So my latest journey in optimizing the workstation build has led me into a dark corner that is USB4_FAN. After completely replacing all the fans, rewiring, 3d printing air flow controls, hand tuning curves based on extensive testing of various combinatoric factors; There is a little black box of doom sitting on the NW side of the motherboard, some may know as USB4_FAN.
After some cursory investigation into the situation, attempting to identify if the sensor the fan curve is matching to is exposed anywhere (doesn’t seem to be, nor is the sensors location documented). It seems at least to be within a reasonable std dev of other instances of the reading Ive seen online.
In order to influence the fan I have tried tweaking the preset curve through the bios, unfortunately it is not exposed in the ipmi fan control. That did seem to drop it by about 500 rpm though, to me it still seems unusually high considering I only have a keyboard and mouse plugged into the USB.
Next I tried to start disabling USB features in the bios like mass driver support and what not. Nothing really seems to impact it.
So long story short, any one have a tale to share or understand why this thing is spinning a trillion miles a second?
this fan is inside the rear io shield. they never figured out how to manage the heat from the chipset. are you wanting to turn it off/is it making noise? it’s not easy to get at and is kinda important
With the rest of the box being so quiet, its high pitch whine stands out now. Wondering why it needs so much air, how I can reduce the temps there by reducing the need to spin so fast, and if other people are seeing similar rpm’s
I was able to fix the m.2 heat issues (also causing high rpms on the m.2 fan) by removing the asus facades screwed to the top of the heat sinks. Maybe I could do something here? or maybe build a micro water cooling setup for this atomic energy plant requiring a fan to buzz at 5000 rpm.
Maybe disable USB entirely on the motherboard and add in a usb pci card?
the usb4 card is removable, iirc. The whole back plastic I think is unscrewable with no deleterious affects. The puget systems workstations I’ve bought seem to come with this bit of plastic not installed, perhaps for this reason
I really need to get better about time boxing my projects. Anyways for anyone who decides to do this in the future. Heres the skooopeelooo.
Remove the motherboard from your case
Remove the two screws on the back of the motherboard. One is at the very top, the other is on the bottom towards the edge of the board. They are very small and will likely require additional light to find if you are blind like me.
Anyways, that USB4 cube the fan is attached to is something else, I see why they hid it under a piece of black plastic. Also found the temp sensor, apparently a couple degrees above ambient temp is the end of the world and we must spin as fast as we can. At least the high pitched whine is much better with that plastic piece removed.
found this thread while chasing down this noise, since the system is now otherwise silent. it does look like the fan curve for the usb4 fan can be set in the BIOS even down to zero - can confirm visually and through the IPMI sensors page. it appears to be the same for the other tiny mobo fans. I did a custom fan profile for the usb4 fan but still hear some high-pitched noise from that area even if the usb4 fan is not running so not sure it’s the only culprit. I’ve also set the other fans to silent profile but intend to play with those as well to narrow it down.
Mine spins at the same 4560 RPM, if I have to trust the ipmitool sensor readings. Luckily enough, on my mobo within a coolermaster HAF berserk it makes no annoying noise.
