I got my Framework 16 today and so far I enjoy everything but the fan noise. When my computer is under load there is a high pitched (compared to the rest of the fan noise) fan noise that, according to an audio spectrum analyzer, is a ~5400hz sound that is around -54 db around 4 inches from the back of the dGPU module. The rest of the fan noise profile is fine to me but that 5400hz sound is really annoying.
Here is a youtube video that sounds exactly like it:
Hopefully someone from Framework can chime in and let me know if this sound profile is expected or there is an issue with my unit.
Anyone else hear this when your computer is under load?
This sounds like a coil whine issue. Fan noise does not reach such high frequencies!
I’d recommend contacting support, coil whine is always present, but in some cases (such as yours) it’s much worse.
I know some people who get their motherboard replaced (and the old one being fixed, should just be an SMT swap).
Update, I installed my shell module in place of my dGPU module and the high pitched sound is not noticeable unless I put my ear against the exhaust. So it has something to do with my dGPU module.
It would be worth trying to identify which fan is producing the sound and investigating further. People have had problems with fan blades touching tape or glue points and producing noise.
Got the new GPU module today and it didn’t fix the issue. So it looks like there is something up with my laptop chassis that is causing the whistle. It definitely seems to be an air caused whistle sound rather than a noise caused by electronic coil whine. It sounds very different than the coil whine that my cheap higher power USB-C power supply makes.
To improve my ability to test the issue I got ectool working to control the fan duty cycle to isolate any noise from putting the computer under load and the noise caused by the fans. (see this post)
I was able to run my shell module at 100% fan duty cycle and I noticed the whistle. It was more difficult to get the shell module to that fan speed using a stress test and it doesn’t start becoming really noticeable until around 60% fan duty cycle which is why I didn’t notice it previously.
Yes it’s awful. The noise gives me a horrible headache over the day and massively agitates my tinnitus. I was so excited for this laptop but after watching reviews and reading posts about this noise, I’m getting worried that I have no choice to return it… I just wish the highly repairable laptop had some new fan options or some guidance on what I can do to solve this whining/whistling sound!
The Framework 16 has coil whine on the dGPU the moment that it receives a high power load. I had a replacement dGPU Module shipped to me but it also experienced the same coil whine noises as the first.
Framework Support said there is some coil whine expected within a certain range that they allow to pass QC. Unfortunately, it seems like nearly all the dGPU Modules are affected with some amount of coil whine.
It is not coil whine I am describing. This noise occurs when the fans are moving regardless of if the GPU is installed or the blank module. It’s caused by the fin profile of the fans or air intake profile or exhaust profile.
It’s pretty common and usually doesn’t affect performance. If it’s bothersome, try adjusting the fan curve in your BIOS or checking for firmware updates.
This isn’t about performance. It’s more about an annoying sound from the fans making using the laptop less enjoyable.
This is noticeable for me whenever the fans are above 30%. It’s noticable under 30% if I am in a more sound isolated environment.
It seems like a flaw in the fan design because there is a spike in one specific frequency that is an outlier from the rest of the sound of the fans. What makes fan noise annoying is if there is outliers in loudness in the sound frequency curve.
Support replaced my laptop and it ended up still having the issue. I think it just needs to be better documented because there are some people who are sensitive to this type of sound profile. It would save support time if it’s clearly documented.
What do you think you are going to do? I’m on the fence about returning mine but I can’t be giving myself headaches just because I want to support the framework mission and have an upgradable laptop.
I have been experimenting with covering the back exhaust outlets with my hands, and it does take away the whine. I don’t think the whine is in the fans but instead comes from blowing through the rear heat sink. I wonder if adding some kind of foam at the back would help. Right now I am doing an insane thing and covering the rear exhaust with electrical tape and that actually takes that high pitch away very well (of course I am also suffocating the computer, but at least it’s not massively overheating)
From my testing, it seems to be the fins closest to the plug on the back of the dGPU. If you cover just those, it’s a noticeable reduction to whine. However I have found a better solution for me which is to use a lower power charger to trick the firmware into keeping the fans lower. Either 90W or 100W USB-C does nicely and the fans throttle back. However it does impact dGPU performance, while surprisingly not impacting CPU performance. I think this will be the solution I use until framework lets us control the fans.
Ya, something about the fluid dynamics of the exhaust/intake creates a resonating whistling sound at around 5400Hz - 6200Hz. It was around 5400Hz on my first gpu module and they replaced it and it is now around 6200Hz.
This seems similar to the noise that I hear when using my Framework 16.
I guess I’m not as sensitive to these noises, so it doesn’t really bother me as much.
I also primarily use my Framework remotely through another computer, so while it is sounding like a rocket engine, I don’t have to listen to any of it.
That’s so strange…to me, that (if you didn’t tell me about the fins), sounds a lot like coil whine. It’s very constant, not a lot of fluctuation. But, yeah, if it’s mechanical due to the fins interaction with the air flow…then I wonder, can you, say, put some tape along each fin (i.e, not taping over the holes) to interrupt (but not impede) the airflow.
