I’m currently developing a card with two USB-C ports but for now only with USB 2.0.
I don’t have that much experience with high speed designs so I’ll first try to get it running at all before dealing with the advanced stuff.
I’m not sure if everything fits on that relatively small PCB area since there has to be all kind of stuff done to get the USB-C part correctly.
Also two ports fit next to each other but not the kind that is used in the original design so I’m not entirely sure if it fits within the height restrictions.
I’ll post it here when I have something substantial to show .
Thank you for updating us! Good luck
Are there any other preexisting products we could tear down and re-house in a new case that would work, even if it doesn’t fit exactly? It could be good to make it work and then make it fit any way we can, or even adhere to the expanded size of the ethernet card and take advantage of the expanded height and length
I just realized that the dual USB-C is useless unless you can actually plug in two USB-C stuff at the same time.
and the easiest way to do something like that would be to have a lead to the device from the port but at that point you’d just be living The Dongle Life™ in a slightly different way
I was playing around a bit recently and did roughly place things, there aren’t even all ICs placed that are needed, a few transistors plus an amount of passives are still missing. When I got them all implemented I’ll try to get everything on one side to avoid double sided assembly. Also I’m not sure if I can connect everything using a 4-layer board or if I have to use (more expensive) 6-layers.
The voltage converter chip is a dual rail chip, so it could handle both voltages required for a future USB 3 solution, until now it’s just USB 2.0. USB 2.0 could be implemented a little simpler (e.g. using two physical ports instead of four) but I’d like to try the concepts needed for a USB 3 solution.
It’s quite elaborate to implement USB-C according to the specs, the three ICs underneath the ports are basically just handling the ‘-C’-things.
The USB-C ports are spaced 15.6mm, a tiny little more than on my MacBook.
There were a few (early) concepts in this thread, did anyone actually complete one or did at least get a working prototype and can share his/her experiences?
FYI, JLCPCB currently (and since january, so it’ll likely stay for a few months longer) only charges $2 for under 50x50mm up to 8 layers.
USB-C connectors need to be mid mount otherwise they will not fit properly.
I can understand that. However, from what I see in that PCB, there’s still a lot of space available to put more components, even space between the USB-C plugs for the smallest chips. Still, I’d be quite tempted with having the more expensive 6-layered option and have USB3 included, even better if using 1x1 but no PD. If you need and can fit an 8-layer into the volume, I’m welcome to it too!
I’d have usages such as connecting an HDD and a mouse to it or maybe a mouse + keyboard combo (which I’d use as a makeshift dock) or just 2 external drives.
I wouldn’t think about charging the laptop through those.
if those ports could be made to be spaced EXACTLY as the MacBook ports that would open a huge range of accessories, from the dual USB-C only MacBook expansion hubs to ones involving Thunderbolt 3 where typically only one plug is actually used and requires TB3, the other one is a stabilizer or used just for data or power delivery.
They could also be spaced exactly, I basically just did the largest spacing without violating minimum distances to the board edge. I’m pretty sure tho that no thunderbolt solution fits in there. On the other hand what probably would work is a dummy connector and a working pass-through like the original USB-C-card. But I don’t have such a dock or information how they are wired internally.
Yeah probably not.
Honestly I think a 1 usb2 and 1 usb3 config without pd is probably the most realistic.
No it wouldn’t. It’s not just a plug distance issue. It’s also the USB capabilities missing that would cause issues. It would make people frustrated if they successfully plug in. (maybe they also would be if they can’t plug in?)
I think it would never work. Better space so 2 plugs plug well (even if the cables are slightly fatter) than spacing exactly like a macbook due to the accessories made for such spacing.
First it would require TB on, at least, one port… I think that’s irrealistic. I don’t even know if they are misconnected (even though it fits) could cause issues because the “language” of the hardware on how it seems to fit.
@Brunoais that’s a fair point, given that the dock probably uses every feature like PD and alt-mode like display as well as TB, it most likely won’t even connect properly.
Hm, you convinced me to go straight so USB 3 .
Spent like half of today with playing around and optimising the schematics. The routing of the USB-lines is complete (not finally but to get an idea of the space needed. Power is at like 0% (the capacitors scattered on the top corner are part of that and have to be placed (and there have to be more of them). Also the programming lines of the hub-IC has to be done entirely (and I have to figure out what exactly to do before ).
I’m using this hub-IC: USB5744 | Microchip Technology
It’s the smallest one I could find. Also I’m implementing the USB-C ports as four seperate ports (which can be done but is probably not best practice).
The major parts alone are currently around 30€, plus the PCB, plus manufacturing (which I’m not sure how expensive that could become), so I’m not sure if it finally is economically worth it for a second USB-port (but I like the project).
Thank you !
The top is already cluttered but there seems to exist a lot of space available on the bottom side near and around the male plug. Would they fit there?
Also, I like how you made use of the space between the plugs. That is engineering! Maybe you can rearrange better the bottom part. That side seems to be less optimized by comparison.
That is an interesting approach… I’m kinda on board with it . Implementing each USB-C as 2 USB is a very interesting approach and it might just be our ticket for the success while making the plug reversible.
I’m on board until 40€. I only know if this was manufactured in batches of 1000, given the parts are around 30€ + the PCB + machine manufacturing price would stay very close to 40€ (except in the most expensive countries such as Germany or Switzerland).
This is pretty unrealistic considering manufacturing is expensive, as well as shipping, tax, and lastly and most importantly profit…
I said this before, YOU NEED A MID MOUNTED RECEPTACLE.
The usb2 + usb3 (usb2 from a hub and 3 straight through) would be a lot cheaper and there are small 2 port hub chips that need pretty much no extra components except for a crystal.
Are you sure about that?
There’s no way (in my experience so far) in terms of available board space and distance between the sockets using a mid-mounted receptacle. In my opinion the only way getting some additional usable height is offsetting the plug using a daughter board with a board-to-board-connector on top of the “main“-PCB.
It could work (with roughly placing a receptacle I had laying around) very tightly (there’s zero-ish space between the top of the socket and the top of the module) but I’ll examine that with the prototype once it’s ready .
Mh I doubt that would have a significant impact, the hub-IC itself is like 5€ vs. 8€, the plug and sockets are not overly cheap, the ESD-protection, power controller and voltage regulator will be necessary anyways (if you want to make it somehow according to specs) and the price of the passives is negligible.
I doubt that that’s a realistic end-price (also we’re talking about the electronics only, there’s no enclosure so far). Prices drop when you’re getting in 1000 or 10000 regions (which I also doubt it will) but in the end I think it’ll be a fun project but won’t make much sense economically.