No freaking way this is happening...
It's happening!!
Gondo to the rescue! As someone who owns multiple HHKBs and loves them to death, this is amazing. I can't wait to see and hear more.
1. I don't care for having a USB hub built in.
2. Yes to QMK and VIA.
3. I think USB C is neat but not a necessity. Plus, I don't want to file my HHKB case for fitment.
4. Pricewise, what's feasible? Like $50 per piece?
Would love to see this for the Realforce 87 and 86u. Especially the latter. I have a "84u" that depends on AutoHotkey and Karabiner Elements for it to work the way I want it.
I would be interested in such a project.
1. I don't personally need a USB hub built in.
2. QMK and or VIA wold be excellent.
3. Bluetooth is not a requirement.
Great to see this.
This is amazing! While I'll skip the first round because I don't have an HHKB that I want to swap out, I would be seriously interested in the next round if you created a new design in a 65% (or 75%) formfactor. I really liked the feeling of Topre when I've tried them at meetups, but haven't picked one up for exactly the reasons you mentioned above.
This is amazing! While I'll skip the first round because I don't have an HHKB that I want to swap out, I would be seriously interested in the next round if you created a new design in a 65% (or 75%) formfactor. I really liked the feeling of Topre when I've tried them at meetups, but haven't picked one up for exactly the reasons you mentioned above.
- First, I need to know if there is interest for this kind of PCB -- a QMK compatible HHKB Pro 2 replacement PCB.
- Second, I need an input on the features we are going to put there. The first idea is to make a simple PCB that can operate Topre and replace the USB hub
- However, we are also able to add native bluetooth and a very nice USB3.1 hub in there; these will ramp prices up because the bluetooth controller alone costs 9 dollars, and the USB3.1 hub controller almost 15. So our idea is to make a first operating version and, futurely, an "ultimate" version will all the big boi features. In this sense, I need to know if you guys agree with this or definitely want a first PCB with bluetooth or 3.1 USB or both.
Finally, I want a target price you guys would be comfortable to pay for an HHKB replacement PCB. I know this is not ideal but this PCB will not be cheap as a DZ60 and, as you have seen, there is a serious tradeoff between features and price here.
Show Image(https://media.tenor.com/images/d46785e4ab591e604e61c42e970ea016/tenor.gif)
Cannot express how excited I am for this. If this project is successful, it opens a whole new door to custom Topre boards, and boy oh boy is that a door I want to pass through.
- First, I need to know if there is interest for this kind of PCB -- a QMK compatible HHKB Pro 2 replacement PCB.
Hell Yeah!
- Second, I need an input on the features we are going to put there. The first idea is to make a simple PCB that can operate Topre and replace the USB hub
I don't care about the USB hub to be honest. Programmability with QMK is all I want from this PCB!
- However, we are also able to add native bluetooth and a very nice USB3.1 hub in there; these will ramp prices up because the bluetooth controller alone costs 9 dollars, and the USB3.1 hub controller almost 15. So our idea is to make a first operating version and, futurely, an "ultimate" version will all the big boi features. In this sense, I need to know if you guys agree with this or definitely want a first PCB with bluetooth or 3.1 USB or both.
I don't need Bluetooth and the PCB with all its ports should fit in the HHKB case without any further modifications needed.Finally, I want a target price you guys would be comfortable to pay for an HHKB replacement PCB. I know this is not ideal but this PCB will not be cheap as a DZ60 and, as you have seen, there is a serious tradeoff between features and price here.
As long as it is cheaper than the already available Hasu controller I'm fine!
Count me in for this one, every feature is welcomed for me, i feel like celebrating now ^^
Will it work (supposedly with some small modifications) with niz Switches? They are on par with topre except for rubber dome forces (which are easily available elsewhere). The niz switches are plentifully available.
1. Definitely don't need 3.1 hub on board.
60$ for hhkb size pcb is fully ok.
Sent from my iPhone using Tapatalk
I'm all in on this!
Very interested. I would like to see a version without the usb hub in addition to the regular one too for Pro1 compatibility.
Very interested!
Oh God... Endgame is finally coming
Wow this is actually finally happening... someone did it. What probably 20+ "ic"s have tried to do over the years.
I'm just now reading through all of the accompanying PCB documentation but already so far I've learnt a ton in 1 page. Thanks so much for this Gondo.
Very interested. Just got a Pro 1 so if there's a pro 1 version of this I'd be interested in that too
Very interesting!! I will keeping my eyes on this!
I’m ok with USB mini as I don’t like to modify the case. I don’t use the USB hub on the board but I would appreciate if there’s USB A ports to fill the gaps. So even if it’s 2.0, I don’t really care. Purely for the aesthetic.
Bluetooth will be a nice complementary.
Sent from my iPhone using Tapatalk
I have an fc660c so won't buy into this round, but I think this is super cool work. If you ever do one for the 660, I will be in.
I also don't care about having a USB hub built-in and wouldn't recommend USB-Cs for HHKB Pro 2. I'd just focus on making an HHKB Pro 2 PCB replacement first, and maybe an 86/87U later on.
Far future/Wishlist: a 65% custom board.
I would like at least a 2.0 hub as I usually keep an HHKB and wireless mouse together. Bluetooth would be awesome as it would allow me to make my HTPC setup wireless, but isn't exactly necessary. I would prioritize a hub over nothing at all though. Definitely interested and probably in for 2 depending on price.
I don't care for the USB hub or bluetooth. USB-C is available on the new version of Hasu's controller (https://geekhack.org/index.php?topic=103495.0) with a castellated daughter PCB. It works perfectly without filing the case.
Just F.Y.I, the BT module made by Yang, sold on KBDfans does have a USB C variation. I saw that on TaoBao.
Sent from my iPhone using Tapatalk
Just F.Y.I, the BT module made by Yang, sold on KBDfans does have a USB C variation. I saw that on TaoBao.
Sent from my iPhone using Tapatalk
Hhhmmmmmm do you have a link? The only one I was able to see was this one https://kbdfans.com/products/hhkb-ble-mod-upgrade-module?_pos=19&_sid=cf4104854&_ss=r
It is in mandarin...sorry. There’s pictures in the description.Just F.Y.I, the BT module made by Yang, sold on KBDfans does have a USB C variation. I saw that on TaoBao.
Sent from my iPhone using Tapatalk
Hhhmmmmmm do you have a link? The only one I was able to see was this one https://kbdfans.com/products/hhkb-ble-mod-upgrade-module?_pos=19&_sid=cf4104854&_ss=r
This is a Chinese version of AliExpress. So YDKB offer this service where you buy the BT module from him, he can mod the mini usb to USB C for you. I think in the link, he has also been experimenting with wireless charging with Qi.
https://m.intl.taobao.com/detail/detail.html?id=600934489355&spm=a21wu.9600033.recommend.1&main_itemid=590221409485&go_item_id=600934489355&pvid=d532cb6d-08cb-44d7-ba2b-b41976df12fd&utparam=%7B%22x_object_type%22%3A%22item%22%2C%22x_object_id%22%3A600934489355%7D&scm=1007.20269.110938.1002003000000001#modal=share&timeStamp=1583720863480
Sent from my iPhone using Tapatalk
It is in mandarin...sorry. There’s pictures in the description.Just F.Y.I, the BT module made by Yang, sold on KBDfans does have a USB C variation. I saw that on TaoBao.
Sent from my iPhone using Tapatalk
Hhhmmmmmm do you have a link? The only one I was able to see was this one https://kbdfans.com/products/hhkb-ble-mod-upgrade-module?_pos=19&_sid=cf4104854&_ss=r
This is a Chinese version of AliExpress. So YDKB offer this service where you buy the BT module from him, he can mod the mini usb to USB C for you. I think in the link, he has also been experimenting with wireless charging with Qi.
https://m.intl.taobao.com/detail/detail.html?id=600934489355&spm=a21wu.9600033.recommend.1&main_itemid=590221409485&go_item_id=600934489355&pvid=d532cb6d-08cb-44d7-ba2b-b41976df12fd&utparam=%7B%22x_object_type%22%3A%22item%22%2C%22x_object_id%22%3A600934489355%7D&scm=1007.20269.110938.1002003000000001#modal=share&timeStamp=1583720863480
Sent from my iPhone using Tapatalk
Don’t get me wrong... you have support for this project. I think the success of this project will possibly open door to high end custom topre board! End game topre board is not a dream.
Sent from my iPhone using Tapatalk
I am also adding my vote for Pro 1 support. I own a couple of different ones. They're my better Topre boards and I'd be ecstatic to see them used at a higher potential. I can help test if it is necessary.Very interested. I would like to see a version without the usb hub in addition to the regular one too for Pro1 compatibility.
I figured we'd be able to port this easily to the pro 1. However I don't know if the inside measurements of the PCB are compatible, and I have an HHKB Pro 2 here, so that is what I'll start with.
Wow!!! This project definitely gets my attention!
Custom topre time gamers 👀👀
Wow, this looks great! Custom Topre 2020 ;D
Wow this is some serious electronic engineering! You're my mvp today! Hoping for a realforce 88u version sometime in the future :)
Custom topre time gamers 👀👀
This QMK compatible PCB will give me a reason to use my HHKB Pro 2 again. I’ve been meaning to get a HASU controller but this solves everything by staying USB mini and having the USB hub. I’m excited for it!
Any chance you could partner with Niz to make QMK compatible PCBs for their boards?
The Atom68 and Micro82 would be great ones to start with since they use widely available keycap sizes.
As a recent backer of the Atom68 Alu, I would LOVE to see a QMK PCB to take it to the top level. Then we could finally have a capacitive keyboard with an aluminum case and brass plate with a QMK-enabled PCB and MX compatible sliders.
LONG time lurker, but found out about this through Man of Interest. Having just gotten my first topre board HHKB Pro2, I've already ordered a Yang BLE Mod. But I'd still be VERY interested in this.
I purchased the Yang BLE Mod for its Bluetooth capabilities, remapping, and battery life (over the Hasu).
However, I would pick up the Faraday60 in a heartbeat, if it could address these issues, at least for me:
- Hasu/Yang Bluetooth multidevice support is limited by Adafruit Bluetooth module (no real Bluetooth profiles for switching between more than 2 devices)
- USB 3.0 Support all the way through (would settle for 2.0 -- but at this day in age, why settle?)
- Full re-programmability (check)
- under $100 (flexible with the assumption of full Bluetooth multi-device support and USB 3.0)
I would like at least a 2.0 hub as I usually keep an HHKB and wireless mouse together. Bluetooth would be awesome as it would allow me to make my HTPC setup wireless, but isn't exactly necessary. I would prioritize a hub over nothing at all though. Definitely interested and probably in for 2 depending on price.
To be quite fair the main barrier for a good bluetooth implementation is QMK. Now I'm not bashing QMK, let me explain.
Bluetooth is historically a sketchy technology and feature because the Bluetooth SIG (who manages licensing of BT products) has some harsh requirements for implementation, among them hefty licensing fees.
Now, QMK was able to implement BT with the Adafruit BL feather and that is amazing, but it has the issue that it is an older protocol and is not BLE; hence it dries out batteries quite fast. Why won't they support a newer protocol, you ask? Well, there is a very problematic issue between the QMK licensing and Bluetooth licensing terms, for instance the Nordic's (famous BT chip manufacturer) licensing terms.
Second, the BT protocol driver in QMK is limited to AVR chips, and honestly I'd rather wait until someone ports it out to ARM because I'm not a fan of going with an ATMEGA here. I don't know the platform ok, and I am proficient im ARM.
So I think that for now BT is out of the question, but if time comes we can make a BT version no issues.
Absolutely 100% interested in this. I love my HHKB!
Now this sounds intriguing, even possible to go the extra mile and use Niz switches and BKE domes, sourcing springs would be a bit more challenging :eek:
This is genuinely amazing news! Although I don't use Topre, there is no doubt that QMK support for Topre boards is more than welcome by the community!
I've been looking for this ever since I first bought my HHKB pro 2!
I'm in!
I'd buy a new HHKB Pro 2 just for this. Very exciting stuff. Topre forever.
Would also love FC660C support as well as as Plum 84.
One very distant step closer to my fantasy: Topre Southpaw 65.
Unfortunately there is nothing much I can do about the Adafruit Bluetooth. It's what QMK supports, and it's not for lack of good reasons, mainly licensing as I have explained on a comment above:
Any chance you could partner with Niz to make QMK compatible PCBs for their boards?
The Atom68 and Micro82 would be great ones to start with since they use widely available keycap sizes.
As a recent backer of the Atom68 Alu, I would LOVE to see a QMK PCB to take it to the top level. Then we could finally have a capacitive keyboard with an aluminum case and brass plate with a QMK-enabled PCB and MX compatible sliders.
Of course! I will have to make contact with NiZ and I think he will be open to some collaboration. I have received immensely good feedback from everyone and I think this will be a blast!
Any chance you could partner with Niz to make QMK compatible PCBs for their boards?
The Atom68 and Micro82 would be great ones to start with since they use widely available keycap sizes.
As a recent backer of the Atom68 Alu, I would LOVE to see a QMK PCB to take it to the top level. Then we could finally have a capacitive keyboard with an aluminum case and brass plate with a QMK-enabled PCB and MX compatible sliders.
Of course! I will have to make contact with NiZ and I think he will be open to some collaboration. I have received immensely good feedback from everyone and I think this will be a blast!
That would be amazing! Between their new cases and components and your PCBs we would be well on the way to custom capacitive keyboards.
Niz boards are also priced much more competitively than Topre, which would help achieve your stated price goal.
Any chance you could partner with Niz to make QMK compatible PCBs for their boards?
The Atom68 and Micro82 would be great ones to start with since they use widely available keycap sizes.
As a recent backer of the Atom68 Alu, I would LOVE to see a QMK PCB to take it to the top level. Then we could finally have a capacitive keyboard with an aluminum case and brass plate with a QMK-enabled PCB and MX compatible sliders.
Of course! I will have to make contact with NiZ and I think he will be open to some collaboration. I have received immensely good feedback from everyone and I think this will be a blast!
That would be amazing! Between their new cases and components and your PCBs we would be well on the way to custom capacitive keyboards.
Niz boards are also priced much more competitively than Topre, which would help achieve your stated price goal.
I just joined the Atom 68 Alu GB, too. Custom PCBs for the Atom 68 and the Micro 82 would be amazing!
Any chance you could partner with Niz to make QMK compatible PCBs for their boards?
The Atom68 and Micro82 would be great ones to start with since they use widely available keycap sizes.
As a recent backer of the Atom68 Alu, I would LOVE to see a QMK PCB to take it to the top level. Then we could finally have a capacitive keyboard with an aluminum case and brass plate with a QMK-enabled PCB and MX compatible sliders.
Of course! I will have to make contact with NiZ and I think he will be open to some collaboration. I have received immensely good feedback from everyone and I think this will be a blast!
That would be amazing! Between their new cases and components and your PCBs we would be well on the way to custom capacitive keyboards.
Niz boards are also priced much more competitively than Topre, which would help achieve your stated price goal.
I just joined the Atom 68 Alu GB, too. Custom PCBs for the Atom 68 and the Micro 82 would be amazing!
I emailed Niz to ask about the possibility of a collaboration. The good news is that they had already been in contact with the OP yesterday, but the bad news is that it sounds like it this will not come to fruition at this time. They said they are going to keep talking though, so hopefully we will see more collaboration on custom Electrostatic Capacitive keyboards between these two parties in the future! I would love to support it.
...
Ex: Custom CNC alu funky case (let's say alice style) + topre pcs with BT&VIA + Niz 2019 sliders + well tuned Costar stabs = love
Thanks to Manofinterest for bringing me here. As a rubber dome peasant myself I will gladly support this. Have plan to buy a HHKB BT for work and travel but this PCB will change it haha.
- My personal preference is USB-C for everything, it is the way. But fully understand if it is not available in the round 1 due to cost and fitment etc.
- Don't care about USB hub really, bluetooth is a must for me
And my pipe-dream is an Alice-layout custom Topre board. Dream on :p :p :p
I’ll be waiting warmly for your work on Realforce R1 and FC660C.
If it is of any help, you are welcome to use the work from my usb3 controller, this is not very cheap but so far it is working great. You may be able to do a more cost effective solution using different chips and connectors.
https://geekhack.org/index.php?topic=93970.0
A bluetooth and USB 3.1 controller would be fantastic and collaborations with NIZ sounds promising :D
joric has done work with the nRF52 and HHKB for a more battery efficient bluetooth solution, but there is still some issues
Personally I don't see a need for this when I already have Hasu's BT controller in my non-BT boards but if I didn't already have those I would be very much interested in this.
Well...I literally just bought an hhkb Type-S on yahoo.jp and was about to pull the trigger on the hasu's controller. Think I may wait now...
1. I love the usb controller and would like it. If it's not available I won't cry though (ok maybe I will, I really love the controller...) .
2. QMK YES PLEASE
3. How much?
4. USB-C is cool but at this point it's my least used connector for customs.
Love the keeping the functionality of the USB controller. That's a nice touch.
Yes please to QMK/VIA
USB-C would definitely be nice.
Agree with having the USB ports included. Mostly don’t want 2 open holes in the case but the functionality is a nice bonus.Love the keeping the functionality of the USB controller. That's a nice touch.
Yes please to QMK/VIA
USB-C would definitely be nice.
I feel bad for leaving the USB controller unpopulated on the HHKB, it's a very useful feature.
Nice! Thank you for your feedback!
Awesome! Very excited about this!!Well...I literally just bought an hhkb Type-S on yahoo.jp and was about to pull the trigger on the hasu's controller. Think I may wait now...
1. I love the usb controller and would like it. If it's not available I won't cry though (ok maybe I will, I really love the controller...) .
2. QMK YES PLEASE
3. How much?
4. USB-C is cool but at this point it's my least used connector for customs.
The USB hub and AMK are a given. For now I think we are going to implement USB 2.0 and mini-B connector, but in the future we do have plans for USB3.1 .
As for the price, really difficult to tell, but my rough estimate is 60 USD for the first version.
So down for a project like this. Its going to open up a lot of topre customization that was sorely needed, for all the thockers in GH. I think having a basic version just to make sure everything is working is the best way to go about this.
1. I'm pretty sure every die hard hhkb owner would love a chance to be able to reprogram them, including myself.
2. The hub isn't really a big concern, but it sounds like you are pretty sure you will be able to include that.
3.Having bluetooth is also something secondary, but would love to see it in a later edition.
4. Also USB-C would be nice for future custom projects using this pcb, but shouldn't be something that makes into consideration for the first run of this board at all.
Hope this project gets rolling asap will be sure to pop my head back in another time to peek for updates.
Ah, this is great! I wish such a thorough documentation was available when I first started with Conundrum (which btw has evolved to a 40% planck with our lovely switch). It took me months to get everything tested and get QMK to play together, it would've been just soo much easier haha.
Feel free to hit me up if you'd like to talk - I'll be polishing everything QMK-wise as soon as I get pre-GB round of PCBs+cases (that I'm in the process of ordering) but I've already had success on bare PCB prototype.
thock thock thock
100% interested! 50-60 definitely sounds reasonable for the simple/base version.
In order of importance of features for me:
USB 3.1
USB Hub
Bluetooth
Good luck!
Will the pcb have RGB underglow? That would be awesome!
...However, since the idea is to get the know-how and technology to make Topre custom PCBs, I think we are definitely impelemtning something of that in the future...
...However, since the idea is to get the know-how and technology to make Topre custom PCBs, I think we are definitely impelemtning something of that in the future...
When you get round to those future plans I could do with a half-dozen replacement novatouch PCBs ;)
...However, since the idea is to get the know-how and technology to make Topre custom PCBs, I think we are definitely impelemtning something of that in the future...
When you get round to those future plans I could do with a half-dozen replacement novatouch PCBs ;)
...However, since the idea is to get the know-how and technology to make Topre custom PCBs, I think we are definitely impelemtning something of that in the future...
When you get round to those future plans I could do with a half-dozen replacement novatouch PCBs ;)
OMG. If you made a QMK Novatouch pcb I might actually die. plsdo
I don't think you quite understand the USB hub feature -- USB hub can be 2.0 or 3.1. The 2.0 variant supports lower speed devices and does not support power delivery, but is really cheaper. It's the stock HHKB feature. So USB 3.1 and USB hub are the "same feature". I think that by USB3.1 you meant USB type C maybe?
Thanks for your feedback though, it's appreciated!
You sir, I am interested!
I don't think you quite understand the USB hub feature -- USB hub can be 2.0 or 3.1. The 2.0 variant supports lower speed devices and does not support power delivery, but is really cheaper. It's the stock HHKB feature. So USB 3.1 and USB hub are the "same feature". I think that by USB3.1 you meant USB type C maybe?
Thanks for your feedback though, it's appreciated!
Ah sorry - you're right! In my head, I half assume most keyboards will be using the USB-C interface moving forward, so I was more focused on the 3.1/hub aspect. I think many people looking to upgrade the PCB would gladly consider paying a little bit more for a more future proof PCB - if you're intending to keep the USB hubs, 3.1 would be amazing if feasible.
Did the Stream announcement ever happen? Any other updates?
So our idea is to make a first operating version and, futurely, an "ultimate" version will all the big boi features. In this sense, I need to know if you guys agree with this or definitely want a first PCB with bluetooth or 3.1 USB or both.
So our idea is to make a first operating version and, futurely, an "ultimate" version will all the big boi features. In this sense, I need to know if you guys agree with this or definitely want a first PCB with bluetooth or 3.1 USB or both.
Great project!
I would definitely prefer an "ultimate" version over the base initially. I can't imagine being able to justify the cost and waste of displacing multiple PCBs. I'd be much more willing to pay extra to know any PCB I invest is the best version possible. USB 3.1, Bluetooth - those features would definitely bring the HHKB to endgame territory and would make it superior to the new Hybrids. If cost is a concern, in the grand scheme of things, I honestly believe that anyone interested in buying an aftermarket HHKB PCB wouldn't bat an eye over a $24++ difference - people gut $200++ Novatouches just for the sliders to mount their $100++ GMK keycap sets.
But perfect is the enemy of good!
I know you're working on a billion things - and projected timelines on this?
I love you
I'd be interested in a future version using USB-C over Mini USB, and maybe Bluetooth LE. I would want to use it in the norbauer hhkb case whenever that gets around to being released. Unless there's some reason to switch from a stock HHKB PCB to this (I already have a Hasu BT controller) I don't really see a reason for myself to get this one yet. The big things missing if this gets made would mainly be sliders and springs, I think. BKE domes at least can serve as domes for now, and some of the previous aluminum Topre cases had the housings built-in to the top of the case.
Look forward to seeing updates.
I'd be interested in a future version using USB-C over Mini USB, and maybe Bluetooth LE. I would want to use it in the norbauer hhkb case whenever that gets around to being released. Unless there's some reason to switch from a stock HHKB PCB to this (I already have a Hasu BT controller) I don't really see a reason for myself to get this one yet. The big things missing if this gets made would mainly be sliders and springs, I think. BKE domes at least can serve as domes for now, and some of the previous aluminum Topre cases had the housings built-in to the top of the case.
Look forward to seeing updates.
Indeed, I think that for people that already have the Hasu controller this PCB is redundant. The idea here is not so much make a PCB for HHKB per se, but acquire know-how and methods to build:
- Custom Topre PCBs;
- Integrate USB hubs in a unified daughterboard;
- Making USB 3.1 hubs;
- Convert USB mini PCBs into USB type C;
As for the hardware (domes, housings, sliders) I am hoping I can work with some makers to make those a reality. Unfortunately, for now, all we can have are replacement PCBs for people that already have those pieces.
I'd be interested in a future version using USB-C over Mini USB, and maybe Bluetooth LE. I would want to use it in the norbauer hhkb case whenever that gets around to being released. Unless there's some reason to switch from a stock HHKB PCB to this (I already have a Hasu BT controller) I don't really see a reason for myself to get this one yet. The big things missing if this gets made would mainly be sliders and springs, I think. BKE domes at least can serve as domes for now, and some of the previous aluminum Topre cases had the housings built-in to the top of the case.
Look forward to seeing updates.
Indeed, I think that for people that already have the Hasu controller this PCB is redundant. The idea here is not so much make a PCB for HHKB per se, but acquire know-how and methods to build:
- Custom Topre PCBs;
- Integrate USB hubs in a unified daughterboard;
- Making USB 3.1 hubs;
- Convert USB mini PCBs into USB type C;
As for the hardware (domes, housings, sliders) I am hoping I can work with some makers to make those a reality. Unfortunately, for now, all we can have are replacement PCBs for people that already have those pieces.
We can provide domes, housings, sliders, and any other parts if needed.
I bought from them the parts before ( also a full Atom66) and used them in a build where i experimented with the sensing, but never finished the project (tried to make a columnar staggered as that is my area of interest).I'd be interested in a future version using USB-C over Mini USB, and maybe Bluetooth LE. I would want to use it in the norbauer hhkb case whenever that gets around to being released. Unless there's some reason to switch from a stock HHKB PCB to this (I already have a Hasu BT controller) I don't really see a reason for myself to get this one yet. The big things missing if this gets made would mainly be sliders and springs, I think. BKE domes at least can serve as domes for now, and some of the previous aluminum Topre cases had the housings built-in to the top of the case.
Look forward to seeing updates.
Indeed, I think that for people that already have the Hasu controller this PCB is redundant. The idea here is not so much make a PCB for HHKB per se, but acquire know-how and methods to build:
- Custom Topre PCBs;
- Integrate USB hubs in a unified daughterboard;
- Making USB 3.1 hubs;
- Convert USB mini PCBs into USB type C;
As for the hardware (domes, housings, sliders) I am hoping I can work with some makers to make those a reality. Unfortunately, for now, all we can have are replacement PCBs for people that already have those pieces.
We can provide domes, housings, sliders, and any other parts if needed.
Oh, main point wanted to tell is, that the 2019 revision of the switch parts is superb!I bought from them the parts before ( also a full Atom66) and used them in a build where i experimented with the sensing, but never finished the project (tried to make a columnar staggered as that is my area of interest).I'd be interested in a future version using USB-C over Mini USB, and maybe Bluetooth LE. I would want to use it in the norbauer hhkb case whenever that gets around to being released. Unless there's some reason to switch from a stock HHKB PCB to this (I already have a Hasu BT controller) I don't really see a reason for myself to get this one yet. The big things missing if this gets made would mainly be sliders and springs, I think. BKE domes at least can serve as domes for now, and some of the previous aluminum Topre cases had the housings built-in to the top of the case.
Look forward to seeing updates.
Indeed, I think that for people that already have the Hasu controller this PCB is redundant. The idea here is not so much make a PCB for HHKB per se, but acquire know-how and methods to build:
- Custom Topre PCBs;
- Integrate USB hubs in a unified daughterboard;
- Making USB 3.1 hubs;
- Convert USB mini PCBs into USB type C;
As for the hardware (domes, housings, sliders) I am hoping I can work with some makers to make those a reality. Unfortunately, for now, all we can have are replacement PCBs for people that already have those pieces.
We can provide domes, housings, sliders, and any other parts if needed.
Sent from my iPhone using Tapatalk
I wasn’t able to watch the stream. What’s the latest developments on this and how it will proceed?
Keep up the good work! :thumb:I wasn’t able to watch the stream. What’s the latest developments on this and how it will proceed?Ok so here goes a summary of what I the changes I announced and partially designed last stream:
I tried to use galvanic isolators once before. They claim to offer high speed isolation.... but even with absurd levels of PCB designing to isolate the noice, they scream in an EMC chamber. No product can realistically use those and pass things like FCC and CE compliance. It was so bad my wireless mouse in my lab would lose connection when they were transmitting data
Keep up the good work! :thumb:I wasn’t able to watch the stream. What’s the latest developments on this and how it will proceed?Ok so here goes a summary of what I the changes I announced and partially designed last stream:
Very exciting!
Wait....will this work for the JIS version too (did I miss the announcement of this somewhere else in the thread)?
Did not expect it to be so detailed and technical. I know some of those words. Haha keyboard go brrrrrI wasn’t able to watch the stream. What’s the latest developments on this and how it will proceed?
Ok so here goes a summary of what I the changes I announced and partially designed last stream:
- Sensing circuit operation amplifier. In the first version and the documentation, the operation amplifier I am using is the TL081, which is a fast enough and cheap, reliable opamp that is pretty much standard. It does however present some issues:
- It consumes too much current, approximately 13mA per unit. Considering that I need to imbue each sensing circuit for every single switch, meaning 60-100 of them, this circuit with TL081 is not scalable -- 100 of them would require 1.3A which is well above the USB2.0 current delivery capabilities.
- It is not small enough. Each switch gives me a 19.05x19.05mm (approximately 363 mm2) space to work with. In that space I have to somehow fit an AM modulator circuit, an AM de-modulator circuit, a rectifier circuit, a filter circuit, a comparator circuit and an optocoupler. As you can guess that is a pretty limiting space and I need the smallest components I can use. The most space-efficient way to deploy TL081 is using the quad-package TL084 in SOIC-14 packages (https://lcsc.com/product-detail/FET-InputAmplifiers_Texas-Instruments_TL084BCDR_Texas-Instruments-Texas-Instruments-TL084BCDR_C352983.html), which are each almost a centimeter wide. TL084 is also available in TSSOP-8 (https://lcsc.com/product-detail/FET-InputAmplifiers_Texas-Instruments_TL084CPWR_Texas-Instruments-Texas-Instruments-TL084CPWR_C85344.html), but it is considerably more expensive. Remember that we need 100+ units, so every cent in opamp cost means a dollar more that I had to sacrifice somewhere on the PCB or raising the end price ;
- It requires ±15V. Since the TL081 has a very bad peak output voltage swing, I used it with a ±15V power source, allowing for ±10V output swing with good linearity. Problem now being that ±15V have to be implemented on the PCB by using the USB 5V supply, probably through switched-mode converters and some power filters which honestly, in a PCB like this, are just too much.
In order to solve this problem I tried using the TL061, which is an opamp from the same family as the TL081 but with FET input and optimized for low- and micro-power applications, solving the current draw issue. Nevertheless TL061 was way too slow and could not keep up with the circuit, so in a second version I had to relax the 100µs settling time requirement and use 200µs, leaving me quite uncomfortable because that has the serious risk of affecting the activation circuit perceived latency. Still, 100µs more in the sensing time is not that big of a deal I guess: the major problem really is the space, as TL064s are not available in TSSOP-14 (https://lcsc.com/search?q=TL064) package (at least not LCSC) and using SOIC-14 is something I want to avoid. So now for the third version I am using LM324, which is available in a ridiculously small QFN-16 package (https://lcsc.com/product-detail/General-Purpose-Amplifiers_STMicroelectronics_LM324QT_LM324QT_C183029.html) and is even cheaper than both TL084 and TL064.- Sensing circuit diodes. I was using the 1N4148W, a silicon diode which is cheap, available, small, and at this point pretty much a standard among keyboard makers. There are two issues with this diode.
- Space. 1N4148 is not small enough and there are better, smaller alternatives.
- Performance. Being a small-signal sillicon diode, 1N4148 is not fast enough to keep up with 100kHz+ large swing wave signals.
So because of this I decided to use BAT54:
- It comes in a dual-package SOD-523 (https://lcsc.com/product-detail/Schottky-Barrier-Diodes-SBD_Diodes-Incorporated-BAT54STQ-7-F_C459838.html), meaning it’s both less than half the size and half the price;
- It also is a Schottky diode, so it is faster, more reliable, has better parasitic performance and less forward voltage, despite that does not mattering that much due to the precision-diode rectifier feedback topology used;
- Wien bridge oscillator topology. To generate the 100kHz sine wave, I was using a Bubba Oscillator topology, which is a very common quadrature-phase signal rotation oscillator. The LM324 was however unable to work with this topology because its output impedance was just too high at 100kHz, making it impossible to reliably set the circuit to generate a 100kHz sine wave with a precision that I am comfortable with -- with every design I have to consider the components tolerances, because oscillator circuits tend to be highly sensible to that. To replace the Bubba topology it I considered a myriad of other oscillator topologies, ranging from Colpitts crystal oscillators to solid-state ones. I ended up deciding to just use a good, albeit difficult to work with, oscillator topology known as a Wien Bridge oscillator. It still uses an operational amplifier, but this time around I got one suited for the job, the TLVx316, which is better than LM324 in every aspect but way more expensive. Regardless, since I need only one for the oscillator and I already saved some budget by adopting the LM324, this is acceptable. The dual-package TLV2316 allows me to integrate both the oscillator and a third order Butterworth-based filter topology to further clean up the output wave. I was able to produce (in the simulation, at least) a 0.5% THD sine wave, which is very good.
All in all I'm pretty happy with the direction that this project is taking and I feel vey proud I was able to make it this far. Everything works like expected and I will probably be polishing things out and starting the PCB shortly.
Did not expect it to be so detailed and technical. I know some of those words. Haha keyboard go brrrrrI wasn’t able to watch the stream. What’s the latest developments on this and how it will proceed?
Ok so here goes a summary of what I the changes I announced and partially designed last stream:
- Sensing circuit operation amplifier. In the first version and the documentation, the operation amplifier I am using is the TL081, which is a fast enough and cheap, reliable opamp that is pretty much standard. It does however present some issues:
- It consumes too much current, approximately 13mA per unit. Considering that I need to imbue each sensing circuit for every single switch, meaning 60-100 of them, this circuit with TL081 is not scalable -- 100 of them would require 1.3A which is well above the USB2.0 current delivery capabilities.
- It is not small enough. Each switch gives me a 19.05x19.05mm (approximately 363 mm2) space to work with. In that space I have to somehow fit an AM modulator circuit, an AM de-modulator circuit, a rectifier circuit, a filter circuit, a comparator circuit and an optocoupler. As you can guess that is a pretty limiting space and I need the smallest components I can use. The most space-efficient way to deploy TL081 is using the quad-package TL084 in SOIC-14 packages (https://lcsc.com/product-detail/FET-InputAmplifiers_Texas-Instruments_TL084BCDR_Texas-Instruments-Texas-Instruments-TL084BCDR_C352983.html), which are each almost a centimeter wide. TL084 is also available in TSSOP-8 (https://lcsc.com/product-detail/FET-InputAmplifiers_Texas-Instruments_TL084CPWR_Texas-Instruments-Texas-Instruments-TL084CPWR_C85344.html), but it is considerably more expensive. Remember that we need 100+ units, so every cent in opamp cost means a dollar more that I had to sacrifice somewhere on the PCB or raising the end price ;
- It requires ±15V. Since the TL081 has a very bad peak output voltage swing, I used it with a ±15V power source, allowing for ±10V output swing with good linearity. Problem now being that ±15V have to be implemented on the PCB by using the USB 5V supply, probably through switched-mode converters and some power filters which honestly, in a PCB like this, are just too much.
In order to solve this problem I tried using the TL061, which is an opamp from the same family as the TL081 but with FET input and optimized for low- and micro-power applications, solving the current draw issue. Nevertheless TL061 was way too slow and could not keep up with the circuit, so in a second version I had to relax the 100µs settling time requirement and use 200µs, leaving me quite uncomfortable because that has the serious risk of affecting the activation circuit perceived latency. Still, 100µs more in the sensing time is not that big of a deal I guess: the major problem really is the space, as TL064s are not available in TSSOP-14 (https://lcsc.com/search?q=TL064) package (at least not LCSC) and using SOIC-14 is something I want to avoid. So now for the third version I am using LM324, which is available in a ridiculously small QFN-16 package (https://lcsc.com/product-detail/General-Purpose-Amplifiers_STMicroelectronics_LM324QT_LM324QT_C183029.html) and is even cheaper than both TL084 and TL064.- Sensing circuit diodes. I was using the 1N4148W, a silicon diode which is cheap, available, small, and at this point pretty much a standard among keyboard makers. There are two issues with this diode.
- Space. 1N4148 is not small enough and there are better, smaller alternatives.
- Performance. Being a small-signal sillicon diode, 1N4148 is not fast enough to keep up with 100kHz+ large swing wave signals.
So because of this I decided to use BAT54:
- It comes in a dual-package SOD-523 (https://lcsc.com/product-detail/Schottky-Barrier-Diodes-SBD_Diodes-Incorporated-BAT54STQ-7-F_C459838.html), meaning it’s both less than half the size and half the price;
- It also is a Schottky diode, so it is faster, more reliable, has better parasitic performance and less forward voltage, despite that does not mattering that much due to the precision-diode rectifier feedback topology used;
- Wien bridge oscillator topology. To generate the 100kHz sine wave, I was using a Bubba Oscillator topology, which is a very common quadrature-phase signal rotation oscillator. The LM324 was however unable to work with this topology because its output impedance was just too high at 100kHz, making it impossible to reliably set the circuit to generate a 100kHz sine wave with a precision that I am comfortable with -- with every design I have to consider the components tolerances, because oscillator circuits tend to be highly sensible to that. To replace the Bubba topology it I considered a myriad of other oscillator topologies, ranging from Colpitts crystal oscillators to solid-state ones. I ended up deciding to just use a good, albeit difficult to work with, oscillator topology known as a Wien Bridge oscillator. It still uses an operational amplifier, but this time around I got one suited for the job, the TLVx316, which is better than LM324 in every aspect but way more expensive. Regardless, since I need only one for the oscillator and I already saved some budget by adopting the LM324, this is acceptable. The dual-package TLV2316 allows me to integrate both the oscillator and a third order Butterworth-based filter topology to further clean up the output wave. I was able to produce (in the simulation, at least) a 0.5% THD sine wave, which is very good.
All in all I'm pretty happy with the direction that this project is taking and I feel vey proud I was able to make it this far. Everything works like expected and I will probably be polishing things out and starting the PCB shortly.
I can't believe this is real this is awesome.
please take my money, but first tell me how many monie this is.. thank
I’m not sure if I already commented in this post but I’m in!
I own a FC660C, a Novatouch, an HHKB Pro 2 and soon a Niz Atom, so I’m super interested.
so... Topre alice when?
(https://media1.giphy.com/media/sDcfxFDozb3bO/giphy.gif)so... Topre alice when?
This is endgame. Im currently talking to xondat about possibly making something like this
so... Topre alice when?
This is endgame. Im currently talking to xondat about possibly making something like this
Never done this before and generally find the replies like "in" and "can't wait" kind of pointless, but since this seems to have lost some momentum I wanted to say that I have a yang controller, tmk compatible, and still would buy this asap. Easily my most anticipated IC. The potential for custom capacitive pcb's and eventually cases is actual innovation and change in a sea of "gAsKEt mOuNT 65%/75%" Can't wait for a good wkl or standard 60 capacitive with hhkb sliders and domes.
This has absolutely not lost momentum; I'm still actively designing for and incorporating into it. The issue is that I recently had some ****ty life situations happen and I'm just detoxing from the hobby for a while.
There are also behind-the-scenes stuff I can't talk about right now but in a couple months you guys will hear from this again!
The modularity of the circuitry is awesome. With your open-source track record, Gondo, I would hope this would be made public. That, combined with the affordable and MX-compatible "switches" from NiZ makes me very hopeful and excited for the future of custom EC boards. Good luck with this project, hope it's the stepping stone we need to make quality Topre customs not only accessible but existant too!
Sent from my SM-G930V using Tapatalk
Sounds amazing.The modularity of the circuitry is awesome. With your open-source track record, Gondo, I would hope this would be made public. That, combined with the affordable and MX-compatible "switches" from NiZ makes me very hopeful and excited for the future of custom EC boards. Good luck with this project, hope it's the stepping stone we need to make quality Topre customs not only accessible but existant too!
Sent from my SM-G930V using Tapatalk
Everything will be open sourced, and I'm already talking to NiZ to make everything compatible with his stuff. The idea is to make it so that makers can offer a single plate that supports MX and "Faraday style" electrocapacitives
Is this project still active? I am desperate to be able to use QMK with my HHKB
so... Topre alice when?
This is endgame. Im currently talking to xondat about possibly making something like this
Is this dead?
Moreuses a dedicated chip for capacitance sensing.Is this dead?
No. The first version of the circuit was too complicated and not very scalable, so I am desgning a second version that
I will give more information as I have more results to show.
Glad to hear this project is still going, best of luck with the new strategy
Do you know about this? It's sold out anyway:
https://www.jels.au/products/gb-usb-c-hhkb-controller?variant=41738239377562