Idle power differences between expansion slots

Framework Laptop 13
1

Hi, everyone. I was doing some power consumption testing on my Framework (11th-gen i7, batch 4) which led me to discover a rather odd behavior. On my laptop, the top left expansion port consumes more power both when charging the laptop through it or when having any expansion module attached to it (including the USB-A modules, but excluding the USB-C passthrough modules, of course).

Using the same power adapter and USB-C power meter on each port, I got the following readings for idle power consumption during S3 sleep (charger on referenced port, all other ports empty):
Top Left: 1.55 W
Bot Left: 1.22 W
Top Right: 1.22 W
Bot Right: 1.22 W

I also tested with the HDMI expansion card and the USB-A expansion card in all of the different slot configurations, and having either card in the top-left slot caused a similar difference in overall system power consumption compared to the same module being in a different slot.

But here’s where it gets even weirder. My friend also has a Framework (11th gen i7, newer batch than mine), so I did the same tests on his. For his machine, there were also ports that exhibited a higher idle load, but it was the bottom-left and bottom right ports. Here are the measurements for his (charger on referenced port, all other ports empty):
Top Left: 1.07 W
Bot Left: 1.42 W
Top Right: 1.09 W
Bot Right: 1.45 W

Both laptops are running BIOS v3.07. Mine was upgraded to it, while his came from the factory with it installed.

Can anyone else corroborate these findings? And can anyone from Framework comment on why this might be the case? I’m more than happy to do any additional tests or experiments, or to test any beta firmware / BIOS updates.

3 Likes
2

Could you explain more what you were measuring and what you mean by idle power?

If you were to attach a USB power meter to the buildin C ports, I would expect them to display zero without a load or charger attached.

3

I was attaching a USB-C PD power supply to the laptop with the power meter in the middle to measure the power delivered from the supply to the laptop. The laptop was in S3 sleep mode during the tests, and I waited for the battery to stop charging so that I was only getting the idle consumption of the machine itself.

1 Like
4

Very interesting. To me this seems to indicate that your top left slot is somehow more inefficient when stepping down the PD voltage to the internal system voltage.

Are you perhaps usually using the top left slot for charging? This could mean that the circuitry has degraded through repeated usage and be maybe overheating. Which obviously doesn’t sound great and might he related to the melted chip problem reported elsewhere here.

5

I intentionally avoided speculating on the cause of the issue in my post. I would need a better understanding of the power system of the computer to make any sort of guess, and I feel it’s counter-productive to just throw random ideas out there. To your theory: no, I don’t use the top-left port for charging. I’ve mainly used the bottom-right.

That said, if Framework were willing to supply me with a detailed schematic of the power input subsystem, I’d be more than happy to take an oscilloscope to my mainboard and try to suss out any component-level variations. It’s certainly a plausible avenue for investigation.

1 Like