Gadgetoid

gadg-et-oid [gaj-it-oid]

-adjective

1. having the characteristics or form of a gadget;
resembling a mechanical contrivance or device.

Wuque Studio Flux Hall-Effect Switches

Wuque Studio’s Flux hall-effect switches arrived a little too late to find their way into my Slice 75 HE. Thankfully they offered me some – of both the regular, and deep clacky – to try and I’m… in like? I can certainly see why they’re dropping Gateron like a hot rock and shipping their latest editions of Slice 75 (and 68) with Flux switches. I am, however, left wanting for a variant with a slightly stiffer spring and thus higher actuation force- perhaps there’s room for a little experimentation and customisation.

Two bags of keyboard switches sat atop of a desk. A switch from each is placed in front.

Bags of switches, bags of fun!

@gadgetoid/115159532630843439

Nonetheless the WS Flux HE switches include a number of improvements over Gateron and other HE switches I’ve encountered thus far. The most obvious is the octagonal stem.

Octagons are the bestagons

When I first encountered the datasheet documenting this change, and how it reduces stem deflection at 45 degree angles I found myself nodding along- “that’s basic trigonometry,” I proclaimed. Whereupon I fell down a rabbit hole, and I’m dragging you with me- Indeed if you take the gap between the corners of a rectangular switch stem, you can imagine a right triangle with the hypotenuse as the diagonal gap and the opposite and adjacent sides as the other gaps.

Text reads:
According to the size analysis in the above figure: there is no
difference in the moving distance when the octagonal shaft is
pushed at 45° and pushed from the front; when the quadrilateral
shaft is pushed at 45°, the moving distance is 1.4 times that of
the front push; the above shows that the octagonal shaft will
have better shaking, anti-shake and consistency than the
quadrilateral shaft.

Wuque quote the diagonal deflection of a square/quadrilateral switch stem as 1.4x that of the cardinal directions. It's all triangles, baby!

@gadgetoid/115159569792445871

Now let’s take an octagon. With eight sides you’d think this makes things complicated, but you’ve got yourself another right-triangle. This time with a known angle x and a known adjacent length.

Why do we know x? Well the angle of a single triangular slice of our octagon is simply 360° / 8, or 45°. Since our little triangle in question is at the corner of one of these slices, its angle x is 45° / 2 or 22.5°.

Now we must find the hypotenuse given the adjacent and our angle. We know that the cosine of our angle is equal to the adjacent divided by the hypotenuse (since those trigonometric formulas tell us so) so we re-arrange and end up with hypotenuse = gap / cos(22.5°). Or 1.08. So the worst case angular deflection with an octagonal stem is roughly 1.08x the gap between parallel faces of the stem and its aperture.

A bad graphic showing an octagon and a square switch stem with a triangular slice taken out of each, revealing they have a common, generalisable construction we can use to calculate ratios using trigonometry.

We can generalise a rough formula to give us the performance of angular deflection to cardinal as a ratio, calculated from the number of sides. It naturally converges on 1:1, but at a certain point you exceed manufacturing tolerance anyway.

@gadgetoid/115159628240320212

Because a square is basically just regular polygons we can generalise this formula to give us the ratio for the number of sides n, to (given in Python code):

1.0 / math.cos(math.radians(360 / 2 / n))

Let’s try it:

Sides Angular Deflection
3 1.99x
4 1.41x (square)
5 1.23x
6 1.15x
7 1.10x
8 1.08x (octagon)
9 1.06x
10 1.05x

(I calculated the above in Python, but here’s a Wolfram Alpha plot for the curious.

By those last few values you might be noticing a trend- if you were to plot this function you’d notice significantly diminishing returns. Adding more sides to the stem doesn’t get us a better result, and eventually we’d just end up with an approximation of a circle anyway. It is perhaps around 8 sides – the octagon – that the difference falls below manufacturing tolerances.

This matches the claims given in Wuque Studio’s own datasheet, where they – rightly – state that rectangular stems can have 1.4x as much play on the diagonals versus the cardinal directions. With an octagonal stem, Wuque have dialled this in to be near as damn 1:1.

The pure maths gives us ideal values, but in practice the stem aperture has little nubbins around its perimeter that serve a dual purpose: reduce the contact surface for lower friction, and reduce the gap for less all-round deflection.

But what good is all this rudimentary maths and proof if it doesn’t translate to better switch stability? Don’t worry; it does. Both the Flux and Flux Deep Clacky switches have noticeably better stability than the Gateron Jade Pro’s that were supplied with my Slice 75. This applies across the full range of the keystroke, since the stem is housed by the octagonal outer profile. A small difference in switch deflection, once you’ve got a tall enough keycap installed, can result in a considerable difference in key stability… hey, that’s trigonometry again! Sneaky maths. You can also think of it like looking through binoculars- a very small movement is amplified across distance. Flux’s stability, therefore, is paramount for delivering a consistent typing feel from inconsistent key presses.

Does it fix the stability of 1.25, and 1.5u keys such as tab and caps-lock? Sadly not. Is it noticeably better? Yes.

So why not a circle? That would be a pure 1:1 ratio, since the gap around a circle and a concentric circle is uniform. Right? Well circles, unfortunately, do not resist twist or torsion. There’s a good reason we use them for shafts and axles. Switches, on the other hand, well we don’t really want those to spin. While the stems are fixed into rails which guide them across the full stroke, twisting against the rails could result in unwanted friction.

Magnets, how do they work?

The crowning jewel of every hall-effect switch is the tiny magnet in the base of a stem that supplies the magnetic flux measured by the keyboard’s hall-effect sensors.

Wuque are particularly proud of their magnets, and it’s easy to pop apart the switch and see why: they’re huge! The 3.6mm diameter magnets are noticeably larger than those in the Gateron Jade Pro switches. Bigger is better, right?.

A pair of terrible cheap calipers measuring a pair of magnets at 3.6 diameter.

Bigger magnets: confirmed.

@gadgetoid/115159511370855492

Indeed the sheer size gave me an idea- since I had many, many Flux switches to hand and I’d been taking them apart and inspecting them I wondered if I could borrow a few magnets. What for? To hold the top onto Chilkey’s Slice 68, which I’m currently testing but holding back for launch. I harvested eight magnets from flux switches, drilled out the screw holes in the Slice 68 and used epoxy putty to glue the magnets into place. The result? Certainly not enough strength to hold the keyboard together when picked up or inverted, but enough to have the top shell satisfyingly click into place and stay put under normal use. No more pesky screws!

Two hall-effect keyboard switch stems on a desk mat. One with a larger magnet in the base than the other.

Magnet size comparison versus Jade Pro!

@gadgetoid/115159635401670893

But that aside aside, these larger magnets have a practical purpose. Wuque, in particular, are proud of their magnetic flux stability, with a 96 hour live stream airing in China running the Flux switches at a whopping 85 degrees celsius and, apparently, producing a negligible change in magnetic flux, versus competitors dropping up to 12%. What the practical upshot of this is, I don’t know. Less drift over time and need for recalibration? Better repeatability across temperature gradients? (Notably magnetic flux normally drops as temperature increases.) If your keyboard is hitting 85 degrees then you might have other things to worry about.

Prying them apart

Perhaps not so useful (if you’re not obsessively disassembling and experimenting with them), but Flux top housing extends over the bottom with two side clips which are easily released with a fingernail or thin, metal pry tool. These are very easy switches to get open without special tools, versus the Jade Pro switches which require a four point switch opener.

A disassembled mechanical keyboard switch, its parts arrayed out on a dotty deskmat.
Another disassembled mechanical keyboard switch, its parts arrayed out on a dotty deskmat. They comprise a little light pipe, the stem, the top housing, a spring and the bottom housing.

Both the regular flux and the deep clacky have near identical construction, large magnets, large springs.

@gadgetoid/115159523067631742

What’s not so easy is customisation- the larger magnet size in the Flux switches means a larger stem and spring diameter too. Frustratingly the springs in the are not much larger than those in the Jade Pro, with it being difficult to tell by eye. Smaller HE springs fit over the stem itself without much issue. What they don’t fit over is the circular part of the bottom housing where the stem travels up and down.

Regular electrical switches are a different story, their tiny diameter springs don’t stand a chance of fitting around the HE housing. If you want slightly stiffer switches, you’re going to have to wait for a Flux variant.

Both Flux variants also have a separate, removable light-pipe component.

Acoustics

Both the Flux and Flux Deep Clacky switches share a very similar construction, though the stock Deep Clacky have a slightly different colour of stem, so slight that it suggests a material swap versus the stock Flux. I’m told, however, that it’s the bottom shell material that differs.

Two switch bottom housings. One a sort of beige and one pink. The pink one - the “Deep Clacky” - has a viable cross shape in the bottom of its stem housing and plastic so thin that the deskmat colour slightly shows through.

The sound profile difference is all in the bottom housing which is reassuringly visibly different.

@gadgetoid/115159508120472261

Deep Clacky’s bottom shell have a different design with much thinner plastic, reinforced by a cross. The base is so thin that you can see these details through it. This seems to serve as a cushion or damper, cutting out some of the higher frequencies from the switch’s bottom-out sound. It’s not a subtle difference, either, it’s easy to tell by ear which are “deep clacky” and it’s also possible to confirm via audio recording and fast-fourier transform exactly what frequencies are attenuated.

A graph showing two plots of sound pressure over frequency. They are marked

A graph comparing the sound of Wuque Flux switches. Deep Clacky has a redesigned bottom shell and a very noticeably deeper, cleaner sound.

@gadgetoid/115159481874822718

The graph shows the difference, with Deep Clacky having a falloff around 5KHz where the regular switches have a more complicated profile.

Four keyboard switches arranged in a line, the middle two are hybrid swaps of parts from the outer two, though the only visible difference is a slight change in colour of the stem.

Swapping stems let me create two hybrid switches. The sound is all in the bottom housing, though, so the middle two sound effectively identical to their counterparts.

@gadgetoid/115159497042009283

I swapped stems between regular Flux and Deep Clacky to produce a total of four switch variations- two stock and two hybrid. It felt like this gave a range of clackyness depth from the higher, more complicated regular Flux, through the hybrids, to the dampened Deep Clacky. Testing revealed this was probably mostly my imagination, with the bottom shell – not the stem – indeed making the biggest impact in sound profile between switches.

A graph showing two plots of sound pressure over frequency. They are marked

A graph comparing the frequency spectrum of Wuque's Flux switches to Gateron Jade Pro.

@gadgetoid/115164914536970328

Adding Gateron’s Jade Pro to the mix reveals a quieter switch with an emphasis on higher frequencies, more or less confirming what my ear tells me. They’re a much sharper, clickier sound than either Flux switch, perhaps due in part to the smaller base of the stem.

Visit my SoundCloud for a sound comparison between Gateron Jade Pro, WS Flux and WS Flux Deep Clacky (in order, with a little audio blip to separate them).

Overall

Wuque’s Flux switches are refined inside and out and positioned to be their flagship, with both the Slice 75 and Slice 68 adopting them. They feel better and sound better than the Gateron Jade Pro they supplanted. If you’ve got a Slice 75 with the Jade Pro switches then they should be on your roadmap for an upgrade.

Suffice to say Deep Clacky are by far my preferred variant of the Flux switch, they are indeed deeper and clackier.

A mechanical keyboard with most keycaps removed and fully populated with pink switches.

A Slice75 keyboard with pink Flux “Deep Clacky” switches.

@gadgetoid/115159537622911688

What’s not so thrilling is the price- Wuque’s Flux HE switches are sold in either packs of five for $9.90 or thirty-five for $69.90M. At those prices you’re looking at $2 a switch. That’s, uh, between $170 and $210 for a full swap on Slice 75, the former if you make use of the five pack to avoid buying three sets of 35. Maybe I shouldn’t have destroyed eight for their magnets, ha… uh…

If you’re thinking you can buy a whole Slice 75 for that price, complete with flux switches out of the box, then yes, yes you can- you wont get those wonderful Deep Clacky, but you’ll have a spare keeb to pass the bug onto a friend. Nonetheless they are a solid upgrade.

Saturday, September 6th, 2025, Mechanical Keyboards.