Thick brass plate in TKL layout (Housing design unveiled - more update)

[Dominus]

Hi people, I want to open a discussion about thick plates. What are your thoughts or experiences with thick plates?
I will be glad if you can join this discussion and share your thoughts.

So, I have a serious vision, shared with some other enthusiasts, about an aluminium keyboard housing with thick plate. Currently, we want to experiment with brass (copper-zinc alloy) as the material and are trying to use a thickness of 3.5 mm. The housing’s design is in progress (still making sure the dimensioning & tolerance and discussing manufacturing constraints) and we will have a discussion about it very soon here.

The reason behind this attempt came from tons of discussions about plate material. According to many enthusiasts, brass is the “thockiest” material. By experimenting with denser materials to achieve better sound and stability, a question emerged: What if we added even more mass to the plate? Adding thickness increases the mass and inertia. Polymer plates are too bendy, more prone to chipping, and has static issue. Composite plates (mostly and specifically carbon fiber) are too loud (high stiffness-to-weight ratio). Last, tray mounts or integrated plates are too rigid, transmitting too much vibration to the housing and eventually to the table. After few years in the hobby now I’m ready for this expensive experiment.

A solid brass slab is expected to be machined for about 28 hours said the CAM engineer. This plate design is about 900 grams when finished.
Well, that number scares me hahaa imagine if we go all-brass. That would be very heavy and too expensive… This https://youtu.be/iq653ZKbLbA?t=263 but TKL…

The plate’s bezel, as seen in the pictures, helps to center the plate. Inside the housing, cushions will keep the plate centered, prevent contact with the housing wall, and help keep dust on top for easier cleaning.

[Dominus]

I forgot to add this info :
Here’s the mod diagram of the plate

If the design requires no rework, the machining might start on May 29th at 6:00 UTC.

[mwb]

I don’t know if your stabilizer holes will work at that thickness. Typically you need to cut away
more of the plate to account for the increasing thickness of the stabs closer to the pcb

[Dominus]

No worries I already made clearance for that, it won’t interfere.

[Dominus]

Well, once I submit the design to the CAM engineer, my mind keeps asking, 'WILL IT INTERFERE? WILL IT INTERFERE??!!"

Stab dimensioning is standardized right? I have checked the dimensioning many times.

[Deadeye]

I really like the idea of the added mass around the edge with isolation mounting… what a gorgeous piece this will be.

Now I’m not sure about brass having anything to do with “thock”, but it’s definitely the most solid feeling thing I’ve used as a plate material. If anything, I think it makes sound trend towards higher-pitched and sharp, where I’d associate the T-word with deeper, more rounded sounds. In my experience with all other things being equal, a flexy POM or polycarbonate plate is more helpful for chasing the rubre-like sound.

Here’s a couple good examples of the brass plate sound I’m familiar with, on some pretty nice keebs:

Project Keyboard Tengu / Brass Plate / Konpeitou Sound Test

AEBoards Constellation with lubed Alpacas (brass plate) Typing Sounds

I’ll be honest in that I’m not too fond of the way my one brass plate build sounds - but I do appreciate it for feeling completely different from all my other keyboards. I have some that weigh more, but none that feel like such an edifice during typing. I worry about the longevity of my flex-cut PCB polycarb plate builds - not this thing. No mild errant physical force will ever disturb its function.

Here are a couple of those flex-cut w/ soft plate configurations that have the deep sound I like, but whose PCBs I worry for in the long run:

YUNZII AL66 | 99.99$ budget aluminium keyboard | sound test

Ikki68 R2, metal topped entry level 65% keyboard with the brass weight

I will say - that one keeb I have with a brass plate is the only one I’ve accidentally ESD’d more than once. Three times, actually… It’s still alive somehow. Brass is beautiful - but also pretty conductive. It won’t make ESD any more likely vs more common materials like aluminum or steel, but it also won’t make it any less likely than with plastics that can carry static charge like standard polycarbonate.

While I don’t anticipate this experiment will produce “thock”, I bet it will produce a striking and unique typing experience with a likewise striking and unique sound. I’m imagining it will be crisp, high pitched, and have rapid decay - with each keystroke having a clear voice in the clacky choir. Interested to see if I’m anywhere near the mark, there.

[ListlessLlama]

There are about a half dozen popular stabilizers in the custom MX keyboard world right now with both PCB and plate mount stabs: Knight, Wuque, TX (at least four variations), Durock, C3 Equalz, Zeal, and Cherry. I have no ideal if they’re the same overall dimensions, but I know there are some slight differences in the housings. May be worth buying a few of these and measuring dimensions before fabrication?

[ListlessLlama]

I can’t seem to find the vid, but keybored (I think it was him?) had an in depth video on YouTube going into the physics of different plate materials a while back. Lots of math, but from a material science standpoint, brass plates should have the deepest sound. In practice, I too find POM or other soft materials to seem deeper.

[Dominus]

o wow… i need to diggest it… i’ll reply to you soon… DON’T GO ANYWHEREEE!

[Dominus]

Exactly, Keybored pointed that out, but @Deadeye has a point… Also, don’t forget that I have added thickness. This will increase stiffness, which we can’t avoid, resulting in a higher pitch. That’s why I don’t just go straight 5 mm in thickness (maximum). I expect a deep-clack tone at least. I hope the heft will overcome the stiffness .

[Dominus]

Thanks for this I’ll check that out. Honestly, I am only familiar with durock, cherry, and TX.

[Surefoot]

Resonant frequency is ω0=√(k/m) where k is Young modulus (stiffness) and m the mass. That’s without dampening - with b as the dampening factor the frequency becomes ω0=√[(k/m)-1/2(b/m)^2].

In simple terms: if you want lower pitch despite a stiffer material, you need more mass and/or more dampening. A dense gasket dampening your brass plate could be the key to lower pitch maybe ?

Also you can work on the sound cavity (below the plate/pcb). Having for example pyramidal shaped “spikes” on the interior weight surface would affect the sound quite a bit, and their size would impact which frequency gets absorbed (or more precisely, diffracted).

[beanns2]

I does the 28 hours include programming and setup time? Because 28 hours of actual machine time seems VERY high.

[Dominus]

thank you for your thorough response, @Deadeye I have similar thoughts on this design, especially when you said that it would produce a rapid decay. The thickness can fill some of the extra space that a 1.5 mm plate can’t, so it will not resonate as much.

I have a similar hypothesis that it can’t produce the pure thock that POM plates do. Adding thickness will change the moment of inertia exponentially, making it stiffer. Imagine you cut the plate and examine its cross-section, here’s the moment of inertia formula:

The “h” variable is the thickness of the plate.

The goal is for this plate to counteract the incoming force that will produce high-frequency vibrations since more force is needed to accelerate a larger mass. It might not differ much if you are a gentle typist, but it will make a huge difference if you are a heavy typist who frequently bottoms out. So it lowers the chance.

Next, about that sound characteristic of those keyboards you sent me, other than the fact that they used normal thickness plates, isn’t it because of the hollowness? There are a lot worse designs than these but anyway I haven’t shared my keyboard design, so here’s a rendering image of the bottom part’s internal:

The internal is not flat. I have added mass to as many free spaces as possible beneath the PCB to help reduce the possibility of disturbing resonance.

Anddd about your ESD experience. How that’s even possible? Is your keeb’s housing made out of polymer?

No worries about this design. I designed my plate to be able to make contact with the PCB that connects to the ESD protection. You just have to make sure about your house’s grounding.

[Dominus]

Well, I was surprised the first time they told me that it would be pure machining cycle time. The duration of the contouring is about 10 hours and the finishing part alone is about 6 hours (two sides), high rpm with SUPER slow feed rate machining. It will be pristine.

[beanns2]

Do you know whay kind of machines they have? As now I’m curious.

[Surefoot]

@Dominus

The internal is not flat.

I think that in addition to the plate itself, this is just as important, as it’s the “sound chamber” of your musical instrument. One line of thought is maximizing the surface area of your brass weight so it can absorb (and modify with its own resonant frequency) the sound waves. That ridged surface looks good, TBH for that insane machining time you could go with having small pyramidal shapes milled into it or something in that spirit (look at anechoic chambers for inspiration ? there are some different patterns you can use, not just pyramids). I’d try and experiment with different sound chamber designs (with spare parts and blue tack maybe lol) in order to hear the difference if there’s any.

[Deadeye]

That’s really interesting about the plate mass! It makes total sense - I’m fascinated to see how much of this factor can be leveraged in-practice by the design.

As for those keyboard recordings - I don’t think I’m enough of a materials / audio expert to tell you exactly why each of them sound the way they do - but I don’t believe either of them are particularly hollow. My own brass-plate build has almost zero internal space and sounds very similar. Wild guess - maybe that high pitched stuff is mostly coming from the switches themselves, where a softer-plate build might be adding its own lower frequencies into the mix while absorbing those high ones? Total speculation.

Internal space and its shape certainly affect sound - I had a GK61 (basic tray mount 60%) whose sound became much deeper when I put it into a plastic case with lots of internal air space. The plastic vibrates with the keystrokes and then all that resonates in that relatively large chamber. By the same token, some of my other deepest-sounding keyboards have almost no space in them at all, they’re just made of really soft stuff in lots of places. For years my deepest-sounding build was a KBD67 Lite with MT3 caps on it; I think the biggest resonance chamber(s) were the underside of those keycaps.

I don’t know what shape is best, but my experience does jive with the notion that not-flat is much better than flat if you want to dampen sharp sounds. IMO, the relatively dainty silicone weight in the original TKC Portico does a better job than the absolute solid chonker from the NK65EE - despite having much less mass, the honeycomb pattern that faces the PCB does its job well. Say that to say I do think the shaping you’ve got going on will help.

As for my ESD experience - now that you mention it, I think all of those happened at work which I do know to have iffy wiring, and I believe they all happened in the dead of winter when the air was bone-dry. It would not surprise me at all to find out the place has inadequate electrical grounding. Add to that poor climate control with rapidly-changing temperatures and humidity levels, a whole bunch of machines moving sheets of paper around, and it seems clear my workplace is a harsh environment for keyboards. But hey! Great for testing sub-ideal conditions, right?

The thrice-zapped keyboard in question is a KBD67 v3 (shown above in a different config), which aside from the brass plate has a solid aluminum top and bottom shell with a brass weight screwed-in. In this configuration, I don’t think the plate actually touches the top shell, but there aren’t any gaskets or dampeners that would otherwise insulate it. The assembly is held in place by a silicone layer between the plate and PCB that extends in places as tabs around the edges, sandwiched in between the aluminum housing layers. From there it has silicone feet between the aluminum shell and the desk, and it is wired to USBC. The PCB design ostensibly has ESD protection with a little pad for connection, but it didn’t seem to work as intended in this case and I seem to remember reading that others had a similar experience with that particular keeb.

I have also used a keyboard with a full polycarbonate case there for quite some time, but haven’t zapped that one yet. (shown above - MelGeek Mojo68; two-part injection-molded polycarbonate case, isolation-mounted plastic plate of some kind and a big silicone internal weight)

[Dominus]

It’s just a not-too-old 3-axis CNC machine with a tool magazine that has limited storage. Some of the tools need to be managed manually it takes a lot of time for tool changing. Also, chamfering the 87 switch cutout and the part with the complex contour for stab clearance is a time-consuming job.

[beanns2]

Ok thats fair haha.