suggested modifications to the 3d print

[itsnoteasy]

I've built this and you've seen the image i linked on twitter. however there's a few issues,

one is that there's no way to move the trackball without clicking the buttons. this could be remedied by raising the outer wall to 1 1/2 times the existing height on the bottom stl. this requires the buttons acute edge to be truncated so that they do not hit the wall. i have filed down the buttons on my build so that i can lift it and it works.

another idea is to allow the use of the internals from the two ball trackball. this would allow high definition scrolling. this can be accomplished by rotating each of the 4 base standoffs 45 degrees, along with the matching holes in the ring stl. the room this makes would allow the second pmw3360 sensor to be placed between the screw holes of either button closest to the usb cable entry point. there would need to be a hole made in the ring stl to accept the lens.

These modifications will work fine with the existing hall sensor, as well as the two ball pcbs, leaving the assembly up to the user.

[frankspace]

@itsnoteasy for those of us who don't have Twitter but are interested in possibly pursuing your proposed modifications, could you please repost those images here? Or at least provide a direct link to wherever you did post them?

[itsnoteasy]

@itsnoteasy for those of us who don't have Twitter but are interested in possibly pursuing your proposed modifications, could you please repost those images here? Or at least provide a direct link to wherever you did post them?

I looked through all of jfedorynski's twitter and my reddit and was unable to locate the images. there were only two, one of the internals and one of the externals.
The images themselves were almost unmodified from the pictures posted in the guide in this repo, however my version was scaled up in size to use a 57.2mm ball.
I gave away the trackball after a friend expressed interest.
I found the ring stiff to use and the hall encoders intermittent. imo it's a flawed design, and you would have better service from the other designs in the repo such as twist to scroll or two ball.
I used his code for the two ball design for a year or so, it worked well, photo here.
https://www.reddit.com/r/Trackballs/comments/12yzdl7/diy_trackball_to_surf_the_net_twice_as_fast/

[frankspace]

@itsnoteasy thank you! Unfortunately, I personally really hate the twist-to-scroll design; I really like the scroll-ring idea on my Kensington Expert Mouse, but I find that particular implementation to be kind of janky. So my options there seem to be limited. Hilariously, I was actually thinking of just designing a PCB and sticking some keyboard keyswitches on it. Anyway, again, thank you!

[itsnoteasy]

@frankspace
another issue i recalled was i pressed down on the ball one day out of frustration and it fractured the pcb. that was an expensive mistake.
I think if the design was to be freely rotatable you would need to use a small bearing, or two like in a fidget spinner or automobile flywheel spigot bearing. you could fix the center of the bearing to the base, and the ball cup to the center also. then fix the scroll ring to the outside of the bearing, like a motorcar wheel hub.
the sensor would be in the ball cup and the wires passed down through the center of the bearing.
Of course the bearing would increase the height of the peripheral, and to be sufficiently strong you would need to make the axle in the bearing center from brass or aluminium, but for a one off custom it would be doable.

[frankspace]

Interesting ideas. There is such a thing as an electrical slip-ring connector (example), if you wanted to pass electrical signals down through something that's rotating. But if I'm understanding this design correctly, a bearing much like what you're describing is already incorporated into it. But anyway, I had been toying with a few ideas for various modifications, and obviously, I'll post something if it actually works.

[itsnoteasy]

the cup wouldn't be rotating so no need for a slip ring. my suggestion was to keep the existing design, but shrink down the bearing size so that it has lower resistance.
since the internal bearing size would be smaller than the ball, it would have to be located below the ball, and the scroll ring would be shaped like a wine glass with a hollow center.
nobody has made one like that before and i think there is the potential for a good performing product.
If you wanted to reduce the cost of making the metal axle, you could use a length of fuel or brake line, but that would be prone to bending.

[frankspace]

Huh. That's an idea. I had actually been considering omitting the cup and bearings entirely in lieu of something a lot like rollers on standoffs, and replacing some of the 3d-printed components with laser-cut metal or carbon fiber and (again) standoffs. But that's strictly at a playing-with-ideas-in-my-head stage, and my ideas sometimes turn out to be bad ones (I'm aiming in my head for something that's easy to clean and maintain, buuuut that might de-optimize for other things). A wine-glass shaped scroll ring is not something that had occurred to me. Thanks!