Page 1 of 1

Trackball "aball"

Posted: 05 Jun 2021, 19:31
by polemon
Hi, I recently came across this: https://github.com/brickbots/aball

https://raw.githubusercontent.com/brick ... l_min.jpeg

Pro Micro based and doesn't use scavenged Mouse parts, which is nice.
I don't quite like the 3D printed cradle with the ball bearings, though.
I'm pretty sure the design can be further improved using ball transfer units in a ring-structure, etc. Also, a Russian Pyramid ball might be a better choice instead of a regular pool ball:

https://upload.wikimedia.org/wikipedia/ ... arison.jpg

American (standard) Pool Ball: 57mm diameter
Russian Pyramid ball: 68mm diameter (the same balls are used in a Finnish game called "Kaisa", and they're still being made - so it's not like we need to raid ancient graves of some Zar or something)

Perhaps some of the CAD pros here can make this thing into something wonderful, who knows.

Further improvements could be made in software perhaps, such that a z-axis rotation can be implemented. Not sure what for, but why not, having a rotation around the z-axis seems to me like something "nice to have", even if it's ignored by most programs and OSes.
Bunch of buttons would be nice, perhaps something for scrolling, too.

(don't wanna link the images directly, and I don't want to re-host them elsewhere, so links to images it is…)

Re: Trackball "aball"

Posted: 06 Jun 2021, 00:13
by Archie
polemon wrote:
05 Jun 2021, 19:31
Not sure what for, but why not, having a rotation around the z-axis seems to me like something "nice to have", even if it's ignored by most programs and OSes.
Z-axis is quite standard signal for ages, interpreted by most systems and applications as scrolling and zooming.

Re: Trackball "aball"

Posted: 06 Jun 2021, 09:02
by polemon
Archie wrote:
06 Jun 2021, 00:13
Z-axis is quite standard signal for ages, interpreted by most systems and applications as scrolling and zooming.
OK, I only know of CAD mice and controllers implementing z-axis rotation.
Regular input devices, I don't know of any, aside from things like joysticks, and perhaps the rotation on Wacom digitizer tablets. But I haven't seen a "regular" mouse or trackball, sending z-axis rotation to the OS, except the aforementioned special cases where it's used for CAD. However these devices require special drivers that cannot be used ootb as a standard pointing device.

Re: Trackball "aball"

Posted: 06 Jun 2021, 09:59
by Archie
polemon wrote:
06 Jun 2021, 09:02
But I haven't seen a "regular" mouse or trackball, sending z-axis rotation to the OS,
https://www.kensington.com/p/products/e ... trackball/

12 years in production.

Re: Trackball "aball"

Posted: 06 Jun 2021, 15:42
by Findecanor
I would prefer a modifier button for scrolling. When held down, you could use the ball to scroll both vertically and horizontally.

There is a way to make the USB mouse protocol scroll smoothly, i.e. in smaller steps. It was introduced ad hoc by Microsoft ... but should be supported by all major OS:es now. There is no support for it in TMK/QMK, so you'd have to add it.
Archie wrote:
06 Jun 2021, 00:13
Z-axis is quite standard signal for ages, interpreted by most systems and applications as scrolling and zooming.
If you'd use the USB HID code for "Z-axis" then it is likely to be interpreted as horizontal scrolling. That's because some mice with horizontal scrolling do this: Apple "Mighty Mouse" and I think also IBM Scrollpoint, while most other use "AC Pan".
(I wonder what protocol Spacemice use)

Re: Trackball "aball"

Posted: 06 Jun 2021, 23:36
by Archie
Findecanor wrote:
06 Jun 2021, 15:42
I would prefer a modifier button for scrolling. When held down, you could use the ball to scroll both vertically and horizontally.
This is my preferred method of scrolling as well. In fact, I rarely use separate scrolling controls, mostly for fine adjustments etc.

Re: Trackball "aball"

Posted: 07 Jun 2021, 01:30
by polemon
Archie wrote:
06 Jun 2021, 09:59
12 years in production.
It's not exactly what I meant. The rotation is encoded as a set of scroll-wheel movements (usually in the guise of "buttons". In my case button 4 is up and button 5 is down). The rotation isn't sent to the OS as an actual z-axis rotation. It is interpreted as such. But in any event, it's nice that it has been implemented before.
Findecanor wrote:
06 Jun 2021, 15:42
I would prefer a modifier button for scrolling. When held down, you could use the ball to scroll both vertically and horizontally.

There is a way to make the USB mouse protocol scroll smoothly, i.e. in smaller steps. It was introduced ad hoc by Microsoft ... but should be supported by all major OS:es now. There is no support for it in TMK/QMK, so you'd have to add it.
Archie wrote:
06 Jun 2021, 00:13
Z-axis is quite standard signal for ages, interpreted by most systems and applications as scrolling and zooming.
If you'd use the USB HID code for "Z-axis" then it is likely to be interpreted as horizontal scrolling. That's because some mice with horizontal scrolling do this: Apple "Mighty Mouse" and I think also IBM Scrollpoint, while most other use "AC Pan".
(I wonder what protocol Spacemice use)
Tbh, I didn't think of QMK. I'd be ready to make a dedicated firmware for this trackball.
And if the standard interpretation for the z-axis rotation is horizontal scrolling, that's not even half bad as a first approach! ¦D

A modifier to change the trackball into "scroll mode" is a nice idea, but that would still allow for some sort of use of the z-axis rotation both in cursor mode and scroll mode (I guess you can call that "scroll lock", even). What for to use it - well, no idea at this point. Getting it to register and track correctly is probably gonna be enough of a challenge…

Re: Trackball "aball"

Posted: 17 Dec 2021, 16:16
by kbdfr
Obviously spam, see here:
search.php?author_id=22064&sr=posts