There is always the m1-s1
Excuse my ignorance, please; which AI slop?
You can check the post edit history of the post I quoted. Typical AI generated content with much cringy affirmation and back patting.
I wouldn’t recommend opening the PSU. Switch mode PSUs got high voltage inside that can hurt/kill (even after unplugged).
There are some experiments that can be done to narrow down the cause of the fan behaviors:
- Try the full ATX PSU on another computer
- Try the Framework FlexATX PSU on another computer
- Try the Framework motherboard and PSU outside of the case (someone mentioned it could be the heat from APU triggering the PSU fan)
Ideally monitoring the power draw too
So after contacting support my Replacement PSU #1 was sent out last week Wednesday with Fedex from Amsterdam and finally arrived yesterday. New in box and sealed, nothing to complain at this point.
And as any long time reader would suspect, the RP#1 behaves exactly the same as the Original PSU, think I’ll name that OP from now on.
I also did low load power consumption measurements, reference is my HDplex 250W GaN.
Win11 22H2 clean, non-RGB KB+mouse, self-powered display on USB-C 1080p60, idle after few minutes
HDPLEX ~6.7W OP ~11.7W RP#1 ~11.7W
This rather large difference reduces to 2-3W somewhere beyond 20W and actually does not change much up to 140W, this is actually respectable for 80+ Silver @ 230V (and yes the label only says 80+ Gold but thats for 110V).
TDP OP HDPLEX
100 142 140
50 77 74
15 29 26
Just contacted support again with new recording of the fan roar-up and images of serial and label on the PSU.
Though I have to say that PSU testing is a bit of a hassle as I have to pull the machine out of service for 1-2 hours and I’m so satisfied with the HDplex 250W I’d at this point prefer to just return the original PSU.
On a positive note I observed no whine or buzzing with either OP or RP#1 - though few seconds before the fan started up there was some kind of clicking, few times every 1-2 seconds or so - maybe that was the fan controller trying to kickstart the fan but failing to do so.
Are there batch numbers or other tracable codes on your two PSUs? Comparisions with PSUs that are noisy with those that are not may yield some insight
Unfortunately not, there is a production month on the label on the sealed box the replacement PSU comes in (mine says 09/2025 so it’s quite fresh). Of course there is the serial number on the PSU itself so I assume FW or FSP have access to data to trace that.
Welcome to the support loop.
I’ve now been asked to send my desktop to a repair center for further analysis. I declined and asked for a refund. I’m fine with just the board and the fan. That’s it. I’m done with the PSU and case.
I’ve been building my own PCs for roughly 25 years now and should’ve known this was going to be a roller coaster. Prebuilt’s just are’nt for me…
That’s a disappointing outcome after the hard work you put in
On the positive side the FW team now have your Desktop and replacement PSUs. Hope that’s all they need to get to the bottom of the noise
They don’t have them yet. Also, my individual parts won’t help them. To me it looks like they’ve been fooled by FSP (or their own team working with FSP).
Or has anyone heard a report of a FW desktop PSU consuming less then 9W in any other state then “Off”? (9W is only achievable in S2 standby, it’s ~11W otherwise)
I was talking about the complete desktop you’re returning for a refund … or did I misunderstand?
My intention is to keep the board and the fan so i’m currently still awaiting their response before i’ll return anything.
I don’t think the parts i got are different from any other parts. They work like they’re designed, but the design isn’t as advertised and seriously flawed (the PSU at least).
Oh boy. Still stuck in support hell and in the process of figuring out my options. Going by the initial responses I’m not very optimistic.
It’s such a shame. I really wanted to believe they learned from their FW13 launch (got burned with the 11th gen Intel cmos battery problem). Currently stuck with a cracked FW12 chassis and a noisy FW Desktop with bent cooler. Seems I don’t learn either ![]()
I love both machines on paper, but definitely can not recommend any products to normal consumers in my circles.
Edit: On a more positive note. FW13 seems to be very well liked now after a few iterations with the main concept holding strong. Hopeful the FW12 and Desktop will get there with us early adopters too. Even though it’s a bit painful at the moment.
Adding fan misalignment and USB issues to the other persistant problems, not least of which is the PSU noise, my early adopter risk-o-meter is nudging the red
Today is the first time I’m seeing this thread. I’m surprised to read people’s experiences with the PSU fan since I’ve never experienced this myself. I have a batch 2 Framework Desktop with the Noctua CPU fan and it’s super quiet even using all cores at 100% to compile code. Sure the CPU fan spins up and emits its normal sound but I’ve never heard the PSU fan in my month of using it. When idle, I can put my ear up to the system and hear nothing. My TuringPi heatsink fans are louder than the Framework Desktop.
I’m not saying this to invalidate people’s experience. I just wanted to note that there are positive experiences as well!
Edit: I encourage people to vote in [Poll] Request to share your PSU noise pattern & experiences so we get some data aggregated.
Edit: I hear the sound people are describing: Noisy PSU fan - #226 by Matthew_Sanabria
Batch 8 and I have the PSU whine. It is a bummer of a noise and I have opened a support ticket. Hearing that Matthew’s doesn’t make the sound gives me hope they just got some bad ones that shipped recently, and can send out better ones.
That’s really good to know, @Matthew_Sanabria Thanks very much for taking th time to comment on your positive experience
From my POV it would be great to have your full kit in FW hands as it has the problem that’s in need of an urgent fix …
The noise problems you’ve experienced aren’t universal (although too many pre-order customers do have them) so having a Desktop with the problem should help FW with diagnostics
I am being quite selfish with this suggestion, tho. You do whatever you need to do
Thank you so much for sharing this! You have no idea how good to hear that is for an affected user. Even though the poll shows a majority of noisy PSUs (at least within participants) it gives a lot of hope that this is fixable. Seeing a few non-affected users is such a relief.
Also great to hear there are PSU alternatives that fit in the stock chassis, if someone wants to take matters into their own hands.
Worst feeling ever living in isolation and uncertainty. So thanks again for sharing!
How is your systems idle power draw at the socket? Are you able to take a measurement?