F104+SSK+122+62+77+50+Ergo orders now open! New Kishsaver+Industrial Model F Keyboards

[engr]

In the future have you ever given any though to releasing an improved layout Model F AT/XT form factor (function pad left, nav cluster right. Basically an improved XTANT?

Isn't that basically F77 model?

[Muirium]

Some folks are just mad for those columns. The only use I have for them on my OG IBMs is somewhere to put the missing modifier keys!

[SunshineRag_DT]

Isn't that basically F77 model?

Not quite as if you look it has 87 key layout, so it has an additional 10 key macro pad on the left. So F77+, basically the same layout as an Model F XT or AT. The layout of a standard Model F XT fits this with a touch of space on the right edge. I am not sure it will fit a 4U numpad layout on the right as I think the F AT is a wider to accomodate this.

The Wheelwriter has this basic layout as my mod illustrated, and so does the F107 if you chop off the numpad like orihalcon did. After the Model F IBM moved the function keys to the top like they are now in modern TKLs.

The only use I have for them on my OG IBMs is somewhere to put the missing modifier keys!

I'm sure you are right. It is just that the XT and AT are the Model F, I would love to have modern layout with these. I actually use function area for my music controls and macro pad and missing modifiers, and I use the modifer numrow layer for the actual function keys. For me it is basically having a built in macro pad. But if you don't want a dedicated macro pad, the F77 and layers can certainly do the job.

Basically the "new Model F" project has never actually produced a updated Model F style layout, basically an AT with all the modern layout updates. I will bet a updated XT key assembly that is drop in compatible with current XT would be very popular.

[SunshineRag_DT]

Also it would be cool if Ellipse could produce a split spacebar mod with Wheelwriter style "Code" size key and a smaller spacebar that will actually allow the RALT to be used. Such a spacebar does not currently exist as the current small space does not fit the area after placing the Code key but you lose the RALT.

[kbdfr]

Some folks are just mad for those columns. The only use I have for them on my OG IBMs is somewhere to put the missing modifier keys!

In the old days, those columns were for the (then 10) function keys. You could use them with your left hand to type any combination of the three left modifiers while not even having to look at your keyboard. It was a simple possibility to make use of left-hand muscle memory to type combos like Ctrl-F4, Shift-F5 or Ctrl+Shift+F8.

And then Microsoft had the silly idea of replacing those two columns of function keys with a row of 12 of them, so that now using function key combos means having to use both hands and/or to look at yoiur keyboard. Certainly not a huge improvement for (touch) typing professionals.

[Muirium]

Isn't that basically F77 model?

For me it is basically having a built in macro pad. But if you don't want a dedicated macro pad, the F77 and layers can certainly do the job.

Indeed! I've found uses for those two columns of keys on my XT, AT and 3278 over the years. Mostly as (Karabiner remapped) media keys, similar to you:



I mean, it's a waste not to use them. But in reality, I don't often use them indeed. That's because in the land of layouts: muscle memory is king. I have the same suite of media keys on the function layer of all of my keyboards—including the 60%s like this one I so often use—that hitting them is automatic. Consistency is key: one function layer to bind them, and rule them all.

And, well, you can hardly go beyond the 60% block for your core muscle memory layer when you yourself prefer and heavily use 60% keyboards!

Basically the "new Model F" project has never actually produced a updated Model F style layout, basically an AT with all the modern layout updates. I will bet a updated XT key assembly that is drop in compatible with current XT would be very popular.

That's actually quite a nifty idea. Drop in upgrades for all those XTs out there! They could sorely use it, that awkward layout of theirs is the reason there's such demand for this project.

And then Microsoft had the silly idea of replacing those two columns of function keys with a row of 12 of them

I've always laid the blame—and the credit—on IBM for the 12 key function row. They introduced it on the Model M, as far as I know, especially in its segmented 3 × 4 key block, with alternating key colours. It seems a bit too well thought out for 1980s Microsoft.

I agree with you that the function keys on top arrangement is worse for chording. They've always struck me as a "one-hit" zone for the keyboard: intended for clustered, single-press functions, with room for a function template overlay to sit on top, to do all the explaining. (Overlays I myself never used, because I seldom ever had licensed software!) The fact that by far the most used function key for me back then was in the awkward combo Alt + F4 was indeed on Microsoft.

[kbdfr]

[…]

And then Microsoft had the silly idea of replacing those two columns of function keys with a row of 12 of them

I've always laid the blame—and the credit—on IBM for the 12 key function row. They introduced it on the Model M, as far as I know, especially in its segmented 3 × 4 key block, with alternating key colours. It seems a bit too well thought out for 1980s Microsoft.
[…]

You are of course right.

[Ellipse]

SunshineRag_DT I have ordered the Code keys, PC AT Big Enter, and non-stepped ISO Enter - feel free to sign the interest form here:

https://docs.google.com/forms/d/1vsamkl ... kDG39r08Q/

I have no plans for Model M kits or recreations of other IBM layouts, though everyone can order parts that are compatible with original Model F's.

[DMA]

Mmkay, I borrowed an USB wattmeter from mah neighbor and measured actual power consumption as a service to community. 5.1V, 30 mA - not even close to requested 200mA (however, if solenoid board is connected, all bets are off. I don't have solenoid board on hand to check).
Then I powered it by mah programmable power supply - 5.00V, 9mA. May be it's lower because mah power supply can't talk USB, but still way lower than 200mA requested.

Mother board USB ports are normally not usb hub ports. They are host USB ports. Host USB ports can exceed 500 ma. The minimum they have to provide is 500 ma.

USB 2.0 standard section 7.2.1

If I had a dollar every time USB 2.0 spec is violated, I'd be richer than Bezos and that something-hathaway guy _combined_. In practice, there's _usually_ a resettable fuse on the host port (it's required to be resettable by 7.2.1.2.1, but see previous sentence..) somewhere north of 500mA (but not exceeding 5A). In any case, hub is never the culprit - one has to cause 18 bit times delay (1.5 microseconds) - which will require about 6 hubs connected by max. length cables (which are about 7 meters - although they specified via max. cable delay of 26 ns (18ns for low-speed devices)).

[Muirium]

measured actual power consumption as a service to community. 5.1V, 30 mA - not even close to requested 200mA

Nice number. Is that an Xwhatsit controller or your own hardware?

I assume there’s a difference. And I imagine there may even be a (smaller) difference between firmwares (Pandrew’s vs. Arkku’s) and even maybe a difference between typing / inactive / multiple keys held down.

But that figure should be the ballpark.

[DMA]

measured actual power consumption as a service to community. 5.1V, 30 mA - not even close to requested 200mA

Nice number. Is that an Xwhatsit controller or your own hardware?

mini-xwhatsit with QMK firmware.

I assume there’s a difference.

Of course. Mine is 24mA full-bore (no power-management effort whatsoever when externally powered).

And I imagine there may even be a (smaller) difference between firmwares (Pandrew’s vs. Arkku’s)

Could be - if MCU clock speed is different.

and even maybe a difference between typing / inactive / multiple keys held down.

Nope. Theoretically, slightly more charge flows via pressed key than released one, but it's not detectable in practice.

[Muirium]

Ooh, I like these numbers!

How about QMK with VIA or Vial on top? I’ve heard that extra layer in the stack taxes the hardware quite badly, but that may well be purely memory rather than actual work.

[SunshineRag_DT]

SunshineRag_DT I have ordered the Code keys, PC AT Big Enter, and non-stepped ISO Enter - feel free to sign the interest form here:

Ellipse thanks, one thing I was noting about the spacebar is to make the code key usable with current layouts it probably needs a inverted small spacebar. I understand that this is likely not going to happen but wanted to note that if a small spacebar was made with an inverted post and stem location it would allow the "Code" key to fit in current layouts without sacrificing the RWin. See the attached pic. Obviously this would require a new small spacebar mold but the size will fit..

Just wanted to note this if ever interested in newer split spacebar layouts.

Small_Spacebar_w_Code_Key.jpg (721.33 KiB) Viewed 13523 times

[Ellipse]

Has anyone modded their spacebar to be inverted as shown? Is it just a matter of adding metal tabs to the other end, to support the reversed wire stabilizer, and then removing the current stem and gluing a Model M stem from a two-part key?

For folks who have signed the code key interest form, how do you plan on using the key?

https://docs.google.com/forms/d/1vsamkl ... QkDG39r08Q

[DMA]

Ooh, I like these numbers!

How about QMK with VIA or Vial on top? I’ve heard that extra layer in the stack taxes the hardware quite badly, but that may well be purely memory rather than actual work.

Computers don't work that way.
Well, _modern_ computers do, but not MCUs. MCU consumes same power, whether it has anything to do at the moment or not.
It goes so far that in modern MCUs power consumption is usually specified in "uA/MHz", so you can adjust your power budget at design time if you need to. Some MCUs even allow to dynamically change their clock speed at runtime.
Instead, MCUs have sleep states (usually more than one even) - but QMK, famously not giving a flying fuck about power consumption, doesn't use any of that. So, via or not via, power consumption stays the same.

[Meowmaritus]

SunshineRag_DT I have ordered the Code keys, PC AT Big Enter, and non-stepped ISO Enter - feel free to sign the interest form here:

https://docs.google.com/forms/d/1vsamkl ... kDG39r08Q/

I have no plans for Model M kits or recreations of other IBM layouts, though everyone can order parts that are compatible with original Model F's.

Are you really charging $99 for one fucking keycap my guy?

[Ellipse]

Meowmaritus as noted on the form, it is $99 for the key set, not one key. About a $20 premium to cover the tooling and mold costs.

[affenstark]

Hi guys,

Trying to compile my own custom layout, which I have to do myself apparently if I want NKRO...

I have followed the instructions in the manual to install qmk (I'm using qmk msys), copied the files and edited rules.mk and config.h, but I keep getting errors on running qmk compile.

First: common_config.h not found, so I just commented out #include common_config.h.
Now I have: "keyboards/xwhatsit/matrix.c:523:36: error: 'keymaps' undeclared (first use in this function)" (and again matrix.c:588:32).

Does anyone know what I can do about this? / why it isn't just working like it seems it should from the guide and from responses here?

[Ordinary Witch]

qmk itself has recently gone through a refactor which largely upends the file structure used for firmware. from what I understand, there's an ongoing effort to port the current qmk firmware to the new structure.
I personally don't know much more than that about it, as it's difficult to experiment when I don't have a QMK keyboard to test on (yet).

[Ellipse]

affenstark maybe try using an older version of QMK. If you download pandrew's git (instructions are in the manual on the project web site) the version of QMK that is included is older and it should work.

[affenstark]

Thanks guys, that was the problem it seems.

At first I thought I need to downgrade qmk by doing "pip install qmk==0.4.0 --upgrade" for example, but instead, after digging through google and reddit for a while, I found that I need to go to ./qmk_firmware and "git checkout 0.18.10", or well, presumably 0.18.17 or prior, because on the qmk_firmware github the next version is marked as "breakpoint", probably the version Ordinary Witch is referring to above.

This downgrading checkout step might be useful to add to the manual. I'm not sure if there's something I'm missing but maybe someone can add some further context if needed.

[Ellipse]

Thanks affenstark - I have updated the manual with these recommendations.

[Owlteeth]

Hi all,
my F77 has started to exhibit a behaviour where typing a space sometimes results in two spaces. It seems sensitive to the speed at which I'm typing and where exactly I hit the key, so I think it's a mechanical issue. I can reduce the problem by limiting key repeat rate in Windows, but it's not perfect and has annoying side effects. Has anyone encountered this and solved it mechanically?

[Ellipse]

Yes please see the manual on the project web site as there is a guide specific to adjusting the key springs and space bar specifically. Also update to the firmware on the web site if your keyboard was shipped a few years ago.

[affenstark]

Hi again,
I have another problem, this time with the solenoid.

My solenoid doesn't work for some reason even though it is all connected. The solenoid controller light flashes when I press keys (and if I toggle solenoid off, it no longer does this). I have of course tried to increase the dwell time with the keybind, which didn't help. Maybe the solenoid itself is just broken?
Is there anything I'm missing?

[dr_xadium]

Could it be insufficient voltage draw off the USB port? Maybe there's enough coming across to light the led but not enough to actually drive the solenoid?

[affenstark]

Thanks for the suggestion dr_xadium.
I do have my keyboard connected to a USB hub, so I thought maybe that could be it, but it still doesn't work even when connected just to the normal USB ports on my PC. I also used usbdeview to check the power draw and it says it's getting 500mA, which apparently is the maximum? So it seems to me like the problem must be something else..

[Ellipse]

affenstark you can also try powering the solenoid directly with a 6v / 9v battery and some header pin wires. It is possible that you may have overtightened the screws while installing the solenoid and punctured through the blue tape - I am aware of reports of this.

[affenstark]

Thanks, Ellipse, I tried using a battery and couldn't get it to work. (using 2 CR2032s because i don't have a 6v or 9v battery lying around atm.)
It looks like you were right though about the tape being punctured though from overtightening the screw from the bracket.
Does this just mean my solenoid is broken or is there anything I can do about it?

[Ordinary Witch]

If nothing else beyond the tape was broken by the screw, you might be able to fix it again by dismounting the solenoid from the keyboard and patching the broken area with electric insulating tape. This is based on the assumption that the blue tape is there as insulation to prevent the coil from shorting. I don't know if it'll work, but it's worth a try