File: keyboardnumpad2.jpg (22 KB, 400x263)

22 KB JPG

first for numpads are for trannies

File: 5306570_0_i1064.jpg (56 KB, 700x326)

56 KB JPG

>>107297682
>trying to fit in so hard that he forgot the numpad existed before the separate navcluster

I've been using my kinesis advantage 2 LF for 3 years now and I'm quite happy. One for home, one at work.

>>107297886
that looks like special needs

>>107297886
exactly, it's deprecated tech that has no place on the keyboard anymore

>>107297969
the navcluster and numpad are combined into one so you get the functionality of both with the numlock key, which is how it should be
>>107297970
retard

Desoldering the Monoface with a desoldering gun was easier for sure, but still more work than I'd like. Although a slight learning curve with the soldering gun had something to do with that.

Discovered that contrary to my assumption the pinging had nothing to do with the alu plate and everything to do with the springs in my mx browns. Surprise, surprise. They were spring-swapped to lubed TX springs but I guess that wasn't good enough. (Oddly enough, it was good enough in one of my other boards. So I don't know what I did differently.)

I only realized this because I was taking apart a hotswap TKL build with stock Cherry Reds in it (for break-in) and I noticed it had the exact same pinging around the Y key, but not when assembled with different switches. Something about the placement of the Y key causes every spring in a wide radius to ping in unison making an ultra-loud pinging noise.

So I don't know what I'm gonna do to make these faggots stop pinging (although I wanted to spring swap them anyway because 70g was too heavy for me), all I know for certain is I'm not fucking soldering it this time until I've thoroughly tested it and know it sounds just right. The back of this PCB is starting to look like shit

>>107297995
>the navcluster and numpad are combined into one so you get the functionality of both with the numlock key, which is how it should be
by the same logic you can combine all that with alphanumeric keys on a 75%. so your argument is pretty much retarded

I only have some MoS2 lube from my motorcycle engine rebuild. will that work in switches?

File: 1697729637028583.png (131 KB, 290x418)

131 KB PNG

>>107297995
>retard

>>107298020
You could, but it wouldn't be as comfortable an arrangement.

File: 1743531075448957.png (51 KB, 480x639)

51 KB PNG

How does this mindbreak a troon so badly that he has to ritualpost every thread now? Not ever using a staggered, worse layout with gayers, either, since I don't larp.

Scissor switch mogs Mechanical Keyboard. Mechanical Keyboard is one of the most overrated concepts in some custom hardware history holy shit. Whenever I type on scissor switch kind of keyboard, it is much faster than on mechanical keyboards.

I spent over 8 years getting used to a mechanical keyboard, but after it broke, I switched to a scissor switch keyboard about 6 months ago.

>>107298135
you probably mean butterfly switch and call it scissor switch because you don't want to acknowledge how gay you are

thoughts on romer-g switches?

the only switch worthwhile discussing is the analog one. the rest do not exist as far as I'm concerned

>>107298148
They're okay, basically just MX Browns, but with a slightly shorter travel distance.
>>107298226
>t. fell for hypebeast retardation and shill snake oil

>>107298226
I feel the same, they are our only hope of breaking out of MX hell

>>107298427
I get it that you invested a lot of money in those primitive switches, which are now basically obsolete tech.
not my problem
cope harder

>>107297628
>1.25u left mods
>1.5u right mods
awful

>>107298226
usecase???

>>107298997
any usecase I mention you'll downplay in a way that makes your obsolete tech perfect for your needs without needing anything more or less.
that falls under "cope", see >>107298488

File: file.png (97 KB, 350x179)

97 KB PNG

>>107297682
60% keyboards are troon coded

>>107299025
i see

no usecase

>>107298990
And 1.5/1/1.5u would have taken the same amount of space in the bottom row. Would like to know why they couldn't do that.

It can be built 1.5u 1.5u 7u 1.5u 1.5u, but I think that looks a bit unbalanced imo, I don't like the spacebar that far to the left.

>>107299171
My bad I meant to say 1.5/1/1.5 with a 6u spacebar. (Or my math was just wrong and I'm retarded, you decide...)

things that might make me start caring about he/tmr:
the firmware not being shit anymore because it still hasn’t caught up
mx hybrid boards with good enough firmware to mix both types of switches
some sort of universal frankenswitch mod to convert a mx switch to an analog hybrid with magnets
anolog switches having even 1/10th of the variety of mx instead of just linears + 1 (ONE) tactile

the most important part of the switch is how it actually feels to use, otherwise why not just use a $10 dell rubber dome or some chiclet piece of shit.

Rec me a TKL keyboard or whatever the cool kids are using these days that has ins, del, end, home keys, pgup, pgdown. I just need something functional, maybe blank keycaps would be cool (think das keyboard). No led bs. My filco ninja broke due to a water spill (is it still repairable?, pressing the number row keys double presses or press the keys adjacent to it)

>>107299212
neo80 cu?

File: smoothmaker.jpg (842 KB, 2140x1901)

842 KB JPG

Prototype Maglev switch for tmr sensor made! After my 200 IQ brain thought about it, the obvious answer was to superglue magnets onto the bottom of a normal HE switch but so the magnet is away ways. Then I took the spring out so the magnet is acting like the spring

Its SMOOTH

Like ultra smooth, butter doesn't quite explain it. Its smoother than anything I've ever tried
But the bottom out is weird, it gets like exponentially stronger the closer to bottom our you get. It takes like half a 150 grams of force to completely bottom out which is insane, it doesn't feel good to bottom out I will be honest

But damn, millimeters 4 through 0.5 are awesome

>>107299171
>>107299174
yeah, 1.5/1/1.5/6/1.5/1.5 would have been the way. hell, even just having matching 1.25u mods on the right and leaving an uneven 1u exploded gap would have been better imo. i didn’t realize that was the brutal v2. it’s a nice looking board otherwise but that would drive my ocd insane.

>>107299212
Smoke ENTR
https://drop.com/buy/drop-expression-series-keysterine-keyboard
WoB pbt keycaps
https://drop.com/buy/artifact-shield-white-on-black-keycap-set
$45usd + shipping

>>107299264
do permanent magnets weaken if you force matching poles together too often?

ok it seems the Keychron HE switches are KS-37B variants. B is for double-rail versions.
There's two standards, they don't work interchangeably. Keychron (KS-37B versions) have north pole towards sensor.
these should be the original switches that Keychron recolored for their versions:
https://meletrix.com/products/gateron-ks-37b-magnetic-switch

Some various versions that are compatible or not:
https://www.hlplanet.com/gateron-magnetic-switches/
and this dude does...something:
https://youtu.be/U7-xZcJVyjU

>>107299196
qmk works fine but via didn't yet do shit for analog stuff. keychron's interface does offer all the functions.
classical switches...meh. complicates design imo. I only have a set of box switches so I don't feel the need for regular switch compatibility, don't really care.
>some sort of universal frankenswitch mod to convert a mx switch to an analog hybrid with magnets
you can replace the stem, if compatible (like dual rails shit)
>anolog switches having even 1/10th of the variety of mx instead of just linears + 1 (ONE) tactile
you need to buy them so they get diverse. nobody buying them means no investment money
>the most important part of the switch is how it actually feels to use
linear, whatever. they can make them tactile but kinda defeats the purpose, as you cannot set tactile point, but you can set actuation point. so it's kinda weird
I just want dead silent magnetic switches, that's what I really want atm, rest is just noise

>>107299272
yes, but with magnets of this size, only after doing it repeatedly for 100 years

>>107299264
yeah I'm the one who tried it with a 2mm cube magnet, but weaker magnet in stem (keychron ones) and the feeling is at least on par with springs. problem is the required magnet is too large, no room for HE sensor under the switch, needs large magnet in stem, and needs TMR sensor inside the switch
but the maglev shit is interesting ngl, I kinda dig the idea

>>107299272
nah. only heat fucks them up

>>107299212
Get another filco and don't spill water this time

>>107299314
>2mm
*2cm

>>107299298
>as you cannot set tactile point, but you can set actuation point
it wouldn’t make sense for just setting the actuation point but being able to set a non-linear analog curve to match the bump would be good for any tactile switch with pretravel

File: he.png (392 KB, 1017x663)

392 KB PNG

>>107299264
bro that's working because that stem magnet if fucking huge, it's really really huge. that switch does not work with my Lemokey P1 HE, guaranteed. that's some other kind of format/standard.
picrel is what Keychron uses, it's too small to get any kind of effect from the size of magnet you are using on the bottom of the switch.
so gotta use as large magnet as possible in switch, literally build the whole stem from magnet, so you need the smallest one on the other side of the PCB, which in turn fucks with TMR sensitivity as the sensor needs to be inside the switch. tho it might work

>>107299334
>but being able to set a non-linear analog curve
ah that's doable. current keychron qmk fork already has a setting for curve points, for controller joystick emulation use.

>>107299378
>that's doable.
not without more tactile he/tmr switches or mx conversion mods

>>107299264
or imagine having coils inside the switch, when you turn on the PC they energize and produce a magnetic field. so when you turn your PC on all your keys would rise at once, when you turn off your PC all the keys would fall.
you could also make your keyboard type by itself. that would be kino af, but would probably draw a lot of power.

>>107299264
>But the bottom out is weird
because you're sticking them right up to the switch. if it was on the other side of the PCB you wouldn't have that problem as they wouldn't go as close.

File: Screenshot_20251123_130724.png (1.21 MB, 3628x1881)

1.21 MB PNG

>>107299231
This looks clean, but I have no idea what to pick or any of the words mean.

>>107298488
>>107299025
I'm not the one coping from sunk cost + FOMO pangs. It's okay to admit that anal-og has no real world benefits for anything that matters, and certainly not its most hilariously hyped up gayming "benefits".
>b-b-but muh contactless muh theoretical longevity
Tests?

>>107299517
it’s a custom board, as in, you get to pick the materials you want. you need to buy your own switches, keycaps and sometimes stabs too though

it might not be for you, but that is how you get the highest quality keyboard. if you want something cheaper, not made out of metal, maybe try a kbdfans tiger lite

File: 1753878190701949.gif (1.12 MB, 250x198)

1.12 MB GIF

I'm using a 3yo Keychron K2 relying on Bluetooth 90% of the time but over the past year it has driving me insane with dogshit battery (lasts like 5 minutes after being plugged in for a week) and connection (keys getting stuck). Its an old model so it doesn't come with a wireless dongle which I imagine would address my issues.

What's a good wireless keyboard that lasts forever on battery?

>>107299518
I don't give a fuck about longevity lmao that has zero relevance in today's world, nobody gives a flying fuck about longevity. not sure how this is a relevant metric anymore with such huge supply and low prices, nobody gives a flying fuck these are not switches you take to your Mars base and they gotta work for 300 years wtf
for usecases I started playing with moving cursor with keys, works fine but need to tweak response
games are nice as well, tested walk/running and it works nice. customizing the response curve is interesting, you can have various use cases.
but the cope is strong in you, so you will simply discard anything that is a clear benefit and focus on exactly what those shit switches give you, a single fucking actuation point, and that's fucking it. and through sheer luck and magic that is the best and only usecase for you, some-fucking-how lol, such that anything else but what you specififcally have is somehow irrelevant and useless

>>107299583
>What's a good wireless keyboard that lasts forever on battery?
literally anything that runs zmk. even if it has a hummingbird battery it doesn’t matter. my wireless agar has two little shitty batteries that mount right under the spacebar and the battery just lasts for months

>>107299583
airv3 I think has very long battery life like 1200 hours with rgb off, like 4 months of daily use for 10 hours a day on a single charger. supposedly

>>107299517
>top case
for looks. Pick what you like

>bottom case
tiny difference in weight and acoustic properties, not really noticeable, pick for looks (brass is cheaper)

>weight
same as above

>pcb
solder vs hotswap and wired vs tri-mode (wireless). Wired solder is cheapest, hotswap makes it easy like legos to stick switches in (no soldering and desoldering). Tri-mode hotswap isn't as flexible for layout so you have to pick whether you want ANSI or ISO.

>plate
the aligning plate that holds the switches in place. Necessary with hotswap (don't choose plateless). Plateless with solder PCB is more flexible and might make for a more flexy typing feel. Material of plate affects its sound and how stiff it feels. Alu is pretty stiff, but has a nice deep sound. Polypropylene is pretty popular right now for its sound. Carbon fiber also stiff, brighter sound than alu, on some boards this helps bring out some 'clack' sound. PEI I don't know about, some new material that they compare to PC, which flexes better than aluminum and sounds a bit different. Glossy-chamfer sounds like a decorative thing.

I'd get brass weight, brass bottom, whatever color top you like, a hotswap wired PCB, and I'd spend a little extra to get a second plate so you have two options that are reasonably different, e.g. aluminum and PEI.

>>107299611
you forgot to mention wk/wkl means has win keys vs has blockers where win keys normally go

Dumb question- are the keycaps on the keychron k10 replaceable? I think the Rohan keycaps would look dope on it

>>107299636
my bad didn't see that option

>>107299588
>for usecases it's just autistic shit (the bad kind) and gayming actions done 10000000000000000000000000000000000x better by other input devices
>n-no u
Those afflicted by sunk cost + FOMO pangs are a miserable bunch, aren't they?

>>107299560
>>107299611
So i'm paying 200 bucks for a metal frame? Where do i pick the switches and caps

>>107299517
definitely don’t get WKL unless you like the idea of not having windows keys
get whatever color you want
get brass because its cheaper
get hotswap wired if you want wired, trimode if you want wireless
alu plate, super duper simple

if after you factor in switches and keycaps the price all makes sense to you and you’re willing to put it all together then go for it. if you just want something complete out of the box and are at all hesitant on any of this then just don’t do it. if you want to see what building one of these things looks like, go look some alexotos videos. here’s one for the neo80 in particular

https://www.youtube.com/watch?v=utf3wmYr-qo

File: disassembly-8.jpg (207 KB, 1800x900)

207 KB JPG

if one were to do something such as drawing up a whole new PCB for their shit,
is there some resource with datasheets for what controllers/stabilizers/hotswap socks one should get?

>>107299641
Keycaps on most switches are able to be changed out for whatever set you want. The only thing you usually have to watch out for is making sure the spacebar(s) in the kit you're getting fit your board, but the K10 is a typical full size, so it should have a 6.25-7u spacebar.
Honestly, the only switches that don't really allow this are ones with proprietary stems/sliders, like some of Razer and Logitech's stuff (some of their optical and Romer switches), some older vintage keyboards and most laptop switches and cheap office low profile keyboards/chiclet keyboards.
Just look up a quick video about how to properly take keycaps off and put them on, so you not only don't fuck up your keycaps, but also make it easier on yourself physically. A word of caution if you keep going into this hobby over the years - if you ever try some older Kailh Box switches, or Cherry's Clears (both the original and Ergo Clears) and MX Purples (basically recolored Ergo Clears) they will almost certainly crack your keycap's mount, the cross shaped hole stem, if they are made out of ABS.

>>107299679
You buy the switches and caps from whoever is selling the switches and caps you want.

Good keycaps will be anything from ~$30 (dye sub) to ~$150

Switches should be $0.25-0.50 each, for a TKL you'll need ~90 of them so that could be $23-45

For me, it's ano taupe with DCS intro beige
https://bowlkeyboards.com/collections/keycaps/products/instock-extras-dcs-intro-beige-and-white-on-black
and some Gateron L'orange from the same vendor:
https://bowlkeyboards.com/collections/switches/products/gateron-lorange-switches

That's how I'd build it. Probably with the PEI plate.

File: xaxis.jpg (338 KB, 1518x1214)

338 KB JPG

>>107299358
The idea right now is to put the sensor in the LED hole. It gives enough room in my tests for the sensor to breathe. All I have to do is snip off the LED diffuser so its simple.

Also I'm using an X-axis TMR sensor for these tests. The Z-axis doesn't make sense for sensors inside the switch but I will test those also. The x axis is also just easier to shield

Im also going to test some wheatstone bridge tmr sensors, I think these guys will be the best way possible to have an in switch sensor

>>107299679
basically that, the pcb, and a carrying case. you are paying for something where every detail has been poured over, from the ano, mirror finish weights, multiple mounting styles from top to poron or silicon gasket, to smaller things like the internals being stepped for more appealing sound when it deflects from the brass. it’s autism.

you can get keycaps anywhere. you can get cheap shit off amazon or gmk clones off aliexpress. kbdfans has good keycaps. i usually use drop just because my favorite profile is mt3 but you probably just want standard cherry profile

i usually just get my switches off milktooth

>>107299703
I've never seen a full, comprehensive reference sheet. I just looked at several build guides and reddit also has some resources in the wiki in ergomechboards. These were splits but the logic is basically the same for full size. You can look at a few github repos for popular boards, the usually list what parts you need to get. I guess one thing splits usually don't have are stabilizers.

>>107299767
>poured over
Pored over. And it's sickly-sweet marketing talk to say they pored over every detail of it. Even if it's in some sense true, it sounds like bullshitting, and doesn't really communicate why there's any value here.

>>107299835
buy another filco if you feel that way. it's just a money wasting hobby, you don't need a custom board

>>107299702
>>107299767
I saw the video and it seemed like a quite a bit of work, would you recommend buying a pre built one? Which it seems that qwerty/milktooth/drop all offers pre built, but which brand is a good one?

>>107299873
I have no problem with spending ~$500 on a very nice keyboard. I have a problem with retards on the internet writing fourth-rate copy trying to sound like a Golf Channel ad

>>107299903
https://www.youtube.com/watch?v=UXYdPQSCaxU
This is the entire process from start to finish. He starts around 30 minutes in and finishes around the 2 hour mark. Even as a first timer it's something you can do in a single evening, easily. The only part that can be a little off-putting for a newcomer is maybe lubing and installing the stabilizers. And really the only hard part of that is the additonal spend to buy lube and something to apply lube with. And if you skip lubing entirely the keyboard will still work, you'll just hear the stabilizers.

>>107298990
U mad bro?

>>107299903
I think if you're buying a custom board you should build it yourself just for the experience. The process is just enjoyable, like, I wish I could build them more often. I don't know anything about pre-built customs other than alexotos, the guy I linked, takes commissions to do builds for people on stream and I think he charges like 50 bucks? Never tried any "ready to use" boards.

File: IMG_5378.jpg (76 KB, 1200x1200)

76 KB JPG

i always thought these were cool
there when you need it but not attached to the main keyboard when you dont need it

>>107299950
Then buy a very nice keyboard, instead of cribbing about paying 200 bucks for a metal frame.

>>107299979
Hey retard, have you considered that maybe you're replying to different people?

>>107299982
You are pixels on a screen in my basement. I do not care.

>>107299703
depends. first you settle on the layout. use https://www.keyboard-layout-editor.com/ to make it, or clone the one you need to replace the PCB for. export json for it.
then use something like https://github.com/jeroen94704/klepcbgen
and use said json to generate the Kicad project which sets all the switches for you. delete the stock micro and use the one you want to use. like check QMK compatible ones, choose one that makes sense, add it in the Kicad project, and start connecting shit up in the schematic.
once that's done, and you edited the switches to add th eversion that you want, with sockets or not, with LEDs etc, then start measuring the fuck out of the original PCB. where USB goes, where holes are, edge shape, use calipers and quadruple check every single fucking detail for shape, else it won't fit.
also check spacing for each switch so it fits the original plate from the keyboard you're making it for.
micro + diodes should be fairly standard. other shit like LED chips you can start looking at how other keyboards did it, use SPI chip for LED matrix, how to connect it, use the same QMK code that others used so you're not wasting time with developing existing shit like a moron

>>107299985
>>107299950
just go buy a Lego set

>>107299982
Same applies to you. >>107299979 isn't even >>107299873 (me) btw

>>107299673
you must be confused, I keep tinkering with it for two weeks and can't seem to stop lmao I must have reflashed it over 50 times while developing the 10bit mod + HID control file
you seem the kind of moron that uses what people like me make, the proceed to give me lip about some retarded shit

>>107299744
ah you are using the TMR sensor, yeah that's should be an ok offset.
whetstone bridge sensors should be the best for this, and they should also be the cheapest

>>107299744
damn, the diffuser really would be the place to mount one of those things. what about encasing it in some PLA that just slots in with it at a 90 degree angle?

File: sot23.png (655 KB, 793x941)

655 KB PNG

>>107299744
a sot23 package won't properly fit on the bottom of the switch, used an actual magnetic switch here, the Gateron Nebula one. though the switch could fit on a small PCB with holes for pins that would poke through the case into the PCB, but doesn't seem like it would work.
so the sot23 package could maybe work vertically, on a pcb. so X/Y axis still. and could need thinner PCB, like 1mm max. and package would be right up next to the stem column

File: alu-kbd.jpg (224 KB, 1600x1600)

224 KB JPG

nah… none of your toy boards have nearly enough function keys for my taste

File: sot23_2.png (263 KB, 588x523)

263 KB PNG

>>107300095
vertically it would fit but only on flexpcb, this is inside the top part of the case, inner plastic goes on the outside of the stem tube, so there's space for the sensor to sit vertically, but would have to be on thin flex pcb, and that flex pcb would have to connect to electric pins poking on the bottom of the switch.

>>107300059
ideally they'd be in switches, and somehow removable, so you can swap them out. this way you don't throw the PCB away when one breaks, and you also don't waste money on buying switches with them included, which would raise the price per switch. having them as a removable addon into the switch seems like the best option for serviceability and PCB design

are ducky keyboards any good?

File: preload.png (413 KB, 761x597)

413 KB PNG

so you could have three pins that poke into the case, and have extensions to the wall of the switch with some preload, and you slot the flex pcb vertically between them and the plastic wall of the switch case. and the flex pcb has exposed pads in those areas, for the three connections. and it just sits there. and you can slide it out and move it into another switch. or replace it when it breaks. no soldering, switches can have three pin sockets into PCB.

>>107300242
software is shit is what I've heard

File: flexpcb.png (65 KB, 697x385)

65 KB PNG

>>107300264
something like this, but on flex pcb. I don't know how to do flex pcb in kicad. imagine that's flex pcb. and maybe add some dielectric lube. this makes sense to me

>>107299712
Thanks. I currently have some Logitech G213 from some time ago but i think its time to upgrade.

Im not settled on the K10 though, i just liked the wood on the sides.

>>107300003
Nope, I'm 100% certain from the way I wrote about it, since I know it to be true. It's self evident. This weird projection out of nowhere seems to be partly caused by your failing in reading comprehension, what with all that ESL babbling, too.
I hope you can get therapy or something, most people usually move on from lackluster purchases instead of coping and raging like this, though.

>>107299673
"I regret buying a wooting HE80" said no one ever

>>107298020
>by the same logic you can combine all that with alphanumeric keys on a 75
and it would be stupidly inconvenient to use, so your argument is pretty much retarded

Which brands does /mkg/ recommend?
Is keychron no good?

>>107299964
issue with those is, they're never in the same place.
sure they're usuable. but not blindly, or less so.

>>107305278
I think Keychron's plastic boards are good I just wouldn't recommend their aluminum boards

Why does Keychron need my phone number when I want to buy something from their shop?
I already get enough scam calls, feels like a red flag when they already have my email.

Which keyboard will help me detransition?

>>107306134
100%

>>107306134
a cheap membrane

>>107306157
I already own a 100% tho

File: 8 bux.jpg (75 KB, 1500x615)

75 KB JPG

>>107306591
Take the next step >>107306570

>>107306071
what're you gettin', anon

>>107305508
>I just wouldn't recommend their aluminum boards
why's that

>>107306884
The K10 HE Special Edition
Been looking for a 100% one in ISO layout and this is pretty much the only one I liked

Do people still buy Leopolds?

>>107307014
been thinking about going over that he cliff myself, but i'm just not sure what i'd use it for. i'd just really like a better build keyboard / upgrade from the ak820 and mise well go he i guess

>>107302608
what is self evident is your coping. nothing will ever change that all of the switches that you ever purchased are now suddenly obsolete technology. embrace the future old man

>>107307014
I was pretty close to getting that exact one myself, but I went with 75% in the end. I think K HE series is the only one with north LEDs, but you can rotate the switches if you don't care about light and need keycap compatibility

>>107307224
>is the only one with north LEDs
I meant the only other HE series from them, with north LEDs

>>107306890
Not competitive with other ones at that price

>>107307203
>the future is using analog switches to emulate binary inputs
he/tmr will never have anything beyond a niche usecase in video games. it’s like the gyro of keyboards

>>107297628
Okay I got the keychron q1 he. I like it and all but recommend me a keyboard that is better if I am looking for any upgrade. Previously I had a k100 from Corsair.

File: lemo.png (174 KB, 731x624)

174 KB PNG

>>107307291
>at that price
doubt you can do better for full aluminium case with hall effect. and QMK

>>107307327
>adjustable actuation point is le bad
why would you cuck yourself out of better technology?

>>107307344
because I have never once thought in my life “oh if only this switch’s acuation point was just a little bit higher” and he switches cuck you out of variety

Do you guys keep all your keyboards in a closet in the original boxes when not in use?

>>107307361
I did, and other usecases like controlling the cursors from keys when lazy af to reach for mouse. or all kinds of actions in vibeo gaems
>and he switches cuck you out of variety
now you're showing everyone how economically clueless you actually are, which figures since you think obsolete tech is worth the money

>>107307377
>controlling the cursors from keys
I would rather have a trackpoint

>economically clueless
you’re going to have to elaborate

>>107307328
To get better than this, you're beginning to get into custom keyboard territory. A good starting point is this post >>107299517

tho if you want a better HE keyboard, the only option is Wooting HE80. It kind of sucks you didn't buy that one first.

>>107307390
>I would rather have a trackpoint
that's extremely obscure tech
>you’re going to have to elaborate
you cannot have diversity if you're not buying it and provide feedback on what you'd like better. plus a set of switches costs similarly as older ones
what won't make for diversity is you yelling "don't buy it" after you've already spent a fuckload on shit you keep in some boxes that you never touch anyway

File: NynBehind.jpg (44 KB, 560x511)

44 KB JPG

>>107307375
I have a wooden chest filled with loose keyboards in the garage....

>>107307408
>Wooting HE80
nice specs but doubt you can really tell the difference between 1kHz and 8kHz polling. no QMK and also double the price of a Q1 HE if you're going for metal case. it's silly expensive.
and I think they're using the south-pole magnet version of switches, which most companies already moved away from. most are using north pole magnets facing the sensor nowadays

>>107307424
>obscure
and yet better at controlling a mouse from homerow than “cutting edge tech”

>you cannot have diversity if you're not buying it
that’s not my problem. its like lecturing me that there are no VR games because I won’t invest in a VR headset, VR would be so much better if I just wasted a bunch of money on something that doesn’t have anything I want to play. Life doesn’t work like that, sorry. I don’t have a usecase for HE switches and I buy switches for the way they feel, not for technology I don’t care about or some vague non-promise that in the future I will get what I actually want if I buy something I don’t.

>>107307471
it doesn't have to be better, it has to be good enough so you don't need some obscure tech instead
>that’s not my problem.
yes, you are clueless. plus a lot of people are already getting in on the magnetic train. get left behind gramps

>>107307478
>get left behind gramps
how does controlling your mouse with your keyboard work?

>>107307493
good enough but need to tweak the response curve
anon, let's get real, once your authority figures signal you should like hall effect you'll start clapping for it like a retarded seal. it's not like you have any agency on the matter

>>107307450
>doubt you can really tell the difference between 1kHz and 8kHz polling.
I can, and so can everyone who uses one, and it's very easy. I'm not sure why people think this is the case. You notice how much faster 8khz is in the first few movements

the real purpose of the wooting is it's software anyways. You don't buy a wooting to use KIA, you buy a wooting because it has the best software out of every HE keyboard in the market right now. It's actually one of the weaknesses of HE imo. It requires refined software to make it work well.

>>107307478
>it has to be good enough so you don't need some obscure tech instead
nigger, magnetic switches are obscure tech, and it does have to be better. that’s the whole point, you use the thing that’s best for your needs. I haven’t tried using he to emulate mouse input but I can already tell just from a glance, it’s not easier than a nub that represents a single vector by itself.

also I’m not the one screaming don’t buy it, I’m telling you I have no usecase for it while you’re screaming mx is obsolete don’t buy it, so cool projection

>plus a lot of people are already getting in on the magnetic train
who cares? a lot of people are cutting off their dicks too

>>107307524
8kHz polling doesn't mean 0.125ms between press and action on your computer, there's more losses into the whole chain between scan rate and action on screen. there's a lot of diminishing returns past 1kHz.
competitive players might be able to tell the difference, but that won't be an advantage that you perceive as a regular keyboard aficionado. it's not worth dumping the money for that.
as far as software goes, Keychron's QMK fork can set all that analog jazz, or their online web based interface. doubt Wooting's can do anything extra as far as analog functions go. if you do have an example please share it with everyone.
also just as a comparison, a P1 HE is $135 see >>107307332 and a Wooting HE80 is $375, both full metal, hall effect, full analog features, Wooting is 8kHz matrix scan rate (not 0.125ms between key press and action on display). if you think that extra 1ms saved is worth extra $240 for your average keyboard enthusiast you're reaching hard

>>107307537
>hardcore coping

>>107307566
>8kHz polling doesn't mean 0.125ms between press and action on your computer, there's more losses into the whole chain between scan rate and action on screen. there's a lot of diminishing returns past 1kHz.
the same thing is true about 1kHz tho. Most 1kHz is closer to 2 ms. While 8Khz is closer to .400ms.

>>107307584
so far your arguments are
>you’re coping
>its current year
>appeal to popularity
and
>you don’t understand economics because you won’t buy something you don’t want
I am not swayed

>>107307595
>Most 1kHz is closer to 2 ms.
see 107199429
it's 1.17ms average. so 1.17ms to 0.4ms is less than 1ms gains. I mean...if you think that's worth it for you sure, spend the money. I'm not bashing Wooting, it's a nice keyboard. it's just very expensive
btw, my 10bit mod shits on Wooting's 0.01mm steps. mine does 0.004mm steps

>>107307674
>see 107199429
>>107199429

>>107307666
don't get it anon.

btw I think I forgot to post the whole thing
https://github.com/gibs2193/qmk_firmware/tree/10bit_mod
mods are in keyboards/lemokey in common and p1_he
the hid file:
https://github.com/gibs2193/qmk_firmware/blob/10bit_mod/keyboards/lemokey/common/analog_hid.c
and python script:
https://github.com/gibs2193/qmk_firmware/blob/10bit_mod/keyboards/lemokey/p1_he/lmk.py
so you can set all analog functions from command line via HID interface, no Keychron web based app needed
you'd have to adapt the Keychron files for other Keychron HE keyboards. I only did it for my P1 HE

>>107297628
My Keychron C2 Pro has random LEDs coming on and flickering, what can you even do at that point?
Should I just take the board apart and try to check for corrosion down there or whatever?

>>107307699
instead of the technology we should argue if the board is even any good. i've heard that damn near every company has been having qc problems. reddit reddit whatever but budget keebs has both nuphy and keychron on the watch out thing and i believe it with nuphy which is a huge bummer. their keyboards look great. wooting too expensive

>>107307792
try reflashing the firmware sometimes that solves it
>>107307796
mine seems fine so far, no complaints. the only thing I don't like is I have no idea what hall effect sensors they used and have no clue where to source one in case any break

>>107307834
just tried re-flashing, no dice.

guess i'll man up eventually once these stray LEDs bother me enough to do something about them.

oddly enough they weren't lighting up in DFU mode so it still might be something data related

>>107307674
what? No, the average is like 3-4 ms. 1.7ms is a lowball. I've never seen a non-HE keyboard ever get below 1.3

Most people are using a Razer board that is like 3 ms+. Boards didn't start being sub 2ms until a year ago.

MX switch keyboards can't get 1.16ms because of debounce. Sub 1 ms is only possible in HE keyboards. It is kind of dishonest of me to use Wooting because they are the best, at .400 ms. Most HE keyboards are .900 ms. I consider wooting the only real HE keyboard atm.

(keychron for example is awful, and has 4 MS response.)

>>107308469
Keychron has 500Hz scan rate (despite their marketing), on HE. I tested it myself. I increased it to 1kHz and I got 1.17ms average on joystick input. whatever happens outside of the USB port is not up to the keyboard, it's the OS settings etc.
plus I can't do much as I don't have usb 2.0 high speed on the micro, only full speed, so max 1kHz polling. seems to work out at 1.17ms with that controller test, which seems to be highly regarded among controller users. see https://gamepadla.com/soft.pdl


>>107308105
could be a broken LED, you could replace it if you can find spares and willing to solder.

played around with The Outer Worlds 2 for a bit, and damn that controller input is really nice for movement. I can move faster or slower depending on how much I press the keys. it's just fucked input settings, mouse doesn't properly work when in controller mode. but movement is so good with keys in controller mode damn

>>107308503
joysticks don't have to work in a matrix nor do they have to deal with debounce like MX switches do. You are comparing apples to oranges.

>>107308515
yeah that's the point, analog input is faster.
they claim 1kHz polling rate because they do set 1ms scanning in QMK, but the rest of the scan settings don't help reach 1kHz. with the adjustments I made to wait times it goes to 1kHz and that's fine for me, I kinda don't care about 8kHz desu, I'm not into esports.

>>107308548
-_-; then why the heck are you arguing no one can notice the wootings .400 reaction time when you fully acknowledge most people are probably using a keyboard with 3-4 ms reaction speed

>>107308578
because nobody(*) can. in a double-blind test nobody can. there's a handful of people on this planet that can notice it. I mean full 1kHz to full 8kHz move.
you do not seem to understand what we're talking about. 1kHz means 1ms, 8kHz means 0.125ms. the reduction, if perfect with no other delays on the way, is 0.85ms, between 1kHz and 8kHz. without anything on top from the OS. I got more reduction in response time by moving from 500Hz to 1kHz, than I'd get from going 1kHz to 8kHz.
whatever tf gains are >1kHz, they are not for mere mortals. it's like audiophiles arguing they can tell 0.0000001% distortion from 0.00000001%

>>107308670
>because nobody(*) can. in a double-blind test nobody can. there's a handful of people on this planet that can notice it. I mean full 1kHz to full 8kHz move.
I can. It takes awhile, but use a 8kHz HID device that is true (the only ones who do it well are Logitech (mouse) and Wooting, and some no-name mouse brands) for a week or two then drop down to 1000hz.

I noticed it organically because a wooting fireware (that I didn't know abouit) update dropped my 8kHz polling back to 1kHz, and it was immediately obvious. whatever double blind study, it sucks. Or they are using some unicorn keyboard that actually does have 1.3ms input and not a more standard 3 ms

>>107308578
I think classical switch testing is limited because you need to record when you actuate the switch and measure until something pops on the screen.
on analog you have a stream of continuous data so it's simpler to calculate, since the software is looking for the next data packet and basically measures the time in between them, which. so only 1 keypress sends a fuckload of data points, and the difference in value between each, and difference in time, makes it able to calculate the actual polling rate. which comes at 1.17ms for analog stream of data.
in keypress mode keyboard micro decides the keypress depending on the set actuation point, once that's reached it sends the keypress.

>>107308761
you aren't able to tell the differnece between 1.17ms and 0.4ms. you cannot do it, unless you are some top esports player.
>but i can between 4ms and 0.4ms
you can't between 1.17ms and 0.4ms. I am speaking about controller mode movement, you can't. it's too fast to tell there's a difference.

>>107308771
no, most testers (including me) are using a servo + high speed camera to record reaction speed. I don't think there is a flaw in this method if you average it out over a few tests

>>107308796
stop bringing your stupid controller into this, we are talking about a fucking 3-4 ms Razer gaming keyboard (which is STILL a lowball, Im just choosing the quicker normie gaming keyboard) vs a .400 MS wooting HE80

anyone can notice a difference between these two.

>>107308824
>no, most testers (including me) are using a servo + high speed camera to record reaction speed
that's dumb, because you are talking about a primitive mode of movment in games, using single key input. you press A it sends the A keypress once it actuates.
in controller mode it's sending a continuous stream of datapoints, so there's 0 need for your nonsense setup. that's for primitive single action keypresses, not sure you understand, you seem confused as fuck.
you cannot possibly actuate a button that fast, more times in a row, to measure the difference in between data packets to figure out how fast the key is scanned. you just can't, thus you need to do it differently, measure when you actuate the key and then measure when it's registered on the computer. and that doesn't calculate the keyboard polling rate, but your whole system's polling rate, OS losses included.

when you have x value at t0 then y value at t1 and z value at t2, you can check if y is not equal to z that means value has been updated, in t2-t1 time. which gives you the keyboard's scan rate.

>>107308824
>stop bringing your stupid controller into this
the discussion is literally about Hall Effect keyboards not your ancient technology bullshit lmao

the whole point of the discussion is you justifying $240 extra cost on a Wooting HE vs a Lemokey HE.
is the $240 extra worth it for going from 1.17ms to 0.4ms? that's the whole discussion here.
let alone the fact that Wooting isn't QMK compatible

real grandpa hours, how was tuning your guys's hall effect boards or whatever? did you just leave it on default pleasant typing experience or did you actually like adjust the mm activation thing to feel the best
i guess since you can set the thing to whatever you want you could just copy other switches like having the uh depth (?) of like a red or something?

>>107308878
Dude Lemonkey HE keyboards are 2.9 ms response time at BEST.

you just don't understand how amazing a Wooting HE80 is. You are HIGHLY underestimating how quick .400 ms actually is.

>>107308845
Yes I don't understand how that is any different than using a high speed camera and a servo.

>>107308922
Most people profile switch it to 2mm for games, then back to 4mm for typing. It's the most superior method atm

File: stock_polling.png (116 KB, 747x397)

116 KB PNG

>>107308944
holy shit you are confused as fuck
listen, picrel is with stock Keychron (Lemokey whatever) firmware, the one that connects to their online shit, the one that has been tested by rtings.com
same fucking test, 500Hz polling rate. 2.35ms polling for KEYBOARD. that's what it does, stock.
500Hz means 2ms. with some overhead 2.35ms. that's the keyboards internal scan rate. actual raw performance. with stock firmware.
this is confirmed by me installing the stock Keychron (Lemokey) QMK fork, and enabling debug and scan rate in debug console, and it shows the same 500Hz scan rate for the analog matrix. this is not debatable, confirmed by both testing and QMK debug data. this is the very best stock performance, internal to keyboard, not to computer.
they did set 1ms scan rate, but they are dragging that figure down with higher wait times and longer ADC averaging (ADC_SAMPLE_56 stock)

File: rtings.png (160 KB, 875x945)

160 KB PNG

>>107308944
>>107309003
now, picrel is rtings.com testing of P1 HE, and it's the best performance out of any Keychron HE models. highest rating out of all.
notice the 2.9ms latency, and 2.35ms from my tests. that 0.55ms is losses between keyboard and final data on screen.
now, I moved from 500Hz to 1kHz polling, confirmed by both QMK debug for matrix scan rate, and testing with the same software, which shows 1.17ms instead of 2.35ms. which is twice as fast. if you want to adjust the rtings test value, it comes to 1.75ms for 1kHz true scan rate. 1.75ms is pretty fucking good. even 2.9ms is pretty fucking good, 9.1 score on rtings.com, global rating. that's fucking good. it says so right there in the fucking text
>exceptionally low latency
with 2.9ms for fucks sake. I moved it to 1.75ms worst case, could be even lower, test software says 1.15ms.
now, is 1.15ms to 0.4ms worth the extra $240 or not? answer this

File: 1759116260408341.jpg (96 KB, 600x662)

96 KB JPG

I want a silent mechanical keyboard with those puffy keycaps that look like candy

what do i get?

i want it to be sturdy and heavy and i like a lot of resistance on key presses

File: amIreGirlretard.jpg (135 KB, 1080x1080)

135 KB JPG

>>107309030
Hmmmm, I don't understand a thing you are saying, its like 1 in the morning where I live and I'm in bed. Im sorry I will read again tommorrow

personally $240 for .400ms is worth it over 1.3ms (which is what the wooting gets at 1000hz) tho. It was gross going from 1.3ms to .400ms. I am a competitive player however, my rig is expensive and specifically geared (360hz monitor, 5.8ghz overclocked CPU) to minimize as few HID interrupts as possible, so you are asking the wrong person that question.

>>107309096
I did kept mentioning NOT esports players. like your average normie looking for a hall effect keyboard.
plus yeah I love to nerd out with modding QMK shit so it's different strokes for different fags. still, $240 extra for just a little bit extra isn't worth it in my case

>>107309096
any way I can optimize linux for those hid interrupts thing?

File: rapeboard.png (371 KB, 808x458)

371 KB PNG

>>107309070
yunzii c75 Cake Meow, no balls you wont

File: 1710762936722388.jpg (61 KB, 479x450)

61 KB JPG

>>107309155
anything that doesn't look so girly?

>>107309165
do it anon. do it

>>107309155
that's a keyboard for >200wpm typists only

File: 1706241279705243.jpg (35 KB, 720x960)

35 KB JPG

>>107309288
i type with 2 fingers

File: 4098fce6393942c5b272c3c27(...).png (346 KB, 737x512)

346 KB PNG

>>107309320

>>107309133
linux is random, you either have a 8000hz HID that is registering more interrupts than Windows, or your computer is crashing.

I only run ubuntu for my server, but I think I might try arch whenever the steam computer comes out and see how quickly it can drive a 8000hz keyboard and mouse.

How's TypePlus stab?

>>107309491
good. all you need is 205g0 and very little of it. i've used typeplus twice and neither time did I have to waste a second afterwards actually tuning the stabs. they're just a little more annoying to put together because left and right side parts aren't interchangeable, they have to match.

Why not just fully dye-sublimate your keycaps to provide extra protection against shine?

File: 1761442892853536.png (477 KB, 1107x1017)

477 KB PNG

Hello lads. If I buy the keyboard (hot swappable if relevant) at bottom can I buy an ansi keycap set and make it look like above? Will enter and shift fit?

>>107312143
After some googling it would appear that I'm fucked

>>107312237
don't look at it like that. :3
Think of it as being granted the opportunity to buy a new keyboard. Lucky!

>>107312143
No.

Reminder: tomorrow at Cannonkeys (US) the keyboard in the OP (Brutal V2 1800, barebones) will be only $150. Very nice very underrated keyboard. Very low front height!

But some anon who wants a 96% is going to wait until two weeks from now to come into the thread asking for recommendations, there's no avoiding it...

File: 1743963123008484.jpg (58 KB, 681x660)

58 KB JPG

>>107312457
Problem is the only I'm looking at has a 44% discount whereas the ansi equivalent is full price. Not cool!

>>107312797
the one*

>>107312797
It's on sale because more people find it undesirable, just as you do. That shouldn't influence your purchasing decision unless you are willing to adapt to a different keyboard layout just to save a little money.

File: gemmy_Temple_OS.png (2 MB, 1070x1070)

2 MB PNG

>>107297628
>threw out an IBM keyboard from the 90s without thinking about 2 years ago and only realised it a week ago. fuck my chud nigger incel life.

>Budget: 200 eur and below (ignoring shipping)
>Location: Ireland
>Preferred switch type: linear or tactile, no preference beyond that
>Layout: ISO UK QWERTY
>Form factor: 100% and it doesn't need to be compressed
>Backlight: so long as each key is "lit" so I can type in low light environments, I don't really care.

wanted to buy a V6, but they're only in ES or DE layouts from Keychron, the rest are sold out in Europe

>>107309165
no balls do it

>>107311578
Shine is good

hello /g/ooners
need a 60 or 65% keyboard, clicky switches, ideally low profile, must not be wireless (work bans them).
what do?

>>107311578
Dye does not protect the keycap from wear; t's just dye. Also there are PBT keycaps that dye the entire surface.

>>107313452
yeah but theyre fairly expensive compared to "standard" keycaps

File: 1000070580.png (10 KB, 650x450)

10 KB PNG

>>107313377
Omega levels of cope

>>107313470
Ok, and?

>>107313400
Nuphy air60 easy
The lofree flow 2 68 is also pretty good

>>107313290
>from the 90s
Good chance it was trash then, don't beat yourself up over it.

>need a backlight
Nevermind what I said before about not beating yourself up. Please hurt yourself

>>107313400
Oh no wireless... God dammit why does everyone have some oddball request to their rec that cucks them out of 99% of the market

You aren't going to find a ton of low profile 60 keyboards that are wired only. A Low profile and 60 keyboard is designed with frequent transportation in mind so all of them will have wireless options. You can just like... plug them in and put it in wired tri-mode. But if your company needs it to be not wireless for security reasons then that won't work.

>>107313779
People type in dark rooms anon. My bed keyboard has to be like a Christmas tree or else it is impossible to see

>>107313400
Bauer Lite is a very nice board. Injection moulded polycarbonate 65%.

>>107313839
(It's not low profile, mind you.)

>>107313400
Keychron V4/Q4 are wired and 60%

>>107313779
people who have issues with lighting are deranged

>>107313852
hard to find low profile only wired. low profile are usually wireless

>>107313779
It gets dark at like 4 pm now, I'd like a backlit keyboard

>>107313981
Just turn on the lights, you'll ruin your eyes after a short few years if you stay in the dark.

>>107297628
i have a keyboard that looks exactly (keychron v1) like this but i have DROP MT3 WoB keycaps i got for cheap
also: which stabilizers should i get? mine suck ass mushy and unstable

>>107314000
I like what I like, please stop being a faggot

>>107313797
If finding the keyboard they want was easy they wouldn't be here. That's why all the recommendation requests are retarded.

>>107313630
You have shit taste pegboardloverfaggot

>>107314593
>t has shiney red GMK keys he deeply regrets

>>107314200
My brother uses his keyboard in the dark, it doesn't have lights. Touch typing + light from the monitor makes it easy to see the legends.

If you go for some type of alternative switch spec that does backlighting well—rubber dome membrabe, romer-g, maybe others idk—fine, ok. I just wish more people understood that MX sucks at backlighting, wasn't designed to accomodate it, and it looks bad 100% of the time

>>107313797
>God dammit why does everyone have some oddball request to their rec that cucks them out of 99% of the market
You don't understand. It NEEDS to be a full-size nord iso low-profile backlit wireless keyboard under 150€. I can't live without even one of those things

File: mrw.jpg (77 KB, 500x761)

77 KB JPG

>>107316115
This but DE

File: Keyboards.jpg (2.17 MB, 3947x3000)

2.17 MB JPG

consoom

>>107317328
Nice boards! I really like the orange one

>>107317598
Dankeschön

File: 25umqgwnqh511.jpg (46 KB, 640x361)

46 KB JPG

fuck you guys, get on my level

Just buy b-stock f1 and be done with keyboard stuff

>>107306071
this nigga doesn't live near a micro center

>>107318793
no ergonomics

File: 1750175655993655.jpg (230 KB, 1595x1080)

230 KB JPG

Keychron K2 HE or Asus ROG Strix II Scope 96 for mainly wireless use

Hey guys, need a big favor -

split key, but it also has to allow mouse options for thumb like thinkpad, ball, joystick, touch pad, ect ect. wireless would be nice too. and i have kinda big hands .

>>107319241
I wouldn't touch "gaming" stuff

File: 1740288541591743.jpg (198 KB, 1000x563)

198 KB JPG

>>107319631
I just want something that lasts long between charges and apparently it's really good at that
Another one would be the Logitech G515 Lightspeed whatever

>>107319667
AirV3 has three months of daily use between charges, 1200 hours on one charge

File: 1736529202581003.png (140 KB, 803x1029)

140 KB PNG

>>107319683
This thing? Looks neat but also more expensive than the ones I've mentioned

>>107319850
K2 HE is hall effect. you want latest tech, and battery life, and cheap. not many options. there's maybe classical switches with longer battery life at affordable prices. doubt you'll find hall effect ones.

File: 1754508210375326.png (12 KB, 243x108)

12 KB PNG

>>107320060
Only reason I mentioned Keychron is because I'm using a 3yo K2V2 right now which shits itself on Bluetooth if I get too far from my computer (like 2m away) or have the backlights on (dies in 5 minutes), I'm guessing it'd be a tad bit better with a 2.4 dongle.

All the ones I've narrowed down here are all the ones with supposedly good battery, some of them are going for a pretty decent price compared to msrp. I'm assuming I want to avoid gaming stuff because they're overpriced but in this case, not so much?

File: 2024-09-16_13-30-08-1536x1037.jpg (128 KB, 1536x1536)

128 KB JPG

Are f104's based? i hate my keychron q5 max. the price tag is bothering me though.

Should I buy a Glare TKL?

>>107320142
HE has high power consumption. QMK also isn't great super great. Whatever some larger copro like logitech is using is likely to be shit too. Your best bet is to get something with ZMK specifically. ZMK, no LEDs mogs literally everything for battery life. I'm talking like months of battery life on a 600ma piece of shit.

Do you have to have a TKL specifically?

>>107320267
>get something with ZMK specifically
Doesn't seem to be anything of the sort from what I've tried to find. I do want backlighting though. I don't really mind about the size and form factor as long as it's great on wireless, battery-wise and signal-wise.

explain to me whats the point of hall effect if i don't game.

>>107321204
>possibly useful functions
>"ah I don't care about them anyway"
you're probably someone who doesn't benefit from this technological improvement. in which case you can settle for older technology

had a look at stm32f405 and if using the HS PHY for 8kHz then that eats into some of the ADC inputs, in which case you cannot parallel scan one row at a time.
seems like stm32f407 would have 15 free ADC inputs, while using an external HS PHY.

>>107320156
>i hate my keychron q5 max
>going from full metal to plastic nonsense
ngmi

File: s-l1600.jpg (109 KB, 800x600)

109 KB JPG

yeah and stm32f407 needs to be 144 pin package so you get to connect high speed USB PHY and have at least 15 ADC inputs. and that package could fit but looks retarded, it's huge. and will waste a lot of extra pins.
see pricrel, yeah that's an ethernet socket. it's that large.
you can use f405 with 64 pin package, for high speed USB, but won't have enough ADCs to scan one row at a time, but could still pull it off somehow. just not as straightforward
so simplest to use is teensy 4.1 but can't properly implement on a PCB, needs socket and takes extra vertical space

STOP FUCKING TALKING ABOUT HALL EFFECT MAGNETIC BULLSHIT THIS IS MKG NOT HEG

>>107321204
outside of turning everything into speed silvers, literally nothing

>>107321204
No chatter, inherent NKRO, no need to worry about contacts wearing out ever.

>>107321465
>so simplest to use is teensy 4.1
for high-speed USB (8kHz). but in that case you might be ADC resolution limited, 12bit might not be doable

>>107321475
stop resisting technology advancements anon. hall effect switches are as mechanical as it gets. things move, mechanically speaking

>>107321492
>>107321490
>>107321480
>>107321511
shill me a TKL prebuilt HE keyboard. I don't have time to build one.

>>107321490
>inherent NKRO
what board doesn't have diodes in their key matrix???? only the absolute cheapest shit imaginable. also, isn't an HE implementation more complicated than a regular mx matrix? how is that even an advantage?

>>107321513
idk, wooting or keychron crap

>>107321513
keychron is having black friday I see.
not tkl but K2 HE is $104 which is crazy price for what it is. non SE version (not wood on sides). SE version is $112
Q3 HE is TKL but $204. not sure it's worth it
Q3 HE 8K is TKL and 8kHz, $230
K8 HE is TKL and $119, wood on sides. non SE is $111
don't know others, Wooting stuff etc

>>107321533
>isn't an HE implementation more complicated than a regular mx matrix? how is that even an advantage?
depends, has an extra sensor on each key. it's not rocket science, it's pretty basic shit.
keyboard PCBs are literal n00b first project tier. they're that basic and simple, even hall effect. you just have more of the same thing (80-110 keys), but overall it's extremely simple and dumb

>>107321552
Actually is 65% keyboards worth it? I use the navigation keys heavily and I don't see a home and insert key on 65%.

The keychron 65% ceramic has piqued my interest

>>107321562
do not get near that shit. it's bad
https://youtu.be/vD3VWr4mPvQ
go for 75% if you want smaller, that's the sweet spot.

>>107321562
It just depends on you. I use 60% and I love it. I just have a fn layer when I hold space that turns my homerow into nav/arrows and it's way faster than having a dedicated nav cluster.

>I don't see a home and insert key on 65%.
As long as whatever you're buying has QMK or ZMK (sometimes the product page will only say via or vial compatible which is essentially the same thing as saying it uses QMK) then you can program the keys to do whatever you want, including complex layer logic, or whatever.

just noticed cannonkeys has a 24 hour flash sale on bakeneko right now

$70/$80 bucks and it includes keycaps and switches but you don't get to pick anything besides board color. pretty good deal for baby's first custom.

SHOULD
I
BUY
A
FUCKING
GLARE
TKL
IN
SS?

>>107297628
Any good mechanical keyboards with a trackpoint?

>>107320191
Fuck yes

what stabilizers should i get
i'd prefer to spend $10-$25

is hall effect the oled of keyboards
keychron k2 he is a hundred bucks

>>107321954
>almost half a grand on a keyboard
dumb move, but I kind of want an enso-e so I can’t fault you

>>107322303
typeplus and knight v4s are pretty much unanimously considered the best. i’ve only ever tried typeplus and they were good, just slightly more annoying to put together than regular stabs. their advantage is you only need a little 205g to lube them

File: file.png (243 KB, 1261x676)

243 KB PNG

>>107322365
this one?
i can get this on aliexpress and the other one is like $25 shipping or at least i couldnt find it locally
picrel is like $20 which is ok, txstabs which is what used to be recommend are closer to $30

>>107322426
also what lube should i get? i don't want mush

my company is giving me 90 bucks to buy a keyboard, what should I get?

>>107299744
I think I found compatible HE sensors for my Lemokey. Seems like Allegro A1308 should be compatible.
Looking at their datasheet, they have 5V for power supply, and 2.5Vout quiescent so no magnet present, which is what I have on mine.
Looking at the KS-37B switch specs here:
https://www.hlplanet.com/gateron-magnetic-switches/
It shows 120Gs to 800Gs between not pressed and fully pressed. With this figure I can look for the required sensitivity, at 800Gs for maximum Vout so highest resolution into ADC I need something like 2.5V (2.5V to 0V) / 800Gs =3.125mV/Gs which is exactly one of the options for sensitivity (A1308LLHLX-3-T). This model seems fine but it's risking it right at the edge. So probably next one at 2.5mV/Gs (A1308LLHLX-2-T) is a better choice. That would make for a total of 2V signal difference between not pressed and fully pressed. So 2.5V to 0.5V. Seems like the best choice.
What is now used is most likely 1.3mV/Gs (A1308LLHLX-1-T) sensitivity, since at 800Gs that translates to 1040mV signal change for full travel, which is what I measure, around 1V.
So I might buy a full set of HE sensors with 2.5mV/Gs sensitivity and replace all of them, just to have that extra signal. But damn, they are expensive as fuck. $93 for 100 pieces. I might check the asian market for compatible ones. I now know what to look for.

>>107322448
>>107322426
yes, dunno about authenticity if you’re using ali but just use 205g0 and only 205g0 and not very much of it. its not like other stabs where you glob dielectric grease on the ends of the rails

>>107320369
there are really only two companies that have good wireless firmware and battery life, Logitech and Nuphy

any other brand is trash, logitech for if you want to have maximum wireless performance, nuphy if you want a better typing experience but have to slightly gamble on QC.

>>107322487
Lol found them cheap on LCSC, like 8 cents a pop. Cheap as fuck for a full keyboard. Now I understand how they sell HE keyboards so cheap.
https://www.lcsc.com/category/1244.html
They gotta be linear/compass sensors, not switches. My keyboard has sot-23w (wide), regular sot-23 won't fit. But plenty sot-23w in offer.
Gotta be around 1.35mV/Gs or 13.5mV/mT (10Gs=1mT) for stock levels, or up to 2.5mV/Gs for highest resolution (up to 2V full press into 3.3V max in ADC).
Allegro A1326 also seems to work, but still expensive. Plenty choices anyway.

The magnetic switch types depend on type of sensor implementation. Either 2.5V to 5V into ADC, with south pole towards sensor, either 2.5V to 0V into ADC, with north-pole towards sensor (KS-37B). That's why they are not commpatible

New Keychron (the HE and 8k stuff) does not have latency problem anymore right? Almost bought them two year ago but the bad latency stopped me
>>107319241
I bought it almost 2 years ago. Just RMA because of swollen battery. Almost ran out of warranty (It actually ran out on ASUS website but I still kept the receipt and claimed though distributor.
It's a good keyboard.

>>107322855
HE non 8k is 500Hz actual polling rate and it's fine in rtings.com testing. HE 8k should be 8kHz but they just got released few days ago and can't find any testing for them. so hold back until someone actually tests them.

>>107322855
removed the battery from mine, will discharge it to ~3.7V and store it away. literally don't care about battery I'm not using wireless

>>107322824
Allegro has a problem with their TMR, that kinda makes them bad. All of their chips I have tested do this thing called "hysteresis", its not a problem 99% of the time, but sometimes their TMR sensors gain a magnetic field through a process I do not understand (but it happens after multiple quick keystrokes)

they tell you how to stop it her, but this is over my head https://www.allegromicro.com/en/insights-and-innovations/technical-documents/hall-effect-sensor-ic-publications/hysteresis-mitigation-in-current-sensor-ics-using-ferromagnetic-cores

File: hysteresis.png (149 KB, 1175x446)

149 KB PNG

>>107322906
that sucks. hysteresis can be useful in certain situations, like a thermostat, else it would go crazy right on the boundary, on/off a lot of times crossing over/under set point.
but for a keyboard switch that may not be as good.
unless you are looking for long battery time or classical switch compatibility then HE seems like the simple choice. they're also cheap as fuck from what I'm seeing, at least the obscure brand ones, which is what is used in these keyboards anyway.
I'll get 20 or so with same sensitivity of what I already have, just so I have some replacements, and also 100 with higher sensitivity, I may replace all for more signal. I should be able to get 11bit resolution, so 2048 steps, or 0.002mm per step.

File: 1746219642347929.gif (934 KB, 320x240)

934 KB GIF

NuPhy Air75 V3 for $120 or Keychron K2 HE for $110? For general typing (not gaming) and wireless use. I looked around and people outside of here don't seem to have nice things to say about NuPhy, and on the flip side I'm not too impressed with my current K2 V2.

>>107322906
bro you need to forget about current measuring chips, they are not adequate for the purpose. look for linear compass sensors, not current measurement sensors. they have different applications. the hysteresis issue is for the current measurement ones.

>>107322475
Is it gonna be a work from home keyboard or a take to the office keyboard?

>>107323014
K2 HE SE gets raving reviews for the price and functions. seems to sound too good for its price. but all seem to have the one with wood on sides, SE version

>>107323015
No its a "flaw" in all of the Allegro redrock series sensors I tested.

>>107323096
if you absolutely MUST have low power look at other hall effect chips:
https://www.lcsc.com/datasheet/C52021063.pdf
this one works on SPI and has 8kHz scan rates. you'll have to make a matrix for cable select pin. has an integrated 12bit ADC, which is weird.
or this uPower hall effect: https://www.lcsc.com/datasheet/C842151.pdf
can drive it at 6.25kHz and has enable pin.
you usually select the sensor based on circuit constraints and performance need. I choose a sensor that fits what I need, I don't design around a sensor I choose.

>>107323096
they probably have those ferrite cores

>>107323068
Reading up on HE, I'm getting the impression that it's mostly good for gaming. But being able to literally touch type sounds kinda neat. Also heard that the battery is kinda bad which is not too good for my case but having too plug it in once every couple of days doesn't sound like that big of a dealbreaker.

>>107307291
who else offers a double gasket design? It's what makes the keyboard kino

File: 02b0e26cacde2b4d0aba800f6(...).jpg (2.55 MB, 2541x3276)

2.55 MB JPG

>>107323234
I wish, the HE sensors I have were 1000 times easier to work with than the TMR sensors. I'm having to use X-axis bridge sensors because I don't know how to make the Z-axis normal TMR sensors work. But for mysterious IRL reasons, this project is deadlocked on using TMR. Its turning into a nightmare. I would have finished this prototype keyboard a month ago if it wasnt TMR.

File: tmr.png (31 KB, 871x534)

31 KB PNG

>>107323438
https://www.mouser.com/catalog/specsheets/Allegro_DS_CT100-Datasheet_GS_FINAL_R3-v0.pdf
see this type of TMR sensor, pure Wheatstone bridge. this one is not adequate as it's +-50mT which means +-500Gs, switches go to 800Gs.
but still, at 30kohm for the bridge it means 166uA but you're still fucked on drift and other shit. like amplifying the signal. which you could try and sort with some opamps, after the analog mux, if there's enough signal.
you will probably want this type of tmr sensor, if you want lowest power and full control of the design. but it's a full trip and won't be easy. and you need lower sensitivity to get into 800Gs range

>>107323438
>>107323575
so if you're going to read one row at a time, for fastest scan times, that means 2.5mA constant, for 15 sensors. and that's without anything else, micro, shift registers or mux whichever way you're going. or opamps.
with a 4000mAh battery that means 1600 hours just for the sensors. since you're scanning 15 at a time, and you only power them when scanning them (with a shift register chip). and again, no other consumer in the mix. you need to add probably shift registers, opamps for amplifying the signal, and micro, at the very least. so it's downhill from 1600 hours with a 4000mAh battery, when factoring everything else.

>>107323438
>Z-axis
well doubt the TMR keyboards are using Z-axis ones, they're X/Y. they might still work, these more sensitive ones, since inevitably X/Y picks it up even if coming on Z. with higher sensitivity ones they may work. but you still need to experiment with position and few different sensitivity grades. with pure Wheatstone bridge ones, once you figure out how much signal you can get out of them, you start thinking about opamp gain, and then the rest of details so you get the signal into ADC, for the least power, at closer to 8kHz scan rates as possible.

File: 1000070597.jpg (609 KB, 1314x1825)

609 KB JPG

>>107323575
There is only one game in town for good TMR sensors. That is MDT. Today is the big day, I'm testing the TMR1362S. If this doesn't preform well, I have one more TMR sensor to test before I throw in the towl on TMR completely. Wish me luck ;_;

Ill post pics of the teensy, key and servo testing apparatus later today

File: latch.png (65 KB, 574x181)

65 KB PNG

>>107323798
that's a composite sensor, it has power management, amplification, and it's a latch sensor. so 0/1. might as well throw it in the trash. you need analog output not latching.
and why even use it anyway? the whole point of TMR is lower power for higher battery life, so Wheatstone + well thought out external circuitry. just get a Wheatstone TMR core and nothing else. if you're not going for lowest power draw (power specs mA values) then why bother with TMR?
sorry anon

>>107324075
You don't understand how brutal it is, im just trying to get ANY tmr sensor to work even if its just a switch. The endgame sensor is to use tmr2617s-aac. But I can't get their z-axis sensor working. I think if the tmr1362 doesn't work as atleast a good on off switch, I will start making plans for HE implementation.

Developing opinions about keycaps is so dangerous. You can end up in a position where everything is unacceptable, no matter what. "Sure it has stepped control, F13, 1u win logo, nav sublegends on numpad, windowed numlock, 1.5u fn, and it's an exact color match for my board. But... icon mods and multiply/divide legends... I just can't do it"

This is happening to me right this very minute.

What keyboard should I get if I'm trooning out but I don't believe I'm a woman?

>>107324597
Re-evaluate your opinions, easy. Unless you're a sheep fag who can't stray from the straight and narrow.

>hotswappable
>tri-mode wireless
>aluminium case
>qmk/via
>TKL (87 or 85 keys)
>the physical switch to change between wired/wireless/bluetooth needs to be reachable without pulling a key cap (very important and disqualifies 90% of the keyboards in that category)

>>107324698
Just get an evo75 if you need a switch

The evo80 uses key combos but I guess that won't work for you either

>>107324698
Keychron Q3 max, I'm sorry to say

>>107321533
It's not just about the diodes. NKRO has to be enabled, and/or done without that method that just makes the board look like two in software.

>>107324621
>change your opinions until you like what everyone else does, don't be a sheep

File: iamretarded.jpg (2.07 MB, 4000x3000)

2.07 MB JPG

*scream music plays*

>>107324075
damn bro you were right this shit sucked its not even worth talking about, it did work but lame as hell

I'm trying a TMR2083S next

i just bought some kaihl v2 white box switches. i want more clicks.

File: 2025-11-17-214637_1157x99(...).png (818 KB, 1157x992)

818 KB PNG

anyone knows where to get cheap corne v3 pcbs that ship to mexico?

if you had to choose one between gasket mount and hall effect which one would you go with

>>107328686
Geon's leafspring is truly amazing

qrd on why i need to spend any money above a membrane keyboard to type? ive been using membranes for 30 years

>>107328850
also though some jewish trick i bought a keycrhon k2 brown. i hate it it sucks i haste typing on and have since day one

>>107328850
you don't anon O_O membranes are fine, it takes like $130 dollars to get a keyboard that is noticably better than a good dell membrane

>>107328968
dude i have a keychron k2 i bought years ago. i dont want to hate because the aesthetic is great. i dont like the major travel. i even have a model m i like that travel but the brown switches were such a major disappointment to me i dont want to pretend to anyone else that this was my favorite purchase.

>>107329050
a keychron k2 is straight budget trash, dont feel bad for hating a trash dumpster fire keyboard

brown switches are trash, only sub humans use browns, throw that thing away

>>107329111
its probably trash i bought it like 10 years ago. i like a good keyboard feel but i cant keep buying 100 dollar keyboards. seems like bad budgeting

Are there any decent slim keyboards? Alternatively, anybody got a wrist rest rec, something in the 16" width range?

talk me into or out of a Keychron k8.
Also undecided on a TKL or full layout. I have no disillusion about my gender and tend to use the numpad for number outputs but im not strong willed as needing it. i'll keep my current fullsize keyboard anyway in case.

this will be my first mechanical keyboard.

>>107329166
buy a dell quiet touch off ebay for 10 dollars and a ps/2 to usb keyboard adapter, and come home

Alternatively, buy a rainy75 and try what a real mechanical keyboard feels like

>>107329245
budget nigger, if you want to buy once cry once then get a set of F1-8X and move on

>>107329245
>Keychron k8.
try to buy a Keychron that is atleast in the $100 dollar range, or you will just be buying trash like that last anon

alternativly, try a real mechanical keyboard like the rainy75

>>107329259
pls dont make fun. i really liked the model m but i hate this new switches

>>107328124
>TMR2083S
that's actually adequate. similar to CT100. 10k resistance, vs CT100 at 30k, so three times higher current but still low overall.
measure voltage change between V+ and V- pins. or X1/X2 on CT100.
Simpler to deal with amplifying this signal, since the HE sensors that are used in my keyboard already have 2.5V output with no magnet.
but having access to V+/V- you can use a differential amplifier. btw CT100 has really low noise 1kHz - 8kHz, so as long as you're getting something out of it you can amplify it a lot of times

>>107329308
o_O im not making fun of you, I know people who only use dell silent touch. Its a legit keyboard.

>>107329294
>>107329285
I thought the consensus here was that at 100 and over, the keychrons werent worth it anymore?

>>107329469
yeah the optimal Keychron price is $100, don't pay much over that

>>107329285
bruh that's >$500 territory for everything

>>107329458
lol than you would hate my keyboard of choice

File: 3768452673452673.png (58 KB, 372x551)

58 KB PNG

>>107329469
full metal and HE it's worth $136 for Lemokey P1 HE. wouldn't pay $200 for Q HE series tho

>>107329507
i--it cant be that bad

File: maxresdefault.jpg (126 KB, 1280x720)

126 KB JPG

>>107328124
picrel is how you're amplifying the signal out of V+/V- pins. I would feed the sensor outputs into dual 4:1 analog mux, and mux output into amplifier, and output of amplifier into ADC. this way you have the least nr of amplifiers for lowest used current, overall.
picrel is x7 gain for the signal. at 0.4V into amplifier you get 2.8V out. so you just have to get something out of that sensor.

>>107329539
its the magic keyboard. apple keyboards feel good. apple track pads feel good. whatever ever engineer at the back of apple is in charge of these two things does amazing work.

>>107329599
Thank you for explaining this, using an OP amp is the hardest concept for me, this is really good stuff.i was thinking about using an INA with the MUX, would that work and make it sp i don't need an op amp?

>>107329673
and the guy that does the trackpad should be a multimillionaire, because everytine i fuck about with a thinkpad, the trackpad fucking kills me

>>107329673
Nah the magic keyboard is legit, esp for the price. It is alittle gay tbf. But an excellent keyboard for its price. Top 3 low profile keyboards easy

File: file.png (714 KB, 1080x351)

714 KB PNG

behold!

File: file.png (782 KB, 1080x584)

782 KB PNG

>>107329707
yes, that's a trackball in the right keywell.

>>107329676
opamps are really simple to use. that example has everything about differential amplification. takes the difference between two outputs and amplifies into single output that you feed directly into ADC. INA are current sensors, I don't understand why you'd use that. I'll make an example schematic using an adequate opamp for you.
btw, another advantage for using a Wheatstone arrangement sensor is you can swap V+ and V-, with PCB SMD jumpers. you can have them stock for north-pole magnet facing the sensor. but if you find some south-pole switches you want to use, just clear the solder bridge and short the other side of it, now V+ and V- are reversed into MUX/Opamp and you still have proper signal, with reversed pole magnet in switches. I'll show an example later
I dig the Wheatstone TMR core, it's flexible in PCB configuration

>>107329707
What the FUCK am I looking at

>>107329707
interesting

>>107329700
its not for everyone, i cut my teeth on an apple iie. to each his own, but sometimes i feel touch typing doesnt need an explosion of travel. i dont get that response from using a pencil but i can still use it

File: file.png (1011 KB, 1080x657)

1011 KB PNG

>>107329718
left half

File: tmr_sensor.png (62 KB, 1336x815)

62 KB PNG

>>107329676
here's one example with an analog MUX, it's just a two channel one. this one is 3:1 two channel, but I've seen 4:1 two channels one, can host 4 switches per MUX. or use something like 16:1 and use two of them for 16 switches.
differential output sensors have two outputs instead of one. but are more versatile. I'll play with different configurations as I might design a PCB using them, just for having more switches compatibility.

Worst general on /g/ and that's saying something. Being into mk is the gayest shit

File: bridge.png (181 KB, 659x484)

181 KB PNG

>>107329676
>>107329853
here's what I mean with using both north-pole and south-pole switches, you can bridge with solder, either 1&2 or 2&3 on the solder bridges, so you can swap V+ with V- this way. you can also have them connected a certain way by default, and if you want to swap them you cut the trace between 2&3 for example, and bridge 1&2 with solder. this would make for a sort of universal PCB, no matter the magnetic switch configuration.
if you really wanna go crazy you could also set the gain of the opamp amplifier, from the microcontroller, such that you can software adjust depending on the switch magnet strength, this way you don't give a fuck about any kind of switch magnet configuration/flux strength

>>107329977
go back on plebbit faggot

>>107329984
I mean, when you are paying >$500 for a keyboard, this is the kind of shit I'm expecting to be in it. whatever the fuck they're doing with the PCBs on those higher end keyboards, it's pleb tier nonsense.
at hundreds of dollars I'm expecting full fucking control, top performance, and zero issues with nonsense like magnetic switch compatibility and shit like that.
people dumping so much money on bullshit PCBs for "bespoke" keyboards is fucking retarded