I have so many unused switches and switch parts. I’ve truly haven’t found a good enough frankenswitch solution to combine anything in years.
I was on the hunt again today. Hours of trying this stem in that switch. I must have at least 100 sets of different switches that I don’t like lying around unused. But I finally found something that really works and is unique… but there’s a catch. It’s very short travel.
Combined v1 linear Naevy housings ( I harvested the stems from 3 or so years ago) with what I think are cherry speed sliver stems. They must be… but I never noticed that they have extra legs on the bottom rails to reduce travel.
Lubed up with 205g0 and filmed, these sound incredible. And the feel is quite unique. I think it’s because the rails are slamming into the resonant switch bottom, instead of the intended pole bottom-out. I measure 3.2mm travel. I know, that sounds off-putting, but the sound and feel of the switch might just be enough for me to overlook it.
I tried them with a dual stage 63.5g switch and it was nice, but I think I can use something more traditional and save those springs for another switch. There’s such little stem movement that I’m not sure a dual stage spring is getting the most use out of it.
Gotta think of a name for it. Not sure it’s practical since most people can’t get v1 Naevy switches anymore.
Well shit. I got everything prepped and lubed the stems. Assembled 12 switches and had the little voice in my head saying, “better try these out.”
Only 2 of the 12 registered in a hotswap board
I thought it was something about the contact leaf and the speed switch stems, but it appears it’s something to do with the pins because the switches register on a multimeter.
’m not sure I want to put in the effort to film and assemble these if I can’t use them in a hotswap build.
I bet it’s the pins on the HS sockets themselves TBH. I just ran into this when I sold my Mahjong. I used some desoldered (but the pins were cleaned very well) MX blacks in it way back when I first got it. Pulled it out to test it & decided to use some HMX switches cause they were the easiest for me to get to at the time. Only a few of the HMX switches registered. Thinking it was the sockets I grabbed some Cherry switches that never were soldered in & sure enough they all registered fine. I did go through & tighten up all the sockets so the person who bought the board could use whatever switches they want. However that sent me on some KB science & I found out the thickness of the pins vary wildly between manufacturers. To the point some switches will push the pins in HS sockets far enough apart to not make contact with switches that have pins on the thinner side (Looking right at Gateron & their new reinforced pins as those are the thickest IME).
I don’t know the exact history of Kailh hot swap sockets, but they strike me as something invented to make life easier for manufacturers rather than enthusiasts. I suspect the idea was to facilitate multiple SKUs for multiple OEMs rapidly, but not specifically for people to change them more than one or twice, if ever, and certainly not with desoldered switches. Like stabilizers, this feels like something ripe for innovation.
This feels entirely plausible to me, if not in whole on why they started, at least in part. And they definitely were not built to withstand frequent swapping. Most hotwap sockets are only rated to withstand 100 changes, and I imagine that’s under ideal conditions.
However, the newest version of Kailh’s sockets are rated for something like 6000 swaps.
I believe that they’ve improved the switch terminals. Where I remain iffy is on the SMD solder joints, especially on discount PCBs. In my own experience, as well as probably dozens of disappointed folks I’ve seen on various forums and social media, that’s what fails, generally upon re-installing switches. If nothing else, the community and vendors should be strongly recommending that wherever possible, hotswaps be changed out with the back of the PCB supported.
Hell, my entire journey with hand-wiring larger boards started after finally giving up on the PCB of an FL-ESports 980 with a lifted pad that refused all my attempts to jumper it. Before that I’d just done a couple of little orthos that I didn’t really want to use.
I don’t know how the hotswap sockets on my first (entry level) mech survived my clumsy attempts at changing the switches. Zero support for the sockets on the first two or three switch changes I did, but surprisingly the sockets are all still working.
There really should be a big warning label on all hotswap PCB packaging that instructs users to support the sockets each time. It feels like common knowledge now to most of us here, but the hundreds of reddit posts asking for help with broken sockets proves otherwise.
My dislike of soldering triumphs over my wariness of hotswap socket durability.
These caramel latte switches are pretty fantastic with lube. If I had one complaint it would be the spring is slightly lighter than I would like. Would like to have about 5g more resistance. Just make sure you trim the stems if you decide to try them.
Custom PCB to use Kailh Choc v2 switches, specifically the whale silent tactiles. External MCU (Waveshare Zero) because I haven’t actually laid out a RP2040 based board before (plenty of AVR based ones tho). Case is a standard GH60 low profile with some light Dremel work.
I have an ISO(ish) version in the works. That build is currently waiting on more hot swap sockets, stabs, and a custom cut plate.
Replaced Ink-bottoms of my Cthulhu frankenswitch set with Mekanisk Ultramarine V2’s polyamide bottoms. It now sounds more distinct from the Cream Soda set and prettier. Mekanisk Ultramarine V2 with Ink-bottom sounds pretty good too.
I swear I see some 0.75u keys lol
