Super ultra high speed board

[chimera15]

So the ultimate typing championship got me thinking that it might be possible to design a keyboard that had everything going for it for super ultra high speed input.  

Overall the keyboard probably won't make a huge difference, but at the highest speeds, it could make the difference between 200 and 205 wpm.  Even a couple extra .'s could matter in a competition.

Right now I'm just brainstorming, but a couple things I've noticed trying all my keyboards.

Keys should have a small area, and have good separation.

I noticed my steampunk hack with old typewriter keys was my fastest board, most likely because of the rounded keys.  

Keys should have some sort of dampening.

Keys shouldn't have a lot of friction, possibly linear switches would be preferred over clicky at ultra high speed because of the speed you're moving the key is going to bottom out anyway, and clicks and tactiles bumps don't make a whole lot of difference.

Keyboard should have n-key rollover.  This could be useful, although I'm not 100% because keypresses may be so close they could be actually right on top of each other.

[chimera15]

So far right now.  I'm thinking something like green alps switches with cream plungers, and bottom switch cases, wired onto a gaming keyboard, with keycaps that have been cut into circular shapes instead of square, maybe with some sort of rubberized, or soft silicone keytop.

Actually the rubberized/soft silicone keytop might take care of the bounce and prevent harshness to fingers, so you wouldn't need cream plungers...

[erricrice]

Something I have noticed recently about bottoming out:  When using my Cherry Black 'board, I would bottom out constantly just to make sure that I was actuating the switch so the stroke would go through.  I noticed that as soon as I stopped bottoming out and only went the actuation point once I had gotten used to it that my typing speed went up more than 20 CPM.

I think this is because the actual time that you spend pushing down the key(increased as you push the key down further) and also the time it takes your finger to come back up and get to the next key(increased if you push down further).  So I have found that not bottoming out has actually increased my typing speed with other 'boards as well once I started making an effort not to bottom out.

So no real empirical data, but it makes sense if you look at it from that point of view.  

I agree thought about the linear switches, once you get to know the actuation point, you really don't worry about whether or not you have pressed it, and with some switches like white alps copies, there can be a lot of friction to the point where you think you have put enough force to actuate the key but it just doesn't because of the angle you're pressing at or just a random increase in friction for some reason.

[chimera15]

Yeah. I agree, around 100 wpm, it does help your speed to try to lighten up your touch a bit and listen or feel for the bump or click, but from what I've seen and felt even at those speeds it's difficult to control how hard you hit each switch. It looks and more importantly feels  like to get past 150wpm, your hand and fingers will be moving so fast that you won't be able to keep from bottoming out, which is why some sort of padding in my theory is  an absolute necessity.  The only thing that I worry about when adding padding is that it doesn't make the key sluggish, so it's going to have to be a very thin strip of silicone if I use that method.

To type at 200wpm, your hands will be moving like a machine gun, it's going to be very difficult to control how much pressure you're hitting a key with.

I'd really love to even talk to or see someone type at 200wpm.  To see the board and techniques they use.  It should be really interesting.

Right now I want silicone keys because the tips of my fingers are becoming sore, even if it does slow me down. lol

[chimera15]

It works! It's soft.  It's a thinly cut wafter of silicone glued onto the keycap.   roflol



It looks like hell, but feels great. lol

[chimera15]

It's going to be the ultimate.  A keyboard that feels like boobies!!!! roflol

[hyperlinked]

I don't know if this is a problem for anyone else, but I sometimes feel like I'm having difficulty retracting my fingers in time to take the next stroke when I'm typing very fast for more than 30 seconds. I almost want something that will allow my fingers to bounce off of it, but needless to say the drawback to that is that I wouldn't be able to type very long on it before sheer fatigue makes the part about registering keystrokes hard.

[chimera15]

Macros. Why type " the " (5 keys pressed) when you can type RightAlt+t (2 keys) ? Etc.

Why not just macro the whole passage if it's a competition...  That's not the point....   Besides I think it actually takes more time to press and hold a switch in combination with another than just tap them one after the other.  There's an abbreviation in one of the passages in the competition "U.S.A." for instance, and the switching between Shift a then a comma which lets up, takes a lot of time off the overall passage.  You have to slow down and think about what you're pressing.

Moving your hand off to wherever you might have macro'd "the" or whatever word, since it wouldn't be in the main section of the keyboard, would take milliseconds off your time too.    It's probably not worth it too since "the" maybe only occurs once or twice in the average passage used.

[chimera15]

New spacebar on my hhk2 hack:

[chimera15]

I don't know if this is a problem for anyone else, but I sometimes feel like I'm having difficulty retracting my fingers in time to take the next stroke when I'm typing very fast for more than 30 seconds. I almost want something that will allow my fingers to bounce off of it, but needless to say the drawback to that is that I wouldn't be able to type very long on it before sheer fatigue makes the part about registering keystrokes hard.

What keyboard did you use to break 120?

[chimera15]

Not sure about that, I use macros for the 5 most popular words in the English language on my DataHand and in speed tests I get a bump of 5 to 10wpm. If you have to slow down to think what you are pressing, you are untrained? If with an ordinary keyboard holding down the modifier is the problem, you can program it into a toggle-once key (i.e. press RighAlt, release, press t).

How fast do you type with your datahand out of curiousity?
What's the top speed anyone has reached with one?

[chimera15]

My new soft keycaps.  No more banging fingers into hard plastic:  



I'm going to use greens for the main keys, but clicky blues for the spacebar, shift, return, and backspace.



I'll probably try these caps on my siig complicated that I've been using during most of the competition to see if I'm on the right track or not.  I may have made the surface area too small, not sure.

Right now this board is uber soft.  I used 10 shore a hardness silicone so it's not super soft, and still somewhat stiff, but still way softer than most mold making silicone.

[hyperlinked]

What keyboard did you use to break 120

I used a Filco-104 w/Cherry Brown. I find that I'm generally only good for around ten short tests in a row. After about ten, I start making a lot of errors because I can't lift my fingers back up fast enough. I don't remember if this was a problem before I got this board, but I also didn't try doing continuous typing tests on other boards either.

I could try switching keyboards, but that wouldn't really be reliable either. I'm so used to this board that I'll make other kinds of errors excessively on another board.

I did manage 115wpm on Keyhero using an Asus EeePC board the other day though. I was kinda surprised at the wpm when I was done. It felt like a slow board, but perhaps all that short travel added up in time savings.

[chimera15]

Transferred.  I'm typing on them right now. My initial thoughts are that they're not too small.  It's definitely a different feel, but working well.  Still to see if they're as fast or faster than normal caps.  At very least my fingers aren't hurting like they were when I was jamming them into the plastic. lol

Edit:  Just from typing that I can tell I'm about the same speed as before, if not better.

[chimera15]

I used a Filco-104 w/Cherry Brown. I find that I'm generally only good for around ten short tests in a row. After about ten, I start making a lot of errors because I can't lift my fingers back up fast enough. I don't remember if this was a problem before I got this board, but I also didn't try doing continuous typing tests on other boards either.

I could try switching keyboards, but that wouldn't really be reliable either. I'm so used to this board that I'll make other kinds of errors excessively on another board.

I did manage 115wpm on Keyhero using an Asus EeePC board the other day though. I was kinda surprised at the wpm when I was done. It felt like a slow board, but perhaps all that short travel added up in time savings.

Ah that's another idea I had, you're right.  I tried my low profile alps and I thought that maybe they did have a bit better time because they were a shorter travel.  The key layout is sort of odd with the m1242, though so, I think that slowed me down, and I didn't get as good results as I might have if they were in a more standard layout.   It's possible that a scissor switch board might actually be a much better choice for speed in general.  

It's really interesting to hear you used cherry browns..  I'm really curious to test the scorpius I have coming.

[chimera15]

It's pretty clear I'm about 10-20 wpm faster with these caps.  It could also be that I got a good rest, since my last tests, but these caps feel really good.  I'm for sure not slower.

[msiegel]

this is a very interesting experiment

[chimera15]

I posted a 108.6 on one of the passages I've been working on memorizing which is one of my fastest times yet, using them with my clicky complicated siig. So I know these caps are definitely not worse than normal ones.

[ak_nala]

Your use of silicone to buffer the bottoming out brings to mind another, if more complicated, solution already implemented many, many years ago - in a typewriter.

My Olympia SG-1s (yes, I have two now) have little individual springs under all the keycaps to help buffer the fingers when bottoming out (which you pretty well have to do with a manual typewriter to get a good key-strike). That coupled with the already silky smooth and light action (for a manual) leads to a delightful typing experience with no loss of precision that might ensue from having an overly soft surface on the top of the keycap itself.

Wonder what the effect would be on a computer keyboard. It's even possible the extra spring action might give a bit of a bounce and help the finger back up as an added benefit. I would think the keypad spring would have to be at least double the switch actuation force.

I definitely think the round keycaps are a great idea. Fewer edges to catch, as you've found. Hemispherical instead of the modern cylindrical indentations might also help, in that it would facilitate automatic centering of the fingers on the keys in all directions rather than just side-to-side. Positional feedback can be invaluable at high speed.

The older typewriter round keycaps run between 1/2" and 5/8" (but usually topping out around 9/16") in diameter, BTW. Even the newer double-shot plastic ones when more finger shaped, or even the more or less square ones (Remingtons of the 50s, for instance), were a bit smaller than modern key tops and had heavily rounded edges. All had a wonderful feel most modern board keycaps just can't match. The closest I've felt are the caps on the Displaywriter keyboard I just got :-)

[chimera15]

Your use of silicone to buffer the bottoming out brings to mind another, if more complicated, solution already implemented many, many years ago - in a typewriter.

My Olympia SG-1s (yes, I have two now) have little individual springs under all the keycaps to help buffer the fingers when bottoming out (which you pretty well have to do with a manual typewriter to get a good key-strike). That coupled with the already silky smooth and light action (for a manual) leads to a delightful typing experience with no loss of precision that might ensue from having an overly soft surface on the top of the keycap itself.

Wonder what the effect would be on a computer keyboard. It's even possible the extra spring action might give a bit of a bounce and help the finger back up as an added benefit. I would think the keypad spring would have to be at least double the switch actuation force.

I definitely think the round keycaps are a great idea. Fewer edges to catch, as you've found. Hemispherical instead of the modern cylindrical indentations might also help, in that it would facilitate automatic centering of the fingers on the keys in all directions rather than just side-to-side. Positional feedback can be invaluable at high speed.

The older typewriter round keycaps run between 1/2" and 5/8" (but usually topping out around 9/16") in diameter, BTW. Even the newer double-shot plastic ones when more finger shaped, or even the more or less square ones (Remingtons of the 50s, for instance), were a bit smaller than modern key tops and had heavily rounded edges. All had a wonderful feel most modern board keycaps just can't match. The closest I've felt are the caps on the Displaywriter keyboard I just got :-)

Do you have any pictures of that spring keytop?  I'll have to look for those.  Maybe it would be possible to cut or take them off and use them with an alps switch somehow?

Yes the results I was able to get on my steampunk hack,  which uses old typewriter keys was what made me discover that the rounded keys help speed a lot.

Hmm

[ak_nala]

That's the one.

I'll have to drag out my camera for shots, and figure out how to up-end the beast without damaging the one achilles heal on the thing, the plastic paper guide (rare to have them intact) - maybe tomorrow.

However, CUT THEM OFF!?! Don't take this the wrong way, but are you nuts? That would be like chopping a 1930s Mercedes Roadster so you could put the fenders on a Honda Civic.

I suppose if you found one that was pretty well thrashed to begin with, but honestly, these were arguably the best full-sized manual typewriters ever made, and they aren't making any more of them!

That said, you can get SG-1's for under a $100 shipped on the Bay in as-found condition if you look hard enough. Not sure if the SG-3 has the same type of suspension on the caps, but they are up there too.

Though, I really think the concept is more valuable than trying to graft on from an original, IMO. I think the needs of the original typewriter and that of a modern keyboard are a little different - we would probably need something a bit springier and perhaps longer travel than the 0.5-1mm keycap spring travel of the Olympia. Just have to engineer and cast new caps after finding the right, readily obtainable, spring.

[chimera15]

Had an idea how to make the soft keys more uniform and less likely to de-laminate with the glue I'm using, so they have a holder, not just resting on the keytop with glue.   Right now, after probably several thousand keypresses they get loose and have to be reglued. This would also make them even softer because there's more silicone in the holder to hit, not just a thin wafer.  I need to buy more antique keys though if I use this method, although it wouldn't matter what condition they were in, they could have broken glass where with my steampunk mod they needed to be perfect.

[chimera15]

Got the perfect set off ebay:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=370298611257&ssPageName=STRK:MEWNX:IT

It even has the square ones I like.  Should be awesome.

[chimera15]

Selectric thread here:

Show Image

Ron Mingo could type super fast using just a baseball bat.  Here he is with the SF Angels.

Show Image

Not quite sure what you're trying to say, but about the fastest typist in the world using a selectric...   I wonder if it is because the keys were powered that it allowed that speed, you only have to lightly hit them, where even the lightest key in a keyboard still takes some pressure not to mention time to actuate.  It may be that a standard layout keyboard with no keypress or actual switch would be the fastest.


I'm coming to the theory that actual relative keyboard speed, keyboard to keyboard is about keys design like the round ones that allow for accuracy, and the actual pressure it takes to activate each key.  If the pressure it takes to activate the key is a definite part of the formula then I'm not sure green alps will be the right choice, there may be better choices for a switch,  like nmb's, montereys, or even acers.

You mentioned how many grams it took to activate a Selectric key before right?

It may be worth time to break this down to a more analytical way, not simply by feel, and time the exact amount of milliseconds it takes to put one or a measured amount of say one word on screen.

I wonder if there's some kind of utility that could time me rather than having to do it by hand...?

A utility like that would be good to use to test layouts too.  For instance how long it takes in qwerty to type the word "the" vrs dvorak.

[chimera15]

60grams, so about the same as a white alps switch for a Selectric? I guess that equates to the same amount of travel time for your finger?   There is a difference though, there's a constant up pressure, where I seem to recall with the Selectric the switch drops away from your finger once it's actuated?  

If I'm remembering correctly, that could be a big difference as far as your finger knowing there will be a letter generated, so it's able to move on to the next key faster.  I find if I lighten my touch up I do get faster, but it generates more mistakes because I sometimes will miss a character.  With the switch dropping away from your finger, there's little doubt that you got the switch activated.

It's interesting I'm getting about 115 average on this site:

http://labs.jphantom.com/wpm/

So the typing test at the competition site are really difficult.

Still looking for a site that will give me a time output for a short word though.

[msiegel]

With the switch dropping away from your finger, there's little doubt that you got the switch activated.

that's exactly what harris's ibm buckling spring patents say

[chimera15]

that's exactly what harris's ibm buckling spring patents say

Ah that's interesting.  I hadn't thought of it that way, that clickies sort of perform that function.  But it's not the same that the pressure decreases, from the key actually physically accelerating away from your finger.


I really want to see a typing test that can read cpm, on only a few words, and that I'll be able to equate to wpm, which will let me see how fast you physically have to go to reach 200.  If I can see with one word the level, maybe I can approximate it with the rest of my typing, and see if it's actually physically possible with the keyboards I have.  Right now it's looking to me like 200wpm exceeds the capabilities of the alps switch.

[chimera15]

This is an interesting board:

http://www.frihost.com/forums/vt-89723.html

One guy states that 200wpm = 3.3 words per second...so damn, that's fast...  they better be small words. lol

I wonder if I'm physically able to type something like, " The boy went down to the river."  in under 2 seconds. lol  I need to try that.

I need to find a clock that will time that and automatically stop at .

Maybe I can write one in flash if i can't find it...

[hyperlinked]

I wonder if I'm physically able to type something like, " The boy went down to the river."  in under 2 seconds. lol  I need to try that.

I need to find a clock that will time that and automatically stop at .

Maybe I can write one in flash if i can't find it...

If you can't find it, it'd be really easy to write it in JavaScript. Just start the clock with your first keystroke and get the script to output your input character count once the timer hits 2 point 00.

...and I'm sure you were just joking about the short words, but in case you didn't know, the wpm count is usually extrapolated from characters per minute against an average number of characters per word. I forget the number. It's something like 5 characters per word.

[chimera15]

http://speedtest.10-fast-fingers.com/

This may give me the raw data I need, it's a much more random test that doesn't throw any punctuation or strange words into the mix.  I should give me an indication of what my maximum speed on each board is and what the possibilities are.  It also gives cpm raw data.  


Although this test is constant it doesn't allow memorization, so times will be relatively slower.  It also throws a flip page into it that slows me down a bit.


Acer:

Black/used:  80-85wpm

I was surprised with my speed with this board. The feeling of these switches is very much more plasticy than alps switches.  They're a lot lot lighter as well however, especially this well used one.

white/fresh/ hhk1 hack: 75-80wpm


This board was uber clicky, which feels good, but I had a lot of problems with speed with it.  This was due probably because of the strange position of the backspace, which is 1 key down from normal backspaces, which would take some time for me to get used to again.  



Ibm model f: 75 wpm

I have a feeling that I got such bad scores with this board because it has the small back arrow I'm not used to, and also the space bar on my unit seems very hard to press relative to the keys, which I think is a result of me having it out of its slot for such a a long time.  I'm also no longer used to the space between the keys which is clearly different than an alps or alps like switch.


Japanese blue alps: 85wpm

 It looks like about 85wpm is my consistent average speed with it with my blue alps Japanese keyboard with about 2 words mistyped.


Cream alps, modified sig: 85-90wpm



  Another slightly above average alps switch for speed I would say.  It's nice not to bottom out so hard, and it's probably about the same speed as complicated white alps.



complicated white siig/silicone keycaps: 85wpm

 I'm about the same with my silicone keycaps on my complicated white alps siig, perhaps a tiny bit slower.  My accuracy seems a significant amount better, I missed maybe 1 word on average.

complicated white normal keycaps/big spacebar: 85-90wpm

I got up to about 85-90 with my bluetooth hack with normal keycaps which uses white complicates, whites, the same as my siig with silicone keys.  One thing I noticed here is that the extra large space-bar on it I think does help considerably.

fake white alps-old typewriter keycaps- 85wpm/good accuracy

  This board was great for accuracy, which can translate to speed overall, I almost never missed a word, but for speed I can't say it's significantly better than the other alps switches.  It's not going to get me to 200wpm.

irocks kr6820e: 85-95wpm

pretty consistent 83 wpm with my irocks kr6820e.  Lots of errors though. It has really comfortable keys at that speed though, very bouncy, short travel relatively.  

Edit:  Just got a 93 with it...  Not sure I can go farther than that with it.  Maybe a bit.  I might get up to 100wpm..  It's a good kb for speed as I thought it might be.  It's domes are very much like scissor switches.

Monterey- 80 wpm

It could just be that I'm not used to this board that I'm slower. It also hurt using it after a few tests.  I bottomed out every keytouch because the click and spring in these switches are so light.

nmb/white - 85/ super accurate

 Another very light touch clicky board, this one has a more pronounced click than the monterey. They have smaller peaks than most alps which likely accounts for the improved accuracy.  My feeling is that I could probably go a bit faster with this board, maybe more in the 90-95 zone, but want to move on to others.  This is a very fast board for a mechanical so far.

Fake alps standard boards:

smk-85 - 85-90

  This is a fast board. The switch is very clicky unlike the complicated.  You know when you hit a key, very low error rate.  


Siig kb1948  85-90

This board, which I tested right after the smk has a completely different feel to it, although it uses the same type of switch as the smk,  probably because of the different quality of the keycaps.  Still, about the same results, most around 88-89 zone.


----------------------------------------------------------------------------------------------------------------------------------------------------------------------



I would say the irocks board is the big surprise so far.   I would guess it's about 10 wpm faster than the average boards.  Probably because of the short depression and nice rubber bounce at the bottoming out.  The only problem I had with it is that I would miss a character every so often.  That error rate would probably bring it down to average board speeds even though the switch activation was signifigantly better.

There's no jump out board that's like clearly 20 wpm better that would allow me to jump up to 200 wpm.

I'm going to try to learn dvorak using that site and bring my speed up and see if I can get clearly better speeds this way.

I think this makes clear to me that it is possible to design a board that is faster overall, but it may not be the direction I'm going.  I think silicone is a good start, and spongeness, because what's happening at 100 wpm plus is really like little collisions that need to be dampened in some way or your fingers will pay the price, not the hard plastic.

[chimera15]

If you can't find it, it'd be really easy to write it in JavaScript. Just start the clock with your first keystroke and get the script to output your input character count once the timer hits 2 point 00.

...and I'm sure you were just joking about the short words, but in case you didn't know, the wpm count is usually extrapolated from characters per minute against an average number of characters per word. I forget the number. It's something like 5 characters per word.

Well my point was with that short phrase was to see if I could get it up to a speed that would equal 200wpm to see if it's actually physically possible with a qwerty keyboard and to experience what that speed is like, even with short words let alone long ones, and how much I have to practice to actually rattle that short sentence off at 200wpm

[DreymaR]

Being such a fast QWERTY typist, I'm sure that trying to learn Colemak or Dvorak to get faster now will bring you a lot of frustration. Just a friendly warning.

It certainly seems to have worked for Barbara Blackburn, but I'm sure she put an extraordinary effort into her typing practice - she pretty much typed all day for years I think.

[chimera15]

Being such a fast QWERTY typist, I'm sure that trying to learn Colemak or Dvorak to get faster now will bring you a lot of frustration. Just a friendly warning.

It certainly seems to have worked for Barbara Blackburn, but I'm sure she put an extraordinary effort into her typing practice - she pretty much typed all day for years I think.

I pretty much type all day, doing one thing or another.  If it does boost a significant amount I'm sure it'll help, especially with practiced memorized passages.   The key thing I need to know, is does dvorak's layout really improve speed, and by how much on average.  If it's 10wpm even that's still pretty significant, if it's more than that, if it's what's making 200wpm possible vrs 150 or something, then it's a must.

[chimera15]

I was thinking.  It might be possible to put a layer of silicone around the bottom of the green alps switch, that would at once limit the travel of the plunger, and second give a nice bounceback all at once...hmmm...I'm going to have to check that out.

Edit:

I opened a green alps switch and discovered it has the little projections in the bottom of the switch that hit cream alps bumpers, so that might be a valid way to go.  The activation leaf however goes all the way to the bottom of the case, so I don't think I can add silicone to it at least not in a soft form.  It's possible I could let it take up the back half of the switch so it doesn't interfere with the leaf....

Testing a green alps switch with a continuity meter I get continuity with the plunger about half way down.  So I think one of the keys will be limiting the plunger to that point, and so it hits softly and returns.  That's only about a 2-3 mm travel, which should be less than even the irocks board, that should equal speed since it should be a consistant mechanical hit that won't have misses unlike the rubber dome irocks board.

[chimera15]

I did it!! I managed to afix a piece of silicone between the case and the plunger on one side.  It now only goes down about 3 mm and then hits softness, activates the switch and returns!!!!!  This should be awesome! lol.

If this board doesn't give me 10-20 wpm faster, it'll sure be soft. lol Someone could type on this thing with broken fingers and not even feel it. roflol

[ak_nala]

Do you have any pictures of that spring keytop?  I'll have to look for those.  Maybe it would be possible to cut or take them off and use them with an alps switch somehow?

Yes the results I was able to get on my steampunk hack,  which uses old typewriter keys was what made me discover that the rounded keys help speed a lot.

Hmm

Show Image

Here's the underside of those keys (sorry for the dust bunnies, but at least they're not board chow ;)


Note that even the Shift Keys are sprung.

The keys on your pic were black and mine are green is the only difference as yours is a definitely a pic of an SG-1.

The spring is housed inside the double-shot plastic key caps, with a steel key shaft coming down through the bottom of a steel cup that is attached to the key bar and cradles the bottom of the spring, so the spring is trapped between the cup and the key cap. There is a slot in that key shaft that allows the key cap to go up and down (and limits the travel). It also appears to be the attachment method for the whole assembly.

I haven't measured it, but I'd say the limit of travel is around 1-2mm with a medium stiff spring action (significantly heavier than an Alps).

BTW - that's all stainless steel - shame they didn't make all the screws in the housing SS as well :(

[ak_nala]

I've checked the one Olympia portable that is at hand (the others are a bit buried at the moment). It is an SM-9 from the late-60's, and it doesn't have any springs in the key caps at all (straight connection to the key bars). There are some white Olympia key caps on eBay right now, but they also appear to be from an SM-9, so won't have the springs.

I have a couple SM-3's that date from the 50's to early 60's (same time period as the SG-1's - the one I took pics of dating from around 1960) that are supposed to have some kind of spring action in them (though reportedly for leveling the keys during the key-press rather than cushioning them). I'll try to dig them up to take a look and report back.

[chimera15]

I figured out rather than trying to redo another matrix on these in case these keys actually suck I'm just going to wait for the siig minitouch I have coming from Canada that I'm going to make a wireless mini from.  I'll have a spare siig minitouch pcb I'll be able to wire these greens right into then.  I'll be able to make 2 compacts at that point.  One wireless with white simplified switches and the other this speed keyboard with a siig minitouch pcb.  That way I don't have to wire any more matrix's, cause that's a pain the but. lol

[chimera15]

Here's the underside of those keys (sorry for the dust bunnies, but at least they're not board chow

Show Image

Note that even the Shift Keys are sprung.

The keys on your pic were black and mine are green is the only difference as yours is a definitely a pic of an SG-1.

The spring is housed inside the double-shot plastic key caps, with a steel key shaft coming down through the bottom of a steel cup that is attached to the key bar and cradles the bottom of the spring, so the spring is trapped between the cup and the key cap. There is a slot in that key shaft that allows the key cap to go up and down (and limits the travel). It also appears to be the attachment method for the whole assembly.

I haven't measured it, but I'd say the limit of travel is around 1-2mm with a medium stiff spring action (significantly heavier than an Alps).

BTW - that's all stainless steel - shame they didn't make all the screws in the housing SS as well :(

Wow that's really interesting, so what happens? What triggers the spring? The little locks? How does it get trapped?  I may have to really watch for one of these on ebay.  I did look them up and couldn't find any.

[ak_nala]

Here's one that's mis-labeled as a "portable":

http://cgi.ebay.com/OLYMPIA-DELUXE-PORTABLE-MANUAL-TYPEWRITER_W0QQitemZ220519839982QQcmdZViewItemQQptZBI_Typewriters_Word_Processors?hash=item335801b8ee#ht_500wt_1182

There were a couple more last week, but they are gone.

Note that shipping on these is pretty bad, because they are HEAVY. Gotta laugh at the above when they call it portable. Hernia inducing is more like it.

[ak_nala]

What triggers the spring action is actually bottoming out. You thwack the key down to give a good, consistent registering of type on page, but instead of a sudden halt as on most typewriters, the key cap acts as a cushion.

The spring is trapped because the end of the little shaft that comes through the cup that holds the bottom of the spring has a slot in it that hooks over the end of the key bar and settles into a notch that prevents it from coming off. In that notch it is free to travel the length of its slot before the key bar hits either end. I think there is a bit more play in it at the bottom than you'd ever use in actual practice that allows just enough wiggle room to lever the shaft off of the bar for disassembly, but not enough that it would ever happen in use. Haven't tried to take it apart.

[ak_nala]

BTW - The art of typing with a manual typewriter (so it doesn't kill you, that is), is to thwack down on the key with just enough force to make the type hit the platten and then back up. It's much more of a quick down-up action than you usually do on a computer keyboard - more like staccato on a piano - allowing momentum to do a lot of the work rather than just hitting the keys full force. It's much more in the wrist than when 'boarding.

And the REAL art is to do all this at speed and with nice, consistent force so all the characters are crisp and the same shade of blue-black.

Nice thing about the Olympia is it takes out any trace of the bottoming out jarring that remains even with good technique, and helps even more when you aren't hitting the keys perfectly.

[chimera15]

Interesting.  I changed my silicone caps to my mini hhk blue alps hack and I started getting times in the 90's, from 85's before. It's possible these caps really do make a significant difference.

[DreymaR]

About the layouts:

I don't think anyone have solid scientific evidence one way or another. But the anecdotal evidence suggests that Colemak and Dvorak will be faster once you learn them well enough. The big question is: When is that?

Dvorak promotes hand alternation, which is good for speed. Colemak has less hand alternation than Dvorak but more rolls which is also good for speed. From playing tambourine I have the feeling that rolls are the fastest way to travel - I did an astonishing 240 WPM on a simple roll test once, corroborating this theory. Unfortunately, I didn't do a similar hand alternation test so I can't compare directly. And at any rate it's not all rolls all day long so your speed total will depend on a bunch of other factors too.

Either layout should be supplemented with some alternative fingering for best speed results - just like I'm sure you're doing in QWERTY already. (Example: NK would be a one-finger digraph so when typing 'think' I slide my hand in so that the middle finger does 'n' instead of the index finger.) This is one of the things you'll have to spend effort on: Learning how every common word is best typed in the new layout, and automatizing the common words again.

If I were you, I'd go with Colemak simply because relearning fewer keys should make you faster in less time. Maybe even consider my Tarmac transitional layout to learn it one hand at a time?

[erricrice]

Interesting.  I changed my silicone caps to my mini hhk blue alps hack and I started getting times in the 90's, from 85's before. It's possible these caps really do make a significant difference.

I think I'm going to have to try some kind of version of this.  I really think I would love it, short throw alps would be great since the actuation point is right at the top!  I think to make it a little easier though, I am just going to put a chunk below the keycap on both sides to stop the key from bottoming out.  Probably not as effective, but meh I'm just trying it for now.  

BTW, where might you come across silicone like that?  I have no experience with this kind of thing so I have no idea where to even start.

Does anyone know what the usual key travel is on laptop(scissor) keyboards is?

[chimera15]

I think I'm going to have to try some kind of version of this.  I really think I would love it, short throw alps would be great since the actuation point is right at the top!  I think to make it a little easier though, I am just going to put a chunk below the keycap on both sides to stop the key from bottoming out.  Probably not as effective, but meh I'm just trying it for now.  

BTW, where might you come across silicone like that?  I have no experience with this kind of thing so I have no idea where to even start.

Does anyone know what the usual key travel is on laptop(scissor) keyboards is?

I'm using ecoflex from smooth-on:

http://www.smooth-on.com/

The bouncy keys are a wonderful feeling.  I'm considering making a sheet of silicone and making some kind of tool that will cut normal key shaped pieces of silicone out to glue to the tops of my normal keys. heh.  It may not improve speed, but the feel is so much better than hard plastic, and I think I've proven at this point that it doesn't hurt speed.

[chimera15]

I still like the keyfeel of a Selectric.  Whiiiiiiiirrrrrrr......  SLAP.  Someday on Ebay I'm hoping one of the rare Selectric Terminal INPUT (not printer) ones like in this link.

It is a cam driven 110VAC motor that does all the work and your fingers just glide along hippity hoppity.

So in other words - if you want the ultimate FAST keyboard you have to motorize it.

Show Image

I know, I was thinking about a motorized keyboard for a long time now, but I don't see any feasible way to do it unless I really get into the mechanics of electrics and find some way for the arms to actuate a keyboard switch instead of typing a letter. lol I think that'd be the easiest way to make a prototype.

I spent almost all day today typing on that darn test and didn't break 103 wpm all day.  I can't even reach my high score yesterday which is frustrating.  

At this point I've started trying to learn dvorak more.  I'm still gaining speed but still in the low teens or high single digit area for speed. My thought is this has to be the key to get a lot closer to 200.  Qwerty just isn't cutting it. Even if I do a perfect run on the shortest of passages my speed doesn't go past 120.  For 200, there's no way I physically can do it.  It would be a whole nother two or three levels beyond what my hands are capable of moving with my current boards.


I hope the green alps will take me up another step, and the dvorak will take me up at least another step, at least then I should hopefully be in the 120-130 average range.

My mantra right now is, "What one man can do another can" from "The Edge" lol There's just something I'm missing. lol

[hyperlinked]

I still like the keyfeel of a Selectric.  Whiiiiiiiirrrrrrr......  SLAP.  Someday on Ebay I'm hoping one of the rare Selectric Terminal INPUT (not printer) ones like in this link.

Oh yeeeeah! Who needs a clicky keyboard when you can have a slammy keyboard? I had forgotten how "easy" it felt to type on those things. Whoever brought up the short travel of the Seletric keys, thank you! It's a fond memory I had forgotten about.

"whuk-KLAM... whuk-KLAM-KLAM-KLAM...whuk-KLAM... whuk-KLAM"

[chimera15]

My thought is that the Selectric allows for faster typing right now.  I'd really like to try one again.  I was thinking though, is there a physical difference and different movement anywhere in the typewriter that would allow for the placement of an alps switch to be depressed on all the keys?

Edit: Actually you might not even need to introduce a switch somewhere.  All you need is someplace where two pieces of metal might come together on the individual letters somewhere were it wasn't before.  You could use that like a switch too.

[chimera15]

I still like the keyfeel of a Selectric.  Whiiiiiiiirrrrrrr......  SLAP.  Someday on Ebay I'm hoping one of the rare Selectric Terminal INPUT (not printer) ones like in this link.

It is a cam driven 110VAC motor that does all the work and your fingers just glide along hippity hoppity.

So in other words - if you want the ultimate FAST keyboard you have to motorize it.

Show Image

Oh I didn't read that link.  So there's a conversion kit for selectrics?  How would that work I wonder?  All you need is to be able to turn the keypresses into a switch somehow, and I would be able to rig a matrix to it...

I need to reread your explanation of how the selectric generates a keystroke again.  Is there already a part that acts as a switch?  As I recall it uses cams to decide what key has been activated?

[chimera15]

It would be pretty hilarious if I actually built one of these things, and actually managed to type 212 wpm on it, and took it down to the ultimate typing championship and walked in with it. roflol

I think it's too late now though, there's less than a month left to the close of the competition.  Why didn't you post that earlier Ripster! lol

[chimera15]

It's pretty interesting.  Apparently Blackburn's record has been broken several times now.  The fastest typer in the world is around 309 wpm.  I need to read more about what they're typing and typing on.

[chimera15]

You are aiming too high, I think. While all this might make the difference between two champs with equal skills, I think it is a limited gain in the end compared to lots of training. Work, work, work, and gain 10wpm in a few months. There is no quick success. Or you need a new paradigm, like typing in syllables or popular words using macros, which is not cheating like in macroing parts of a text, but changing the atomic unit. Consider Japanese, where the syllable is the atomic unit, which is actually quite smart and effective.

I still don't think using macros is fair.  I don't see improving much past 110 wpm without changing the board or format I'm typing on.  I've shown that key type can make as much as a 10-20 wpm difference in raw speed now.  It's possible that a board could make as much as 30 wpm difference.  

  I'm typing as fast as I can on one phrase to get that result.  I doubt much more training with the boards I have now is going to make much of a difference to that speed.  I think there is a max speed to a board switch/cap type, just like a car.   It'd be better to find a board that has the highest max speed, and then work like hell on that to max out its speed.  It's kind of like you're saying, take this vw bug, and train on a track to get it up to 200wpm, it's just not going to happen.  You need a combination of a well performing high speed car, and a skilled driver to get those results.


If you told me what board the current record holder of 309 wpm used, and he was actually typing real phrases like on the typrx website, not using macros, then I'd be satisfied with that, I'd buy that board and train on that, but lacking that information I'm going to try to build a board with the max possible relative speed.


I'm viewing this a lot like nascar, or f1, where new devices are thought of to improve driving for everyone.   I've already hit on an invention of using silicone just for normal keycaps that I plan to use in the future.  A lot of this is producing results that's useful in just normal typing.

The Japanese example doesn't track.  Those are actually letters, and they're not all syllables in fact.  They have vowel sounds and modification symbols, and one consonant n,  that effectively turn the raw consonant/vowel combination into more consonants which doesn't follow exactly for syllables.   Even if you're talking kanji, a lot of characters are still just a single vowel sound, which isn't a syllable when used in combination with other letters.  At any rate you're talking about a different language with completely different sentence and word structure in general.  Words sound completely different to what would be made with a Latin based language.  It's so different you can't even count wpm when you're talking about Chinese or Japanese as a source of comparison, you have to talk in cpm purely to have any comparison at all.

You're saying that I should come up with a new English typing format completely essentially.  That would take a lot of work just by itself.

It's actually probable that this has been done hasn't it?  I think it would still be disqualified in a competition situation since it would take alternate drivers or macros to run on the computer.

[chimera15]

This is a really interesting thread:

http://typera.tk/forum/topic.cws?id=17

[chimera15]

Maybe not in a competition where the atomic unit is agreed to be a latin character. But typing for competition is useless If you'd speed type for business it is just another paradigm and who cares how you get your wpm. I was looking at it from that point of view. Sure it has been done before. Chord keyboards. Velotype, etc. Goes back decades (stenotype).

Show Image

Unlike most countries, Dutch TV subtitles foreign languages. This is what they use for live translations.

So that's an interesting point, how does a stenotype keyboard work.   Why doesn't everyone use it?  What's the fastest typist with a stenotype type keyboard? lol

So wikipedia says the highest speed is 375 wpm with a stenotype keyboard.  But it's in shorthand? So it's not full english words?

[chimera15]

Heheh I wasted years on MUD. That is indeed a great if utterly outdated way to improve your typing. Especially when action oriented, natural speed training. But also, the smartest players and coders used macros and "quicktypers"

Yeah  I used gs4 as a way to get the basics of dvorak down.   I still plan to use it to increase speed. It's still a handy tool.  I spent 15 years playing. It's the main reason I can type at all.  Of course kobold and rolton doesn't come up much in normal typing. lol

I really wish there was a Japanese text mud so I could really get fluent in Japanese. lol  Playing gs was like learning a whole other language practically for me anyway.

[chimera15]

I went with my idea of making a set of silicone steampunk keys....


So they absolutely work.  I went with a harder silicone, poyo putty, it's about a shore hardness of 20-30.  I bounce off the keys at high speed.   I'm turning in averages of 100 wpm now on medium difficulty passages.  

I didn't have enough keys from my last steampunk mod, so waiting for the set in the mail.   I'm short about 10 keys, which you can see.  It'll leave me 46 which should be more than enough to do another set of prime keys for my green alps board as well as leave these as they are.  Next set I'll probably use silicone color to make the putty black.

Anyway, it's totally awesome, the keytops are like magic.  I'm really curious to see how I'll do with the green alps and these.

I think it's also helping that the keytops are completely flat, which may be one of the reasons I was faster with the irocks board as well.  With no little projections you can move your hand closer to the board and not get caught when going for the next key.

Once I get the rest of the key bottoms I should pick up even more speed because then all the keys will be perfectly uniform which should help a bit.

Probably with the final ultra speed board I won't rig up any unnecessary keys, so the board won't have like an function, alt, control, tab, or possibly even a return key which aren't used in competition.  Just like a nascar competition, there's no reason to clutter a board with keys that won't be used and could potentially interfere with advantageous finger placements or keystrokes.

[ricercar]

Complete win! I love it.

[msiegel]

the keytops are like magic

i think you're onto something

maybe one of our resident mold-makers can make a set of these... chimercaps

[msiegel]

aha, a punch
http://geekhack.org/showpost.php?p=139203&postcount=44

[ak_nala]

Following up –


Olympia SM-3 portable, circa 1956 IIRC.


Fully sprung key tops on all alpha-numerics and the shift keys.


Mmmm, chocolate double-shots


Looks like the exact same mechanism as on the full-sized SG-1 typewriter.

The alpha-numeric key tops on the portable SM-9 are running exactly 1/2" wide at their widest point, while the full size SG-1 key tops are just a hair under 9/16" at their widest point.

I read somewhere that starting with the SM-5 they did away with these springs (verified by there being none on my SM-9) and I have no info on the SG-3, so if looking for spring loaded key tops I'd stick to either a full-sized SG-1 or a portable SM-3 or SM-4.


Note that the keys on the SG-1 are actually quite a bit darker than this. Really a very, very dark olive green that looks almost black when viewed from above under normal room lighting.

[chimera15]

I had not tried my cherry g84-4100ptmus before cause it was out in the car with my carputer.  I decided to finally give it a try today, and I'm way above my average scores, most around 95-100.  Short travel linear is definitely the way to go for speed.  I think I'm a bit slower than I could be cause it has a slightly compressed layout and a small backspace.

I also tried the scorpius m10 with blue cherries, I would say it's about an average board for speed.   I averaged around 80wpm with it. I was surprised at the lightness of the blue cherry switch.   I'd say it's about an 80wpm board for me.

[chimera15]

Some update on this.  I couldn't get the silicone even enough the way I was doing it.  I remembered I used some of this foam from Micheal's before on projects and it worked well for things that needed to be soft, so I started experimenting with it.  I deswitched my ortek mini.  I never felt comfortable with it because it didn't have a number pad and it had the European return.  It's perfect to use different switches on though because you never have to press return on any of the passages, nor any function keys.

This has the foamed caps and green switches with silicone in the bottom which shortens their travel, and gives them extra bounce when they bottom out.  I wasn't super happy with the results, but I might find it better the faster I get.


Ultraspeedboard hack mk2




As you can see the foam is much more even than the silicone was.


Finally this is Mk3 incorporating all the advantages from all the tests I've run.  It has complicated whites from the ortek that I pulled for the main row, and the round typewriter key bottoms, and foam.  Some adjustment will still probably be necessary for the foam, I think I need to add another layer so the edges of the keys don't hook my fingers as I travel over the keys. So far I haven't fun a full test with this board.  I also found with the elevated angle of the wireless hack, my speed improved a lot.  Angle seems to have a lot to do with how fast you are, cause it effects overall hand position.  For me and the way I'm sitting with the board in my lap, a high angle really helps, which is beyond what is capable with the feet, so I made a home made riser out of some foam board.


Ultra high speed board mk3:

[chimera15]

So far not bad, I added the extra layer of foam to the keys. It's about 100 times more comfortable than a normal minitouch, I haven't reached the 125 wpm speed I reached on my wireless minihack yet though...

So far:

12   zzephyr   117.9 wpm   
13   ranger296   117.0 wpm   
14   chimera15   114.9 wpm   
15   Mustang   112.5 wpm

heh.

[chimera15]

I gotta look for that foam sometime.  That does look pretty slick.  What brand is it??

I'm thinking it might be a way to muffle key noise by squirting the stuff into the bottom of keys.

It comes in sheets..not liquid. It's called "creatology fun foam".

[chimera15]

nmstin   130.1 wpm   
12   chimera15   129.2 wpm   
13   aoeuaoeu   124.6 wpm   
14   astinus   123.6 wpm

Almost broke 130! On my version 3 with cream dampered switches, with clicky leafs, and old typewriter keys.

Someone should make cream stem/clicky spring alps and make them alps reds. lol  Really awesome feeling switches.

[chimera15]

Heee! roflol



I think right now that It's reasonable to think I could get around 150 if I type perfect with this board, but It'd really take some doing, I almost got that quote perfect as you can get it, I was just a little slow around the beginning of it.

[ricercar]

Awesome. If you're the one with a ton of extra G keys, could they be modded for replacing the square keycaps?

[chimera15]

Awesome. If you're the one with a ton of extra G keys, could they be modded for replacing the square keycaps?

G keys? Which square keycaps? huh?

[ricercar]

Maybe I'm dreaming, but I thought you posted a typewriter key mod pic of keys that showed a normal set, and then another pile of G keys aside the other, of equal size. Of course I can't find it NOW.