Model F angle and curvature?

[fanf]

What is the radius of curvature of an IBM Model F plate? (Are they all the same? Is the Model M the same?)

I found a post on Reddit that said “about 10.9 inches”. That implies (very roughly) the main block of a keyboard subtends about 20°, which is close to the typical difference in angle of the tops of sculpted keycaps. So it seems plausible. Anyone have a more accurate (or authoritative) number?

Also, for the 5150 and PC AT versions, what is the angle of the top of the case when the feet are retracted?

[Muirium]

What Model Fs (if any) do you have yourself? I’m thinking a good diagram showing us what to measure could be helpful. I’ve a few Fs and can measure them sometime, but I’d need to know where.

Circular radius of a short concave arc is tricky.

Oh and what are you up to with these measurements? Sounds like the sort of thing you’d need to know for a good project…

[fanf]

I don’t have any myself, which is why I ask I have scrawled on some pictures from clickeykeyboards.com …

metal plate curvature

New Note.png (618.66 KiB) Viewed 13159 times

I guess the easiest way to measure the curvature is to measure the width w and the distance between the straight line and the middle of the curve h

side view of a Model F terminal keyboard

New Note 2.png (507.53 KiB) Viewed 13159 times

For the angle, it should be enough to measure the height at a couple of places h1 and h2 and the distance between them w.

I am toying with some case ideas, that don’t look like a rectangular billet of aluminium. I wonder how well it might work to have a curved top that echoes the keycap profile. If that doesn’t work out, I really like the gentle wedge shape with round front and back that many Models F have. Seems like a good starting point would be to copy the original.

[Muirium]

Cool. I'll get you some data when I can. I have an AT and a 4704 Kishsaver to measure.

[clickykeyboards]

Some model M and model F backplate reference photos with ruler.

Not sure what the project is or what the measurements are for, but @fanf contact us at clickykeyboards@gmail.com if you want to buy/borrow these spare parts for more exact measurements


IMAGES 1, 2: left side (silvery metal plate from 1988 model M, right side (black metal plate from 1983 model F)

IMG-4899.jpeg (318.99 KiB) Viewed 12953 times

IMG-4912.jpeg (340 KiB) Viewed 12953 times

<BR>
IMAGES 3, 4, 5 show model M plate (1988)

IMG-4906.jpeg (342.2 KiB) Viewed 12953 times

IMG-4907.jpeg (321.26 KiB) Viewed 12953 times

IMG-4908.jpeg (221.47 KiB) Viewed 12953 times

<BR>
IMAGES 6, 7 8 show model F plate (1983)

IMG-4909.jpeg (323.99 KiB) Viewed 12953 times

IMG-4910.jpeg (279.27 KiB) Viewed 12953 times

IMG-4911.jpeg (219.92 KiB) Viewed 12953 times

-----

[fanf]

Magnificent, thank you!

So for the Model M it’s 171mm wide and 12.5mm high, so it subtends an angle of 33° and has a radius of curvature of 300mm, or 11.8 inches (35 barleycorns or shoe size 12 for Chyros’s leftpondian fans). The bigger angle is not too surprising because it has to cover the F key row and other extras that I left out of my guesstimate.

For the Model F I’m going to say 124mm wide (taking off a couple of mm for the flanges) and 8mm high, so it subtends an angle of 29.4° and has a radius of curvature of 244mm (9.6in). Part of the reason for the shorter radius is the difference between base plate and switch mounting plate, but I doubt that is much more than 10mm, so the numbers say the Model F is more scooped than the Model M. Hard to tell from the pictures, tho!

I think my conclusion is that I shouldn’t worry too much about getting an authentic scoop, because there’s a range of values that are plausible and they are in the ballpark of MX keycap profiles. I guess I should get CADding and find out where the idea is on the hideous/heavenly scale.

[ifohancroft]

How's the project going? I'd be interested to see what you made in CAD and so on

[fanf]

Progress has been slow

I’ve retreated from the idea of a profiled top (whether Model F or AEK style) because I think it works best with fairly rectangular key blocks, and I would like to get away from samey rectangles. I have since found out about the Lunar II which does an impeccable job of matching the curve of the case to the keycap profile (Apple ALPS).

I realised that my old idea for a G80-1800ish arrow cluster doesn’t work with the idea of being able to reach the modifiers with my thumb. It makes more sense to have the modifiers closer and the arrows further away owing to thumbs being shorter than fingers. Then a desire for symmetry and some XT envy joined the party and this was the result

keybird42 sketch.png (111.94 KiB) Viewed 10854 times

I want to see how that top plan works with a round front and back, inspired by various Models F, but probably with uniform thickness instead of the back having a larger radius of curvature than the front.

I have started learning build123d which is a CAD system programmed in Python, closely related to CadQuery. I previously tried CadQuery and didn’t get on with it, but build123d is much more to my taste. You can use both styles in the same model because they share the same underlying CAD system, so I have been playing with Matteo Spinelli’s OPK keycap model, translating it from CadQuery to build123d. The code so far is a lot easier to read and a lot shorter: encouraging!

[Muirium]

Lunar II looks very nice, but goodness me that price!

The Matteo of which you speak is Deskthority’s own Matt3o btw. He’s been a highly valued contributor to the community for more years than is polite to talk about in public! I was there for most all of them.

[fanf]

I get the impression that ai03 aims for the higher echelons of keyboard design

I’m a fan of matt3o’s work: I have a nightfox keyboard, and I am currently waiting anxiously to get hold of MTNU Susu. (I missed the groupbuy, d’oh, so I am hoping to pick up extras.)

Both ai03 and matt3o’s design guides were really helpful when I was getting started.

[AndyJ]

I'm not that impressed with the curvature on my Model Fs and Ms. I think it's a relic of the typewriter days, when the "proper" position of the keyboard was much lower than modern desktops.

I don't see anything wrong with a flat keyboard and flat switches. Laptops and some "compact" keyboards are flat, and I don't see people complaining about them.

If I was building a keyboard from scratch, I'd just make it flat. Well, if I was using a custom circuit board and Model F switch flippers and barrels, anyway. And IBM keycaps are all the same, none of that "row curvature" stuff.

[fanf]

It’s a matter of personal preference and people are allowed to differ, of course I have been converted to the way of the sculpted profile for a few practical reasons I wrote about earlier this year https://dotat.at/@/2023-08-05-keycaps.html

[Muirium]

The flatness of typing on my Mac's own keyboard is one of the worst things about it, for me. Sure, the keys are satisfaction-free chiclets, but I like the layout and the backlighting remains the gold standard of that stuff for me. No animating rainbow-snakes here! Yet it's that damn flatness, accentuated by flat caps and flat dullard scissor switches, which make it feel so dead to me. The HHKB I pull out any time I've the laptop out and about, opened up for mobile use, feels so much more alive. Yes, it's Topre's magic, but it's also the shape and position of these keys. They're right there where I want them, not lost in that slab of samey flatland blandness.