Miniguru progress

[lowpoly]

Hi,

working on it again.

Edit:



(Edit: Original thread on GH)

Stick prototype:



The screw will be countersunk in the end.

The stick is made from stainless steel. It is so thin that it will fit through the gaps of the keycaps with hopefully no modification to the keycaps necessary. Not a joystick, it almost doesn't move.

I was never happy with the cat tongue caps as they get dirty fast and the sand wears off. The rubber ones I have are kind of slippery. This one here presses a bit into the skin which provides the friction for horizontal pressure. I'll do another shape, slightly wider, during the next days.

Damned ZIF cables, 0.5mm pitch:

[Muirium]

Looking good. The hard part has got to be pointer speed control, I imagine?

[lowpoly]

Only if I want to do fancy stuff like scrolling when the pressure is higher. It has its own controller that works pretty well.

[matt3o]

are you sure you can fit it in the gaps between keys? there's really little room there (especially with DSA and SA)

[lowpoly]

I don't have these yet, only DCS (Round 3, I think that's correct?).

The stick part of the stick is 1.5mm. There's 1.05mm space between the caps with 18mm caps if the corners are really sharp.

In the worst case you have to file a tiny bit off the lower corner of the D and H caps.

[lowpoly]

Even with sharp corners, there's a 1.32mm diameter circle between 3 caps like G, H, B:



The smaller circle is 1.32mm, the bigger 1.5. Cut corners are 0.24x0.17mm.

Of course, that's all theory.

[matt3o]

that's tiny. well, hope it works.

[damorgue]

Will be following this. Have you accounted for tolerances and wiggle? What I mean is that you probably want a few tenths of a mm there just for clearance, and then there is an additional few tenths of a mm because the stems and caps wiggle a bit.

[lowpoly]

With my DSC caps it looks (and feels) like there's enough space. But only a real test will show whether the keycaps will scratch the stick due to wiggle. Still, if a keycap mod is required it will be much easier than the big hole that was needed for the "old" style:



And while this stick is easier to manufacture it would still be possible to make the traditional 3.5mm square version.

Unrelated, I got "scroll on press" working which rocks. Unfortunately, this requires doing the pointer movement filtering myself.

[Muirium]

Maybe I should send you a few spare SA caps to test. (I don't have any DSA, yet.)

[HzFaq]

I can send some DSA caps, I've got some left over from my ergodox. Send me a message if you're interested, happy to help with this awesome project.

[cookie]

I am curios how this will look on a keyboard

[suka]

Nice to see you back, lowpoly - guess I owe you big time for giving me the initial inspirations to build my own boards after all

Back to topic: I've had good experiences with even thinner trackpoint stems so far - so good, in fact, I still haven't changed the top of the pin in over half a year now - always too much other ideas interfering with finalizing those details.
But I would really try to use this approach with a tiny metal mid-piece to avoid any interference with the caps, if you keep it short enough (e.g. the 4mm travel plus a bit) you should not even feel the flex at all.

Even PtS works with it, but I must admit for prolonged usage I tend to use a separate button instead of hurting my fingertip...

[lowpoly]

Thanks for the keycap offers. I need a case and pcb first or it won't be a realistic test. I'll get back to you then if that's OK.

I am curios how this will look on a keyboard

Me too. I was actually so curious that I was going to drill a hole through a G80 keycap donor, then saw it was an ANSI board and stepped back from that idea. I should do a shorter fake stick on the lathe.

[lowpoly]

I've had good experiences with even thinner trackpoint stems so far - so good, in fact, I still haven't changed the top of the pin in over half a year now - always too much other ideas interfering with finalizing those details.
But I would really try to use this approach with a tiny metal mid-piece to avoid any interference with the caps, if you keep it short enough (e.g. the 4mm travel plus a bit) you should not even feel the flex at all.

Good point that it doesn't have to be that thin below the keycap travel point. I'm not using that to the full extent now although the 'clamp' even enters the pcb. I tried to bend the 1.5mm stick with my hands - no way. Thinner should be possible. I have to see if I can manufacture this on my crappy lathe.

Even PtS works with it, but I must admit for prolonged usage I tend to use a separate button instead of hurting my fingertip...

Yep, the head is 3mm now and pointy. Will try 4 next with a flatter shape and maybe some friction pattern filed/dremeled into the top

[damorgue]

Approximately how large is the movement of the 'pinhead' at the keycaps upper surface? Just curious.

[lowpoly]

I'd estimate <= 1mm with medium to hard pressure.

[lowpoly]

I am curios how this will look on a keyboard

Fake with brand new Filco "dust magnets":



Not sure if I got the angle right... Can be closer to B in the end.

In other news, I got 'tap for left click' working. Still needs a lot of fine tuning to be useable (movement and scrolling too). Kind of convenient for surfing.

[cookie]

I wish I could get used to Trackpoints, your Idea is verry smart not to cut the caps but benefit from the space between them.
Never leave the home row... never...



I am curios how this will perform and if you need to cut the caps anyway.

[lowpoly]

I am curios how this will perform and if you need to cut the caps anyway.

A few weeks from now we should have a prototype. If we proceed as we currently do.

Here's the new stick, 4mm head and 1.1mm shaft:



Starts to look like suka's nail. However, with this material 1.1mm is too thin. I could bend it easily while it was still in the lathe. That means not much gain by making it thinner when there is more bend. It works though, still less than 2mm bend at the top. How much it will bend in the end - have to try it. The 4mm head is quite comfortable, much better than the first version.

Pointing, press-to-scroll and tapping works. There's still room for improvement, like negative inertia, but for now this should do.

[cookie]

I see no chance to get this in a HHKB
I think I like the flat hat more then the round one!

[Muirium]

Hmm… I wonder if this could be hacked into a NovaTouch? Or really anything without discrete MX or Alps switches.

[cookie]

I think the main problem will be the controller... If you attach a custom controller then it could be possible but without I see no chance to make it work :/

[Muirium]

You forget that USB is your friend. Attach both keyboard and pointer controllers to a USB hub inside the keyboard…

I think the harder part with Topre is how the whole rubber sheet construction works.

[cookie]

Okay, you could solve the problem with a internal USB hub as well.
On topre, the rubber sheet is no problem but I don't think there is enough space for the trackpoint to fit into.

On realforce keyboards you have space in the area over the F keys, there you could place the pbc and somehow attach the Trackpoint to the metal backplate and connect them with a wire.

But I can't see this on a HHKB unfortunately :/

[lowpoly]

Okay, you could solve the problem with a internal USB hub as well.

Hub is in the case.

But the problem is build height. Between GHB there is just not enough height to the ground. We have a big hole in the pcb for the clamp (the part with the screw) so we don't lose the pcb height. And another small hole in the case. It is really tight.

In a mechanical board it may work with re-routing some traces. In a Topre it is hard to say how it will affect the capacitive sensing.

You will also need a breakout board for the ZIF cable connector because soldering wires to it is kind of unfunny (see pic in first post).

[mSSM]

Incredible! I am very excited about this. Way before I actually got my first mechanical, I was following the progress on the miniguru with great interest. I was quite sad when you put the project on hiatus.

Now I am typing on an SSK and an HHKB, while my ErgoDox waits to be assembled. Your miniguru (with trackpoint!!!!) might be my fourth.

BTW: 7bit had some plans for his qHack (lots of other q-names for various layouts out there), which was also supposed to incorporate a trackpoint. Maybe you guys can collaborate? Just throwing it out there.

[jacobolus]

Have you folks tried moving the pointer to be directly next to the home row? I think in between the index and middle finger should work well (if it were on the other side of the index finger it might accidentally be triggered when moving the finger to press the inner column).

More concretely, assuming a Sholes/QWERTY layout, I’d put it either at the corner of JKM or at the corner of JUI.

Then you basically don't have to move the hand at all to use the pointer, unlike with the standard Thinkpad pointer positioning. [Or if you want to get really fancy, add one on each hand, and use either one or assign them separate functions and use both.]

[Muirium]

Right stick points. Left stick scrolls!

I'm addicted to full 2D scrolling from my Apple stuff. This would be one way to get it…

[cookie]

Jeah but they did it with a scroll ball, that is actually verry smart and feels natural. Horizontal scroll wheels are the biggest garbage ever invented!

I need to get my hands on a thinkbad keyboard to try the trackpoint... better not, what if I like it? Then I am screwed