Hmm, the TS100 can't pump enough heat into the joints between the two PCBs (at least for the ground plane connections). Gonna have to bring out my old, less fancy but higher thermal mass iron.

I did a quick comparison of the sound of a few different models - the Microsoft Trackball Explorer, Elecom Huge, Ploopy and a quick test part

The Microsoft and Elecom ones have very low "stiction" at low speeds which is really nice. The Ploopy is great at high speeds but slightly frictiony at low speeds depending on direction. Hoping I can mod the design to fit silicon nitride bearings into it!

I designed some quick bearing mounts for the Ploopy trackball today - instead of roller bearings I wanted to use silicon nitride ball bearings. They don't rotate, they're just a low friction surface.

Rather than re-printing the entire massive top part of the trackball, I chose to just make them small "adapter" inserts that you can put into the bearing holders in the existing part.

Anyway, last night I printed the trackball bearing mounts and fitted them into the trackball. They needed a little bit of filing and drilling to get them to fit right, but they work well!

I used the trackball all day today and it's definitely a small improvement over the roller bearings. Unfortunately the snooker ball is too rough so there's about as much friction as before. It's nicer though because it's the same in all directions!

One of the mounts (the one closest to the camera) needed to be reshaped a bit to fit - the pocket is a slightly different shape. But I'm kinda shocked my first attempt works so well!

I ordered an alternative ball for the trackball. It was sold for a Logitech cordless optical trackman and is a bit smaller but still works.

Using it with my silicon nitride bearings with an adapter, it was surprisingly not much of an improvement over the pool ball. However, once I put the original roller bearings back in, it makes a huge difference! There's now very little friction to overcome when going from stopped to moving, which makes precise movements a hell of a lot easier.

So to summarise the current best setup I have for the Ploopy - Use the original bearings and just replace the ball with a glossy one.

I've been talking with someone on reddit about modding ball transfer units into the . They're basically omnidirectional ball bearings.

It was kind of hard to source the small ones in the UK (seem to be available mostly to companies, but I have no VAT number) but I got some eventually.

Here's a quick video comparing them in a test part to the Ploopy's stock bearings: imgur.com/a/Y7NeHqx

How's this for a computer peripheral aesthetic?

This is the first stage of a mod for the Ploopy trackball to use ball transfer units (BTUs) as bearings for the ball.

This just about fits in the shell and feels good, so the next step is to try to modify the CAD file for the original shell.

I'm quite enjoying the ball transfer units in the trackball. They're the lowest friction bearing option I've tried. That means you can flick from one side of a bunch of screens to the other, but it also means it's easier to be precise.

With high static friction, it's hard to overcome it to do a small movement without overdoing the movement.

Yesterday I did a third iteration of the BTU mod for the Ploopy

I cut off just the front section and printed it today - my supports were a bit excessive so just that bit took 7 hours!

I've been using it this evening and it feels good. I moved the bearings a bit further up the walls so it's harder to push the ball sideways by accident.

The trackball print finished in 12.5 hours instead of 13.5, so now I have the world's ugliest trackball!

Also that was a lot of support material - probably could have turned that down a bunch.

My first open source contribution got merged! I added my BTU mod for the Ploopy to the official repo. The creator suggested I do it.

Probably could have done with squashing some of the commits but oh well!

Pull request link: github.com/ploopyco/trackball/

I've been writing a blog post about my various iterations getting the Ploopy to a point where it's just about perfect for me. It's getting very long - I need to learn to be more concise.

As a designer I should probably be more aware about the information the reader actually cares about!

Just spent most of my day off work finishing up the blog post on my various experiments with the Ploopy trackball.

I've decided it's good enough to upload, so here it is:
gbryant.co.uk/posts/2021-02-15

Show newer

@gbrnt what are you're thoughts on the ploppy? How does it compare to the elecom (I swaped the bearing in mine)

@strikersan The ploopy is very usable and feels OK, but not as smooth as the elecom. The roller bearings mean there's a bit of friction at low speeds and that makes it hard to be precise. At higher speeds (e.g. moving from one monitor to another) the friction is very low.

I'm going to try modding the silicon nitride bearings into the Ploopy to see how that feels. I think the ploopy ball isn't quite smooth enough for that, but I'm going to test it anyway.

@gbrnt I repaired an industrial trackball once that used a tiny ball bearing assembly on a shaft as the idler spot. The bearing assembly was broken, tiny balls missing, just two ring dangling on a shaft. Got the bearing at a bearing place, Amazon did not exist yet.

@AskChip I'm having trouble visualising this. Why was the bearing assembly on a shaft?

@gbrnt The ball bearing assembly, that part that contains balls is an assembly. They put a shaft through the hole, mounted it, and the outside of the ball bearing assembly rode the ball as the idler wheel.

@gbrnt it looks, on my phone, like you've used masking tape to secure a few pieces of sandwich meat to the mouse

@gbrnt
I'm impressed that you could remove the support material in one main piece. What slicer do you use?

Did you manually switch filament colours, or is it the same filament roll?

@normandc It was pretty well stuck but it did come off with some levering with a screwdriver! I'm using Cura, with "enable support interface". I think there's a setting to break the support into chunks, which would have been very useful.

The filament is a rainbow one, but I'm not really a fan. You need a big print like this to see more than one colour, but then it changes too quickly.

@gbrnt better than mine! My first contribution was correcting a link in documentation that someone else pointed out, and another person commented exactly how to fix it. It was a hollow win, but I took it. 😁

Show newer

@gbrnt
Just curious, what CAD software are you using? Is that Rhino or Moment of Inspiration?

Trying to guess a CAD program from screenshots is a game of mine. 😄

@gbrnt
I tried MoI back when I was still on Windows. Very nice little program, not as complete as Rhino, and cheap.

Sign in to participate in the conversation
Mastodon for Tech Folks

This Mastodon instance is for people interested in technology. Discussions aren't limited to technology, because tech folks shouldn't be limited to technology either!