Is it just me or does it seem wrong for the Ellipse F62/F77 to leave the factory with Micro USB?

I know the keyboard is (+/-) 3 years in development but doesn’t it seem wrong that it’s not leaving the factory with USB-C? I think it would be much more appropriate (durable) and future proof connector than Micro USB.


Lots of new boards are still using mini/micro, unfortunately. Hopefully the USB C keyboard trend continues to grow.

1 Like

You have to open the case up to disconnect the cable on these boards, so durability matters less

My bad! I take it there is a strain relief system in place?

yup! you can see at 2:39 mark when ellipse opens the case up. woulda been cool to see externally detachable usb-c though I agree.

1 Like

I, for one, will not be buying any micro-usb peripherals in 2018

It has more to do with the controller being used than with the keyboards themselves. Namely, Ellipse is using one of xwhatsit’s IBM Capsense USB controllers. They are tried and true, but unfortunately, since they were created a couple of years ago, when USB-C wasn’t yet all that popular, their native interface is mini/micro USB.

The controller is open source, though, so if someone can update it to use USB-C, I’m sure Ellipse would be more than happy to consider manufacturing it instead of the micro USB variant.

I don’t think it’s particuarly wrong for anything to not have USB-C, especially in the mechanical keyboard market. Hell, up until a few years ago, and even still to this day, many of them are using mini usb, not even Micro. I mean, I guess it makes it more difficult to have only one cable to your compuer to change out boards, but for me, it’s always been the cable goes with the keyboard. If I can use the same cable, fine, but if each of my boards has their own cable, I never have to hunt for one to use.

Also, as @konstantin said, it’s a lot dependent on controllers and their form factor also.

There was a short discussion about this on the Deskthority thread around a year ago.

To quote Ellipse:

  1. Micro-USB will still be in use for a very long time.
  2. USB-C on the other hand is new and not available on every machine.
  3. In the future, because Micro-USB is 100% compatible with USB-C, a simple adapter will suffice.

I agree. Even a small change like USB C would increase costs by thousands across 1,000 keyboards. Given tooling costs it is not feasible to make some controllers one way and others another way. Instead I focused on using the highest quality materials: e.g. a high quality zinc alloy for the cases (even IBM used something like “pot metal” for some Model F 4704 cases)

This is also relevant. It seems that there will be support for USB-C if you decide to add it yourself (which probably entails getting a controller that supports USB-C or using an adapter inside the case).

In the end, if someone comes out with USB-C Capsense controllers in the future, it should be pretty easy to put them in your F62/F77 yourself (the thing is designed to be serviceable by anyone, after all).

Thank you for further educating me. My intention wasn’t to spread FUD.

The connector seemed (to me) to be an unusual choice with the mission of the project, which is to have a keyboard that lasts for decades; even if that means making your chiropractor wealthy should you pick it up the wrong way.

yep konstantinub1d - I still have the same thoughts

dante the controller was designed with micro-USB as the connector is strong enough and also low enough cost (compliant USB-C cables would also be additional cost).

And yes the good thing is that the PCB is open source, so with a soldering iron and PCB design knowledge you could update the xwhatsit in case the ATMEGA32U2 or micro-USB connector become hard to find in the future.

I think the more long-term quest is to find alternative options for the ATMEGA chip powering the xwhatsit which is more likely to disappear in a few decades.

1 Like

A few designers have been jumping ship to ARM. If someone wants to put together an arm-based USB C XWhatsit I’d certainly pick one up!

But can you see where the confusion comes in? ± 8 pounds of Zinc and brand new 1 piece keycaps but then worry about the cost of the connector is just bizarre to me but whatever.

It’s about balance. All products (outside maybe some of the crazy expensive stuff) have compromises. And by compromise, I don’t necessarily mean sacrificing durability, or quality, I mean does Part A need to be used at twice the cost, when Part B will still be functional 50 years from now. Part B, at half the cost, can still be over-engineered for the purpose it needs to serve. Not saying this is what happened, but it’s a consideration all designers/engineers need to make.

@Ellipse could have designed a whole new controller, at a higher cost (a la development costs, parts sourcing, etc), or used the existing and inexpensive xwhatsit controller. I think the correct decision was made, since it brings us the boards sooner cough, and reduced the sheer number of items @Ellipse had to design, test, and bring to production. You have to draw a line somewhere.

Having said that, I have no idea how much time, effort, and cost would be involved in a new controller. The capacitive design used in Model Fs and Ms is pretty well understood, so ‘anybody’ with the working knowledge of designing a controller could create a new one, as has been mentioned. The xwhatsit does pretty much everything you’d want though, so aside from wanting to update it for updating’s sake, there’s no reason to not use it.