I am the proud owner of a Framework 13 with a “AMD Ryzen AI 7 350 w/ Radeon 860M” (lshw). I am running Debian Linux on that machine.
Among the expansion cards I ordered, I bought a SSD-expansion card with the idea to use that as an encrypted data or OS disk for work-related tasks (we have an official bring-your-own-device rule book) and otherwise to use a separate OS for private stuff, all on the same device but with the ability to remove the work-data from my private device when needed. From the onset I assumed that speed and reliability of this plan were a gamble, but one I was willing to take on.
One issue I did not expect, however, was a weird, I think, power-related issue: When I mount the SSD expansion card, for example, in the back-right expansion slot and then attach a USB-A device to the front-right slot (with a mounted USB-A expansion card), this sometimes causes the entire USB-controller for the entire right-side to reset… crash? I cannot tell… At that point, a reboot does not even fix this, only re-insertion of the expansion cards fixes the problem…
Obviously, this is a problem for an encrypted data drive - data corruption is luring. For the device I connect to the USB-A port: Nothing extraordinarily “scary” I would say - mostly just my bluetooth usb-mouse to charge it up. It happens regularly enough that am too anxious to use the side the SSD is connected at all at this point. My YubiKey and low-power devices seem to not cause this problem to manifest though.
I wonder: Is this a known kind of issue with frameworks? I did not see any issues about this kind of behavior, though I do not see many posts from people with SSD expansion cards at all either…
It can supply 1.5A at 5V, 7.5W of power to the usb device. Hardly enough to charge it, and if it tries to use more current than it usb pd negotiated for, it might reset the laptop usb chip.
So, this reset is probably what you are seeing.
General rule, do not charge devices from a FW13/16 usb ports. Use wall chargers instead.
Ha! Thanks for highlighting that very section for me in the docs! But that settles it then.
A bit of a pity because I am not actually trying to charge my mouse, a CoolerMaster MM712. Instead it has a USB-mode where it acts like a USB mouse but then also tries to charge itself at the same time. Clearly it seems to be more down to the mouse not adhering to specs in this case. I can work with this.
Again, thanks for the help!
And just for something “funny”: I actually bookmarked that very page in my personal notes and cut off that important info you linked me:
I did a bit of searching about bluetooth mice and how much they need to charge up.
It seems a mouse might only need 2.5W to charge itself. The MM712 datasheet does not mention how much power is needed at all.
2.5W is below the 7.5W provided by the FW13 laptop use ports.
Even if, via the USB-A, it only supplier 900mA. That is still 4.5 Watts, so above 2.5W.
So, this should not be causing you problems.
I suggest you raise a support ticket with FW via their web page form.
I think OP’s problem is more likely to be a glitch. The USB-A ports in FL13 supports up to 1.5A, 0.9A is only on FL16’s port 3 and 6, not on FL13.
It’s very unlikely for a USB mouse to draw more than 1.5A as higher charging current will damage the battery, with 100mAh to 400mAh on most if not all wireless keyboards and mice, the charging current would be less than 500mA
SteelSeries APEX 3 TKL keyboard (with RGB, link). No reset.
With the hub there are some other tests I could do, like connecting the problematic devices to the hub. I did not try that yet, but actually it seems that even connecting things to the hub does not reset the USB port, which to me seems to support the idea that there is an on-connect issue on the framework-side of things. Anyway, some structured testing for this is needed and I do not really have time for that right now, at this very moment.
I think I will open a support issue on the framework-side now and see if they have some maybe obvious informed advice for me. If I have more info, or collected more data myself I will report back!
About the hub, connecting the power supply to the hub first, then connect the hub to the computer, and connecting the hub to the computer first, then connect the power supply to the hub, may yield different results.
