[VOTE] Framework Desktop Rear USB-C ports power supply to usb devices

Framework Desktop
1

I am interested to find out what, in general, would people expect the rear usb-c (usb4) ports on the Framework Desktop should be able to provide.
Also, note that on this FW page provided no useful information regarding power on usb-c ports.

So, for lack of any information from FW on the matter, what would most users expect?
I asked copilot/chat-gpt and it said the default for usb-c is 15W, but with PD it can go much higher.

For the FW Desktop, rear USB-4 ports.
First poll is for what people would normally expect based on other laptops/desktops out there from other manufactures:

  • 7.5 W
  • 15 W
  • 20 - 30 W
  • 100 W
  • 240 W
0 voters

Second, would users expect both usb4 ports to have the same capabilities. I.e. both ports deliver the same amount of power at the same time.

  • Both ports deliver the same power at the same time
  • One port wins, the other port received less power.
0 voters

Third poll is for what people would wish for in future:

  • 7.5 W
  • 15 W
  • 20 - 30 W
  • 100 W
  • 240 W
0 voters
2

Suggestion: add another poll.

What you wish the computer do if the total energy consumption exceeds PSU power?

  1. Cut or reduce USB PD output
  2. Reduce CPU/GPU clock speed
1 Like
3

What’s even the point of the massively overkill framework desktop psu then?

1 Like
4

I’ve tested this morning the ports with my chargerlab tester and for those who are wondering, the only PD power supporded is 5V@1.5A or 7W. It is just enough to power a USB C portable monitor at low brightness.

Front USBC does not support PD protocol.

5

This sounds fairly limiting. HP Z2 mini G1a’s side usb ports can delivery 10 to 15W. Not sure about its main USB4 ports but probably higher?

I have an AMD2364 USB4 ssd enclosure. Technically some SSDs can draw over 7W under load. It feels unsafe to use…

6

Which BIOS did you test with?
There is a newer bios that mentions something about enabling 3A for the rear ports. Would you be able to test that?

7

I tested with BIOS 03.02, the latest version then.

Now with 03.03 PD is still 7W but I saw in the release notes that it’s a known issue and 3A support will be available in upcoming versions. I’ll test it again to confirm then!

Edit: here is the GitHub issues, they are working on the fix.

8

I see, they removed this line from the release notes for 3.03

  • 6) Enabled 3A PD profiles on the back ports.
9

It’s not even difficult, since 5V1.5A and 5V3A are both 5V. All they have to do is to simply not cut the output when the current draw is bigger than 1.5A

10

It is a bit more difficult than just changing the overcurrent limits.
They need to start with the 3A Rp, and not the 1.5A Rp.
Then they need to advertise 3A being available in the PD negotiation, and then accept requests to draw 3A. Then set the overcurrent limits to 3A.
Then make sure all this api is reachable, and configurable over I2C from the EC firmware. That probably means debugging the PD chip firmware. Then implement it in the EC firmware.

The FW desktop uses the CYPD6228 for the rear USB4 ports.
That chip supports 3A on both outputs. That means 15W is available on both outputs. So, I am fairly confident, once FW implement it, that the FW desktop will eventually support 15W on each rear usb4 port.

11

Yeah not really seeing why they do the whole only 1 3A port complexity on the desktop, on the laptop there are some use cases but I don’t even think it makes sense there but at least I can somewhat understand why they did it.