Interesting! That’s clearly not the case. I wonder if that’s enough to call it a defect of the unit at this point. I wanted to install the fan with the cable pointing out towards the front of the case so that there is no slack on the cable. But after a bunch of attempts that didn’t work at all. The only way it worked for me was with the fan cable pointing out the top. I was thankful that at least that worked so I can get the system up and running.
Yep, I had the same problem. I drilled one hole to make it wider, and then had to tighten screws to make it actually fit.
Any chance you could do a photo with an arrow along the cable? And yours is a Noctua fan, did I get that right?
This is exactly what I had to do (batch 11). I could only get 2 holes to fit at a time. Aligning the fan on top, and turning the fan screws into the heat sink worked flawlessly.
I’m kind of confused by this issue. The nocuta fan fit perfectly for me. It felt like it was only two at first but once they were all actually aligned it fit perfectly.
If it don’t fit - take a small 3d printing rasp and clean out the fan screw holes - one swipe per hole and the fan slides in like a glider in the air. This is not a defect - just new stuff.
After running with the A14x25r fan that fit so nicely onto the mounting posts I decided to try the original fan again.
Before I dremeled it out any more I wanted to try some of the suggestions here.
- It didn’t slide on directly in either of the 90 degree rotations.
- I didn’t try it flipped over because I was mounting it the right way up in the first place.
- I noticed that one side would mount and the other was stranded and I could lift it back off and then the other side would mount leaving the first side stranded. This was always “horizontally”, i.e. towards the front and the back of the case. I never got it to seat both top or both bottom holes.
- I tried really hard to get it to align all 4 holes equally and I’m pretty sure that I was in that “even mounting locations” state several times, but it just sat on top of the posts.
Before I bored the mounting holes out any further I tried once again to get all 4 holes lined up at the same time and randomly I ended up pushing a little more than I had been to verify the alignment - it turns out that I had already bored the holes out enough that it would slide on with a small amount of pressure - nothing huge, just fingertip pushing, but it mounted completely without having to use the screws as a vice. I tried it both ways with the cord at the top and the cord on the M.2 slot side and it was the same - just a small amount of pressure was needed to mount it.
I’m pretty sure that small amount of pressure wouldn’t have worked out of the box because I did try some pressure and wasn’t comfortable with how much it seemed to require, so the dremel tool did help.
With respect to performance, it’s a mysterious bag. The stock A12 is very quiet, definitely on par with what I heard with the A14. The A14 should be able to move more air for a given noise level, but without a shroud it wasn’t as effective for a given CFM level so it likely ran faster than the A12 needed to.
Some quick figures I measured:
Cinebench R23 single core - 2012 @ 62c vs 1985 @ 64c with the A14
Cinebench R23 multi core - 33670 @ 68c vs 34274 @ 74c with the A14
Prime95 for 5 minutes averaged 3.6GHz @ 70c vs 3.4GHz @ 75c with the A14
So, it was cooler all around even with higher clocks on Prime95 - the only issue there is that I was using the “blend of all test types” setting so a lot would depend on whether it was using the same test distribution on the 2 attempts. I would need to run some tests with only a single test type to double check those numbers.
Cinebench is a mystery because it performed better on single core, and worse on multi-core, but both had much better thermals. I don’t have a scientific thermometer next to my desk, so I can’t vouch that 2 runs a few days apart had similar ambient temperatures, but the performance of the stock fan looks really good. Also, when I ran the single core before on the A14, Windows started sleeping which may have affected its performance on the A14. Sorry, I didn’t want to spend another 10 minutes to verify that figure, but I did disable the sleeping for the subsequent tests.
Batch 12 here and I had to do the same thing to get the Noctua NF-A12x25 to fit.
And now…
I took the fan off so I could install another M.2 drive in the front slot. When I was done and remounting the fan, I couldn’t get the screws to install into the mounts any more.
(Note that the fan slid on just fine, it’s the screws refusing to screw into the threaded mounting holes this time.)
I’ve tried basically all of the screws in all of the holes and only a couple of screw+hold combinations work. Even worse, while I was trying them out I took a screw out of hole A that had gone in fairly easily to try it in hole B - it didn’t want to go in so I went to put it back in hole A that it had just come out of, and it balked.
On the theory that it might be the blue loctite on the screws that over-baked in the couple of weeks I’ve been testing the unit, I’ve tried using acetone, hot water and scrubbing with paper towels and a brush but I cannot get the blue gunk off - it’s like it’s made out of concrete.
I haven’t tried an open flame yet, or my heat gun, but it’s really odd that they went in so easily the first time, then they came out so easily when removing the fan, and now they absolutely refuse to reseat - sometimes after having just been seated fine in the same screw socket just seconds prior.
Same problem trying to get them to go in whether through the fan or into the bare mounting posts with no fan in the way.
(I was supposed to be installing a couple of Linux distros and testing them out on the new drive, but I’ve been in screw mounting hell for the last 3 or so hours…)
I’ve ordered a replacement fan mounting kit and heatsink. Even with expedited shipping I’ll be out of commission for about a week apparently…
Due to the screws not wanting to reseat on my old heatsink, I just got a new heatsink and shroud kit to try again.
My Noctua fan did not fit onto the new heatsink either. So, I took some calipers to the various mounting posts and the fan holes. It looks like the inner distance between the mounting holes on the fans are just a fraction of a mm too far apart to fit on the mounting posts without some force. The typical inside post-to-post measurement was just around 101mm with a couple just under by a 10th or so and some over that by a 10th or 2. The inner fan hole-to–hole distances on my Noctua measured around 101.3 or .4 meaning they did not have clearance to slide over the mounting posts.
More time with the dremel and a new cylindrical bit and I was able to shave the inner sides of the holes on the fan so that they were less than 101mm apart and the fan now fits with no force onto the mounting holes.
One thing I noticed when screwing in the new fans and shroud, I never got much resistance on the screws. They just kept screwing in and in but the fan was tight against the heatsink. I decided not to keep turning them as it was already tight enough, but shouldn’t they have seated and provided some resistance? I seem to recall this happening with the original screws and heatsink and now I’m wondering if I may have stripped the original screws and that’s why they wouldn’t re-seat after I removed the fan for the M.2 swap. Anybody else notice that the screws never seemed to fully seat?
This was what worked for me (thanks) - frustrating that it was necessary but an easy fix in the end
I’ve just received all my mainboard and fan bits and pieces
In my case, the Noctua fan fitted absolutely perfectly. I did have to double check that I had it oriented as the manual indicated but then it slipped on like a tailored suit
Setup new Framework Desktop yesterday with Noctua fan. When pointing the cable as indicated in the directions the fan would not fit on the posts. After rotating the fan 90 degrees counterclockwise it fit on perfect. The cable was long enough to still reach the connector. Life is good!
![]()