I'm currently in the process of designing an #ExpansionCard for my new #FrameworkLaptop.
I need the #Unifying #dongle (lower latency), but don't want it sticking out and losing a whole slot.
So I'm going to throw a USB hub and the peeled dongle on a PCB and hope it all works out :D
As a further experiment, I routed the USB A SS lines past the hub (USB 2 D+/D- connected to port 1 of the hub).
Not sure if this will play nice with USB drivers, but it's worth a try.
@LeoDJ drivers shouldn't care, as USB 2 and 3 are completely separate busses :3
also, might be handy here somewhere, do you know you are allowed to switch polarity of the SS pairs to simplify routing? E.g. SSTX+ to SSTX- and vice-versa :)
@littlefox
I think it should work out too.
I only know that I crashed an HP BIOS with something like this at my old work, because of the USB 2 and 3 parts of the device not being at the same place topologically. (Because of a cursed USB fiber extension)
Apparently the BIOS was confused when the device upgraded to USB 3 but was now at e.g. 1-1.1 instead of 1-1.2.1 and froze :D
Then again, a VL671 was involved too, soo idk ^^
Regarding swapping +/-: Ohh, I didn't know that, that's neat, thx!
@LeoDJ but .. USB devices don't upgrade to 3.0, they try to connect on the 3+ bus and if they can't find a host, they try the 2- bus instead o.o
Maybe the controller was confused about something connected on both 3+ and 2- on what was the same port for it?
@littlefox
It's been a few years already, so my memory is a bit hazy.
I think in this instance the device's USB 3 stack initialized some time after the USB 2 stack, leading to the device only being reachable via USB 2 at first.
(And "upgrading" to 3 a few hundred ms or so later)
But I'm unsure of the exact details rn ^^"
Maybe some non-floating USB 3 lines could've been involved too, but I think it was mainly because of the topology thingy.
LeoDJ
Saphira Lohikäärmekettunen 