such a board would be much more difficult.

Offline PieterGen

  • Posts: 135
Re: fixing the ergodox thumb section
« Reply #69 on: Sun, 21 September 2014, 08:53:05 »
This thread talks a little bit about smaller key caps.
http://geekhack.org/index.php?topic=61786.0
Thanks.  Still, bringing the keys closer together is hard (without using different switches, like Alps etc).

suka found, in a way, a nice solution for "raising the key caps of the pinky row". If you look here you see a curved board. Compared to a flat tented board, the pinkies are raised (farther away from the surface of the earth, so to speak)

Suppose you take such a curved board, let's say for the left hand. You put it on the table. You put your thumb keys on the table as well, roughly near the Menu and Meta keys... then you have, I think, a fairly good height difference between the finger keys and the thumb keys.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #70 on: Sun, 21 September 2014, 14:03:43 »
suka found, in a way, a nice solution for "raising the key caps of the pinky row". If you look here
Show Image
you see a curved board. Compared to a flat tented board, the pinkies are raised (farther away from the surface of the earth, so to speak)
This is nice for making a one-off DIY keyboard that is very small and looks pretty, but it’s both ergonomically sub-optimal (if anything the pinky keys should be oriented so they angle very slightly the other way – while still being raised up – since that’s the direction the pinkies actually move) and also very difficult to mass produce.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #71 on: Sat, 04 October 2014, 01:30:12 »
I made a quick laser-cut acrylic prototype of the “mini” version. Ignore the specific legends shown on the keycaps here; I was just trying to get keycaps that were close to the right profile (some of them could be a bit better, but this is okay).




I added an extra .25 units of stagger to the pinky columns compared to the version shown upthread, based on my own hand shape. I found this to be a pretty good layout, but Yao, who has smaller and a bit differently shaped hands than I do, found that for her the main pinky column is staggered a bit more than she would prefer; I think the stagger shown earlier would be good for her. So I’m going to try making another version soon with some tweaks, including scooting the pinky column(s) a bit.

I’ll take some better pictures of this one in the daytime when there’s some light.
« Last Edit: Sat, 04 October 2014, 03:28:15 by jacobolus »

Offline kurplop

  • THE HERO WE DON'T DESERVE
  • Posts: 992
Re: fixing the ergodox thumb section
« Reply #72 on: Sat, 04 October 2014, 07:49:15 »

Show Image



I like this design.  It has an adequate number of keys for a multi layer keyboard.  It looks comfortable and simple.  The thumb keys appear to be well placed and everything is within reach.  By keeping the thumb cluster tight, it avoids the problem of getting too tall when tenting; This is especially important if using a pointing device between the halves.

I would recommend the less aggressive pinky column stagger if in doubt.

Why did you choose the staggered number row?  It seems to me that keeping them in their corresponding columns would keep everything tighter, be more consistent with conventional placement, avoid gaps between them (unless you use larger caps which would be difficult to find), and could allow room for an additional key per side in the same footprint, if desired.

I'm excited about this.  Overall, I think you've got yourself a winner there.  It's what the ErgoDox should have been!

Offline vvp

  • Posts: 886
Re: fixing the ergodox thumb section
« Reply #73 on: Sat, 04 October 2014, 07:52:18 »
Quite nice!
I would like to have the top row packet (no so sparse) but I like it anyway.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #74 on: Sat, 04 October 2014, 13:20:48 »
Why did you choose the staggered number row?  It seems to me that keeping them in their corresponding columns would keep everything tighter, be more consistent with conventional placement, avoid gaps between them (unless you use larger caps which would be difficult to find), and could allow room for an additional key per side in the same footprint, if desired.
This way the number row keys are extremely easy to find and impossible to make mistakes on. They pretty much follow the natural shape of the hand if you just open your fingers out; if they were all tightly packed, then most fingers would end up just slightly off of a key when you open your hand out.

The reason this is particularly important to me is I want to make a layout where I use the top row keys as modifiers (option, command, control, shift; with shift probably also duplicated to the side of the pinkies or on a thumb). This makes it very easy to use keyboard shortcuts, even ones with 2–4 modifiers, without the awkward hand contortions required by a typical keyboard.

The alternative way to get this kind of spread would be to angle the full columns of keys, but unfortunately unless smaller / differently shaped keycaps are used, this ends up taking up more space overall. It also would pull the design noticeably further away from the Ergodox design.

For folks who want the extra keys packed in on the top row, I think I’d recommend just using the bigger version, which is more suited to copying layouts from the Ergodox or a conventional keyboard:
« Last Edit: Sat, 04 October 2014, 13:22:45 by jacobolus »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #75 on: Sat, 04 October 2014, 16:04:07 »
Adding some more images:


Top row still takes a bit of hand reach; would be a bit better with a slightly taller keycap, but this is still much better than a standard keyboard or even an ergodox IMO:


With the index finger raised so you can see the other three fingers on the bottom row:
« Last Edit: Sat, 04 October 2014, 16:16:48 by jacobolus »

Offline sordna

  • Posts: 2248
Re: fixing the ergodox thumb section
« Reply #76 on: Sat, 04 October 2014, 16:17:29 »
This thread talks about fixing the thumb section, but all the proposed designs ruin the bottom row (some of us actually like the kinesis advantage style arrow keys).
Kinesis Contoured Advantage & Advantage2 LF with Cherry MX Red switches / Extra keys mod / O-ring dampening mod / Dvorak layout. ErgoDox with buzzer and LED mod.
Also: Kinesis Advantage Classic, Kinesis Advantage2, Data911 TG3, Fingerworks Touchstream LP, IBM SSK (Buckling spring), Goldtouch GTU-0077 keyboard

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #77 on: Sat, 04 October 2014, 16:24:03 »
This thread talks about fixing the thumb section, but all the proposed designs ruin the bottom row (some of us actually like the kinesis advantage style arrow keys).
Well personally I think the row two rows below the home row is hard enough to reach that I’m not going to use those keys for anything either way, so I couldn’t care less what happens to it (at least, for the keys not within reach of the thumb). However, of the people I’ve asked about it, the majority seem to appreciate arrow keys in an inverted T or diamond shape. I think this is just one of those things that folks are going to disagree about.

I wouldn’t be averse to a straight row across the bottom if more people prefer it that way.

By Kinesis Advantage style arrows, you mean up/down on one hand and right/left on the other hand? Personally I think that sounds terrible, since many of the times I want to use arrow keys come when I’m simultaneously using a pointing device with my other hand, but I’ve never tried it, so I can’t properly judge.

Do you have a preferred layout idea? Care sharing a picture?
« Last Edit: Sat, 04 October 2014, 16:33:14 by jacobolus »

Offline sordna

  • Posts: 2248
Re: fixing the ergodox thumb section
« Reply #78 on: Sat, 04 October 2014, 18:41:04 »
In my case I do use the left-right arrows on one hand and up-down on the other. Basically I use the Kinesis layout on my ergodox except I swap up-down to mimic the use of index/middle fingers on VI.
However other folks put all arrows on one side.
Anyway my main point is that this thread is about the thumb area, but has strayed into other areas of the keyboard that aren't really problematic :-)
Kinesis Contoured Advantage & Advantage2 LF with Cherry MX Red switches / Extra keys mod / O-ring dampening mod / Dvorak layout. ErgoDox with buzzer and LED mod.
Also: Kinesis Advantage Classic, Kinesis Advantage2, Data911 TG3, Fingerworks Touchstream LP, IBM SSK (Buckling spring), Goldtouch GTU-0077 keyboard

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #79 on: Sat, 04 October 2014, 22:54:05 »
Anyway my main point is that this thread is about the thumb area, but has strayed into other areas of the keyboard that aren't really problematic :-)
Well, personally I don’t like the bottom row, right-hand side, or amounts of column stagger on the Ergodox, either. It’s just that the thumb section is the most obviously flawed (IMO).

But someone who wants to build an ergodox using one of the thumb sections proposed here, or some spinoff – and otherwise leave the rest of the keyboard precisely like the existing ergodox – is welcome to.

Anyway though, did you have an opinion about any of the thumb key designs proposed here? Or other ideas for how you’d prefer the thumb keys to be arranged?
« Last Edit: Sat, 04 October 2014, 22:56:22 by jacobolus »

Offline sordna

  • Posts: 2248
Re: fixing the ergodox thumb section
« Reply #80 on: Sat, 04 October 2014, 23:10:58 »
My opinion on the thumb clusters is to simply bring them closer to the hands. I had posted these photos on the original ErgoDox thread as there are 2 possibilities for bringing them closer, one involves moving them inwards and down, and the other inwards but shrinking the innermost column bottom key to 1x so it fits.
« Last Edit: Sat, 04 October 2014, 23:19:00 by sordna »
Kinesis Contoured Advantage & Advantage2 LF with Cherry MX Red switches / Extra keys mod / O-ring dampening mod / Dvorak layout. ErgoDox with buzzer and LED mod.
Also: Kinesis Advantage Classic, Kinesis Advantage2, Data911 TG3, Fingerworks Touchstream LP, IBM SSK (Buckling spring), Goldtouch GTU-0077 keyboard

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #81 on: Sun, 05 October 2014, 02:40:10 »
My opinion on the thumb clusters is to simply bring them closer to the hands. I had posted these photos on the original ErgoDox thread as there are 2 possibilities for bringing them closer, one involves moving them inwards and down, and the other inwards but shrinking the innermost column bottom key to 1x so it fits.
Have you tried either of those out on your own hands, e.g. on a paper mockup? I find that the Ergodox thumb keys are already closer to the body than ideal (i.e. downward along the "y axis"; at least for someone with small to medium hands), but both of your alternative versions are moving them even closer.

So I take it you find the layout of the thumb keys to be effective. If you could design any keyboard shape possible, would you leave it with thumb keys in that current shape? Or you just think it’s okay, and you’re now familiar with it?

What do you use the keys for? I suspect I would still find 3–4 of the keys in your “fixed” layout basically unreachable for routine typing; do you you use those 1x1 keys frequently?

Offline daerid

  • Posts: 4276
  • Location: Denver, CO
    • Rossipedia
Re: fixing the ergodox thumb section
« Reply #82 on: Sun, 05 October 2014, 13:42:04 »
TBH, the ErgoDox would be absolutely fantastic for me if the thumb clusters were just a small distance closer. That way I wouldn't have to move my thumb laterally to hit the space bar, which is what was primarily causing my hand pain.

Also, I think at the very least I'd probably want a 2x space bar on the right side (since I only hit the space bar with my right thumb).

Acidfire's project still seems like it's going to be the jackpot in modularity and customizability.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #83 on: Sun, 05 October 2014, 15:30:21 »
TBH, the ErgoDox would be absolutely fantastic for me if the thumb clusters were just a small distance closer. [...] Also, I think at the very least I'd probably want a 2x space bar on the right side.
These two goals are impossible to simultaneously satisfy in a simple way. The corners of the primary thumb key and the corner index finger keys (N/B in QWERTY) coincide. If you start moving the thumb keys up and in, you have to make them smaller or they collide with those finger keys.

If you relax the 2x1 size constraint though, then this kind of thing would be possible:

Offline vvp

  • Posts: 886
Re: fixing the ergodox thumb section
« Reply #84 on: Sun, 05 October 2014, 17:31:05 »
If you relax the 2x1 size constraint though, then this kind of thing would be possible:
Show Image

Now make the inner column keys all 1x1 and that moves their bottom border a little bit up and then you can also move the top two thumb cluster keys more to the right ... and you will have it almost like on Katy  ;D

Well, for a flat keyboard, I think your acrylic prototype looks better, except that sparse top row  >:D

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #85 on: Mon, 06 October 2014, 00:51:42 »
Small, medium, large:

Offline nacitar

  • Posts: 27
Re: fixing the ergodox thumb section
« Reply #86 on: Mon, 06 October 2014, 11:49:17 »
TBH, the ErgoDox would be absolutely fantastic for me if the thumb clusters were just a small distance closer. [...] Also, I think at the very least I'd probably want a 2x space bar on the right side.
These two goals are impossible to simultaneously satisfy in a simple way. The corners of the primary thumb key and the corner index finger keys (N/B in QWERTY) coincide. If you start moving the thumb keys up and in, you have to make them smaller or they collide with those finger keys.

If you relax the 2x1 size constraint though, then this kind of thing would be possible:
Show Image


This one is the best so far imo.

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #87 on: Wed, 08 October 2014, 20:51:01 »
Tweak to the mini version prototype:


I think the pinky column is probably almost as good for me (the top row key is now a bit more of a reach; pinkies are short!), and better for some other people; overall probably an improvement. The tweaked thumb section tweak is noticeably better: the angle of further keys is closer to the angle the thumb actually sits at, and the other two bottom row keys are slightly less reach; the two taller 1u keys are about the same.

I now find these thumb keys extremely usable. For people with large hands the closer 1.5u key is probably going to be best as a spacebar. For people with small hands (or medium sized hands who angle their hands a bit more relative to the keyboard) the further 1.25u key (in this pic, the blue arrows) might be better. All 6 of the thumb keys should be pretty easy to reach for almost all hand shapes (though certain types of RSI might make some of them harder to use).

For the folks who don’t like the spread-apart top row: don’t knock it until you’ve tried it. I find these keys now very easy to target, noticeably easier than the top row on an Ergodox, because they follow the natural spread of the extended fingers. In general, I advocate assigning them to something other than numbers, and putting the numbers on a layer in either a numpad layout or

Comparison to Ergodox:


Note that because the pinky row is .5u closer to the body than on the Ergodox (“downward“ in the “y direction”), the hand ends up at a different angle to the keyboard, so the two big thumb keys here end up a bit closer in practice than on the Ergodox, even though from this comparison they look to be in about the same spot.

That is, in practice the comparison is a bit more like this:


* * *

I’ll try to prototype one or more larger keyboards sometime in the near future.
« Last Edit: Thu, 09 October 2014, 03:18:09 by jacobolus »

Offline vvp

  • Posts: 886
Re: fixing the ergodox thumb section
« Reply #88 on: Thu, 09 October 2014, 03:07:19 »
For the folks who don’t like the spread-apart top row: don’t knock it until you’ve tried it. I find these keys now very easy to target, noticeably easier than the top row on an Ergodox, because they follow the natural spread of the extended fingers.
Just my opinion about this.
Also that is true for index finger and little finger when in most relaxed straightened position (if we can call straight finger relaxed at all). Index finger and little finger are quite good at small radial movement when almost straight. The more bent they are the more packed the natural position is. So the sparse top row leads you to two less keys in the top row for hardly any improvement in reachability. You are replacing two keys with one key at the position just between them. And you are doing it for both index finger and little finger.
May be a good trade for people who like layering a lot. I'm more in the group of people who like a lot of keys (if they are still reachable without moving the hand).

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #89 on: Thu, 09 October 2014, 03:18:57 »
The people who like a lot of keys can choose from many other keyboards. :-)

Anyway: give it a shot. You might find that you like the gaps. It makes it almost impossible to make errors on these keys, something which definitely cannot be said for the number row on a conventional keyboard.

Also: the potential two extra keys that could be added, one on each end (if we crammed the rest of the top row together), would be quite noticeably harder to reach than any of the keys that are currently on this keyboard. Personally I think the extra index finger key would be especially uncomfortable to reach, but the top row pinky key there already takes some hand movement, and a key scooted half a unit further away would also be that much harder to get to.
« Last Edit: Thu, 09 October 2014, 03:24:22 by jacobolus »

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #90 on: Thu, 09 October 2014, 03:50:50 »
Comparison with split TKL board:


On this keyboard every key is easily reachable, and pretty much every easily reachable spot has a key on it. (At least, to the extent possible on a flat keyboard.)

Really highlights how many of the keys on a standard keyboard are hard to reach (in particular escape, delete, return, right shift, and all the ctrl/alt/meta keys, all very important in modern operating systems), and how much some valuable possible motions are wasted (especially for the thumbs).

The comparison doesn’t even quite do it justice, because on a standard keyboard the hand is usually angled over (because there’s no tenting), making the index fingers and pinkies both less able to reach than if they hands were comfortably straight above the board.
« Last Edit: Thu, 09 October 2014, 03:57:20 by jacobolus »

Offline kurplop

  • THE HERO WE DON'T DESERVE
  • Posts: 992
Re: fixing the ergodox thumb section
« Reply #91 on: Thu, 09 October 2014, 08:29:50 »
I can see how the spread top row keys could be more natural to reach and if you're convinced that it is you shouldn't compromise what you think is ideal spacing.

My concerns are more visual. As it is, it reminds me of a mouth badly in need of orthodontic work, and with the staggered columns a simple 1.25 cap fix would still leave small square cavities in it. Also, while the extra top keys may be more difficult to reach and might be unnecessary with layers, a case could be posited that it could be desirable to have a few keys that are less convenient to access; layer locks for example.


Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #92 on: Thu, 09 October 2014, 17:19:07 »
My concerns are more visual. As it is, it reminds me of a mouth badly in need of orthodontic work, and with the staggered columns a simple 1.25 cap fix would still leave small square cavities in it.
Yep, it would be possible to put 1.25 caps in there (and for that reason it might even be worth scooting the other index finger cap over by an extra .25u to make that uniformly possible, or scooting both index finger keys closer by .125u so that if the other three get 1.25u keycaps there will be no gaps), and I think that sounds like a generally reasonable idea.

This kind of thing:


The reason I have them shown as 1x1 keys is that I want to use SP’s DCS keycaps on this thing (since those are fairly easily available), but not quite the standard way: I want to use DCS row 5 on the top row, so the profile of each column looks like:

(note that I’m skipping the traditional “home row” [row 3] keycap shape, so I can get some increased height to the further-away rows; this makes them easier to reach, and easier to fully press with an extended finger without touching the neighboring keys)

But DCS row 5 only comes with 1x1 size keycaps.
« Last Edit: Thu, 09 October 2014, 17:36:56 by jacobolus »

Offline kurplop

  • THE HERO WE DON'T DESERVE
  • Posts: 992
Re: fixing the ergodox thumb section
« Reply #93 on: Thu, 09 October 2014, 19:00:12 »
If you are sold on the upper row positioning, I think the ergo factor should trump the look and given the DCS availability, you should stick to your earlier plan.

Will all of the components fit within the footprint shown?

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #94 on: Thu, 09 October 2014, 20:40:12 »
Will all of the components fit within the footprint shown?
What do you mean? Like, can a PCB be made within that footprint? Sure. An ARM chip that will drive a keyboard is a square less than half an inch on a side (like .35–.4"?), so it will fit pretty much anywhere.

* * *

Also, the outline here is just what I was using for my laser-cut acrylic mockup. It might be better to make a case for this in a different shape. For instance, it’s possible to make a very simple keyboard case by just using one piece of bent metal, which serves as both the plate and the “case”, with nothing on the other side of the PCB; such construction is relatively cheap because the whole keyboard is just 2 pieces of cut and then bent sheet metal, two PCBs, switches, keycaps, a cable between the halves, and a handful of other components, mostly surface-mount. Or a case could be made from layered Acrylic, or two pieces of metal with standoffs in between, or a big block of CNCd plastic/wood/metal, or.... Different construction methods call for somewhat different shapes, I expect.

For the simple bent metal case, for example, it might be necessary to have some straight sides, for the bend. So maybe it would end up along the lines of:

Or maybe something totally angular. Or maybe like something else entirely, I’m not sure.
« Last Edit: Thu, 09 October 2014, 20:52:54 by jacobolus »

Offline daerid

  • Posts: 4276
  • Location: Denver, CO
    • Rossipedia
Re: fixing the ergodox thumb section
« Reply #95 on: Sat, 11 October 2014, 23:48:26 »
I'm sorry, I forgot to clarify (been on my honeymoon).

When I said 2x1 spacebar, I meant horizontal. Having some lateral distance available for the space bar is probably the most important

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #96 on: Sun, 12 October 2014, 01:12:01 »
I'm sorry, I forgot to clarify (been on my honeymoon).

When I said 2x1 spacebar, I meant horizontal. Having some lateral distance available for the space bar is probably the most important
Aha. You’d probably like Matias’s new 2.5x1.5 spacebar keycaps.

I wonder if there’s a good way to support both skinny and wide spacebars; personally I don’t find the skinny ones to be a problem, provided they’re in the right spot.

Also, congrats! :)

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #97 on: Sun, 19 October 2014, 00:41:42 »
One possible embedded numpad layout, which stays fairly close to the standard numpad.

Offline tp4tissue

  • * Destiny Supporter
  • Posts: 13551
  • Location: Official Geekhack Public Defender..
  • OmniExpert of: Rice, Top-Ramen, Ergodox, n Females
Re: fixing the ergodox thumb section
« Reply #98 on: Thu, 23 October 2014, 06:53:49 »
One possible embedded numpad layout, which stays fairly close to the standard numpad.
Show Image


You guys are just asssuming that those angles for the thumb key will fit.. 

It'd be better if that piece is horizontal. 

Offline jacobolus

  • Thread Starter
  • Posts: 3661
  • Location: San Francisco, CA
Re: fixing the ergodox thumb section
« Reply #99 on: Thu, 23 October 2014, 16:17:54 »
One possible embedded numpad layout, which stays fairly close to the standard numpad.
Show Image

You guys are just asssuming that those angles for the thumb key will fit.. 
It'd be better if that piece is horizontal.
What do you mean assuming that the angles will fit? I built a physical copy of this and wired it up. I like it much better than the Ergodox, at least for my own size/shape hands (fairly medium sized).