If it’s a through-hole molex should be easy enough to buy separate and solder on, but if it’s anything more than that, yeah, better to have a PCB fab do it.
It’s a 0.5mm pitch SMT connector unfortunately. I just drag soldered one and it was doable, but definitely not easy for beginners.
Ouch. Yeah, easier to get it on a PCB. Also something to consider once a final version is done, is a larger batch order, funded by pre-orders and then extra on Tindie. Larger runs should end up being cheaper for everybody. We’ll see how much actual interest there is of course.
Unfortunately there might be regulatory complications with selling PCBs and/or kits, mostly around EMI/RFI testing. Laptops are intentional radiofrequency radiators (by nature of WiFi and Bluetooth), and the likely required testing to validate that the design or modifications fall within legal limits is not cheap (potentially several thousand dollars at a test lab). There are often legal exceptions for DIY projects, but once you start selling parts and kits to other people you potentially start falling outside those legal exceptions.
I was originally looking at just open-sourcing the design, though not having JLCPCB be able to do the connector is a major issue there for other people wishing to build one. Doing small volumes with other PCB fabs is much more expensive there.
Hm, if you were doing some kind of WiFi/RF/BT device, absolutely something to worry about. For a hardwired connection for the keyboard, unless you’re stupid about things, should be fine.
But, I’m perfectly happy for you to ensure you’re not going to be under potential worry, sure.
I am looking for someone who can make me a touchpad module with at least 2 buttons (3 would be fine as well).
https://community.frame.work/t/touchpad-with-2-buttons-instead-of-clickpad/36343/4
Someone suggested I post over here. I can see it’s overall a thread about keyboards and trackpoints so I’ll just lurk.
Cool! I’ll keep you in the loop. @FeiQi_An has also expressed interest. Up to the hospitality of @dmx0987654321, of course.
Yeah, I didn’t mean to dismiss it, what I meant to say was that I had enough examples of the general style of the T410 that I could probably figure out if that was a good direction to go – if the T4xx/L4xx from my collection-to-be seemed like the front runners when all is said and done then I’d probably circle back and take a harder look at that connector.
Ultimate Hacking Keyboard makes a trackpoint module, you might be able to buy one and look at hacking it into an input module? At the very least, I’m sure they’d be quite willing and open to discussing how it’s made, re-wiring one, finding modules to buy from where ever to DIY up one from scratch.
I suppose it depends on the scope of the changes. If the lower chassis has to be redesigned (almost certainly required for the thick T410 keyboard onto the FW13) - that might be a big enough change to affect the WiFi/Bluetooth/unintentional emissions profile of the device and require re-testing.
Though since EMI/RFI testing only has to be done once a design is finalized (unless the design fails testing and has to be redesigned), it’s a cost that can be spread out among all the buyers, and it might be financially viable to diffuse a multi-thousand dollar test lab bill across a lot of buyers if there’s enough interest.
There might be some wiggle room if it’s a kit, though I’m not sure if that’s true for intentional transmitters (again, due to WiFi/Bluetooth).
The alternative is to open-source the design as a DIY project and publish the manufacturing files but not sell anything. End users would have to source the parts themselves from machine shops and PCB fabs (therefore losing significant cost savings from economies of scale), but there are often some exceptions to EMI/RFI testing regulations for DIY projects.
I know there are some modules that are effectively sealed, like the ESP or nfR that the modules is what’s EMI/RFI, since it’s an active transmitter, and can be used with much less or perhaps almost no additional EMI/RFI testing. But I’m unfamiliar with the specific rules of that, and I’m sure it’s more complicated than what I’m talking about.
Even just releasing as open specs/gerber files with instructions on where/how to get a PCB made up is massively awesome, and I’d be happy with.
This probably deserves a separate thread of is own, but: The FCC and Open Source Hardware - SparkFun Electronics
EDIT: and I didn’t realize that @pylon themself wrote an excellent survey: Your custom keyboard is probably illegal - a long post on radio interference testing regulations and keyboards - Learning and discussion - KeebTalk
4 Likes
Oh! That’s really good, thanks for posting that. Not a lawyer, but it might be that the adapter board would fall under one of those exceptions mentioned. On the other hand, I’d never want someone to take such a major risk if they don’t feel comfortable with it themselves.
Oh we can chime in with money? Where’s that meme when one needs it so badly 
1 Like
Since folks were so generous with contributions to the keyboard fund, I went ahead and threw in New Original Laptop Lenovo ThinkPad T460s T470s Touchpad Clicker 00UR946 00UR947 | eBay to play with as well, once Framework gets around to documenting the touchpad side.
But right now Framework and hardware hackers alike are going to have to sit on their hands for a couple of weeks while we wait for folks to return from their trips to/from home for Chinese New Year. I think my “version 2” boards made it out of the factory before everything shut down, but it looks like they didn’t make it all the way out of China and might be stuck at the shipping company until after the New Year break. Alas!
7 Likes
Amazing work!
1 Like
Meanwhile perhaps let’s hack on software – the negative inertia part? I would love to help if I knew how… however, the last time I wrote C code was in the previous millenium 
Thank you for including touchpads.
1 Like
I can provide some pointers – basically you want to take this code from the linux kernel, which configures all the special trackpoint registers:
linux:trackpoint.c
linux:trackpoint.h
And port it to this file in (say) QMK, which doesn’t do anything trackpoint-specific at all:
qmk:ps2_mouse.c
Consulting the IBM Trackpoint Specification for guidance.
I don’t know /exactly/ what the magic sauce is, maybe you just need to initialize the various special registers correctly and let the trackpoint do the negative inertia magic itself, or else maybe you need to reimplement the magic transfer functions described in section 2.6 of the trackpoint spec from scratch.
1 Like
This is I know for certain: the TrackPoint does the negative inertia magic by itself. I think the “tuning for feel” you mention is the “Negative Inertia Factor” in the specs you linked, and I believe this shows up in the Windows Control Panel as the “TrackPoint sensitivity” slider. And from experience I can confirm it’s a huge, huge difference. I do wonder whether the respective OS drivers would recognize the TP as TPPS and use their drivers – or we need to do more for that? And also … how would I test anything here?
2 Likes
