Sitting in front of a computer all day isn’t exactly what the firmware between our ears was tuned to do. We’re supposed to be hunting and gathering, not hunting and pecking. So anything that makes the computing experience a little more pleasurable is probably worth the effort, and this premium wireless scroll wheel certainly seems to fit that bill.
If this input device seems familiar, that’s because we featured [Engineer Bo]’s first take on this back at the end of 2024. That version took a lot of work to get right, and while it delivered high-resolution scrolling with a premium look and feel, [Bo] just wasn’t quite satisfied with the results. There were also a few minor quibbles, such as making the power switch a little more user-friendly and optimizing battery life, but the main problem was the one that we admit would have driven us crazy, too: the wobbling scroll wheel.
[Bo]’s first approach to the wobble problem was to fit a larger diameter bearing under the scroll wheel. That worked, but at the expense of eliminating the satisfying fidget-spinner action of the original — not acceptable. Different bearings yielded the same result until [Bo] hit on the perfect solution: a large-diameter ceramic bearing that eliminated the wobble while delivering the tactile flywheel experience.
The larger bearing left more room inside for the redesigned PCB and a lower-profile, machined aluminum wheel. [Bo] also had a polycarbonate wheel made, which looks great as is but would really be cool with internal LEDs — at the cost of battery life, of course. He’s also got plans for a wheel machined from wood, which we’ll eagerly await.
I am interested in a better mouse, but not in a separate scroll wheel, but still, this looks like quite a nice project if you want this kind of thing.
Some Ideas I had while watching the video:
The device can be made thin by reversing the bearing mount. If you look at the cross section @14:01, then mounting the lid (wheel) on the inside of the bearing, and the outside of the bearing to the static housing, you can eliminate the base going from the outside to the inside of the bearing. It should not be too difficult to reduce the PCB size so it fits inside the bearing too. Maybe you need to stack two smaller PCB’s.
A not about the wobblyness: @16:39, the new device bobbles a lot less, but you can also see the base itself is wobbling a bit on the table. Flat surfaces mate very well with each other, but only if they are both flat, and even then, any bit of dirt in between them makes it wobbly again. It’s probably much better to put it on 5 feet. Chairs with 5 legs don’t wobble. This is why nearly all desk chairs on wheels have 5 feet.
For the broken taps…
30 years ago I bought SKF / Dormer taps, and most of them are still good. Recently I wanted some extra, and bought some more from China. Those are such garbage that I threw the whole lot in the bin. Another cause for broken taps is the wrong diameter hole. Cheap taps are the most expensive tool in a workshop.
I think you had the aluminimum lid 3D printed by PCBway (I did not see the whole 18 minutes of the video). It’s quite normal to order it on purpose with too small holes, and then drill them to the right size before tapping. EDM-ing the broken taps out would be nice, but for this, my approach would be to mill out some new pockets next to the broken taps, and then drill some fresh holes.
Just some disagreement about 5 feet:
Everything having more than 3 feet can wobble because then it’s mechanically over-determined. For example milking stools (the floor is often uneven in a stable) have three legs only, for this reason.
Rolling desk chairs have 5 legs/feet for safety reasons. People fell over quite often together with their chair when they only had 4 legs/feet when they leaned back to far. Have a look at an old (rolling) desk chair and you will understand.
In the case of something like this I’d suggest 3 feet will be massively less stable than 5 feet – the overhang of the triangle defined by 3 feet is significant and thus you will will have so much leverage about that line when you are using it that will pick up the other side. 3 feet define a plane which is great when you want your stool to sit on uneven ground, but 5+ feet even though they are bound to be slightly less perfectly defining a single plane can be distributed in a way that largely eliminates that pivot point effect when sat on a “flat” desk – it might only be sat on 3 of them, but as soon you apply force over that void you have another foot that was fractionally too short now in contact with the desk and the total wobble is effectively zero…
Oh also in this case I’d suggest the best option is that ring of fairly soft silicon/rubber or a dish shaped base – in effect infinite feet with a fairly narrow circle contact patch so any imperfections in the surface you are putting it down on likely end up in that deliberate void or bridged over by the ring.
It’s not often that I want to say Thank you for a comment on the Internet, but when I do, it’s for a comment as informative as this one.
I too want to make a custom mouse with a few fancy features but sadly i cannot find a sensor for sale anywhere. I really don’t want to buy a cheap USB mouse just to harvest its no-name sensor and having to reuse the lens in my design.
I doubt I’ll be able to do it anyway, without damaging the sensor
Here ya go:
https://us.store.bambulab.com/products/wireless-mouse-components-kit-002
“We’re supposed to be hunting and gathering”
Ehmm… what doe you mean with supposed? The hunting and gathering concept gave our species a certain advantage over species that didn’t “hunt or gather”. The features of species are not “by design” or “by choice”, it is simply the result of a mutation/error causing a change in behavior that seems to work. Which causes that species to thrive or at best prevent it from extinction.
And regarding that hunting… ebay and related sites allow me to hunt for great stuff, which I gather and store at the place where I live. Which over time becomes so full that I need to make piles and stacks of all my gathered stuff, piles that are prone to collapse any time soon and if not taken care of properly may lead to my demise (or those that visit my place carelessly). So I guess there is a limit to hunting and gathering.
There is only a thin wobbly line between gathering and hoarding.
Modern magic ( previously called “technology”) is also to blame. It now enables people to hoard Tera bytes of data without anyone noticing.
https://en.wikipedia.org/wiki/Hoarding_disorder
Even a peta byte is becoming smallish. with 24TB disks being available, you only need 42, and with 26.1mm height that is a stack of only 1100mm
Touch input device..
Find a VCR harvest the head drum and motor for a very good bearing and mass. There will be some height but thinness is the cause for wobble in a pancake thin bearing.
I must admit I built a spinner from a mouse scroll wheel sensor and a VCR head.
The VCR bearings are amazing, and these days you can use a $3 RP2040 board for the electronics.
Million ways to skin a cat I guess, but if you don’t stop and look at what you are trying to accomplish then the journey is needlessly complicated.
I speak from the perspective that I’d like the end product to ideally be less than $10 so anyone can make one. IE use a large Yo-Yo for the spinner, or 3D print one and use pennies for ballast. Also no custom PCB required, because the sensors don’t need anything but soldering to a easily obtained project board. Be that Arduino, RP2040, Pico, ESP32, etc etc.
No shade, if he wants to learn PCB and mechanical then this is a great vehicle. I personally want to be able to give away 5-10 of them, so if the cost is much reduced, all the better.