Very Successful Kludge-Fix on my 17" Asus i7 notebook - and a Question.

So this morning my Asus 17" X750JN notebook suddenly started thermal spiking the CPU (i7-4700HQ) to 100°C and throttling down from its base of 2.4 GHz to 2.1 GHz, for no apparent reason. I wondered if the throttling had anything to do with the Windows 10 Anniversary Upgrade (v. 10.0.14393) that I did the day before. After digging around the upgraded OS, I saw that the Windows 10 Updater had installed intel's Dynamic Platform Thermal Framework package, and it checked out fine (via Event Viewer), so that wasn't it.

I figured that maybe the CPU cooling fan was really dirty (it had been 5 or 6 months since I last opened and cleaned the laptop), so I opened it up (long, careful, tedious process), and I immediately noticed that that one of the 4 stainless steel legs going up to the CPU heat sink had popped up off the system board. That would explain the crazy spiking, thermal runaway and throttling. What in the world would cause it to pop off like that? (I've heard horror stories about Asus' notebook quality control).

I saw that the tiny steel machine screw was still in the hole on that leg's foot pad, and the knurled, round nut/socket (about 3mm threaded "donut") that goes into the hole in the system board had popped out, and was still threaded onto the screw I also noted that none of the 4 feet had springs on their screws.

Anyway, for the popped socket, I tried to push the nut back down into its hole on the system board, but it kept popping back out. It simply would not stay put. I think these socket nuts originally might have been epoxied into the board at the factory.

So here's what I did: from the bottom side of the system board, I positioned the nut/socket on the rim of its hole (about 3mm in diameter), from underneath, then inserted the machine screw from above the board, down through the steel foot pad on the spreader leg, through the hole in the system board, and I had just enough threads left on the screw to capture the socket nut sitting underneath. The socket won't even enter the hole that way. It just sits on the rim of the hole, and it's secure. I then loosened the other 3 screws on the heat sink spreader legs, in the 1-3, 2-4 diagonal torquing order (counting the number of full rotations on each screw as I went), then re-torqued them all in the same diagonal pattern, same number of turns...(I kinda fudged it).

Long story short - I got it all back together, it powered up & posted just fine, and I've been stress-testing it for 6 hours now at full Performance Power Profile in Windows 10, using Prime95, PassMark, Cinebench R15, GeekBench, and BOINC clients...it runs and completes all of them just fine, temps max no higher than 88°C, and most important, no more CPU frequency throttling. It "Turbos" all 4 cores up to 3.2 GHz as needed, and up to 3.4 GHz on one, just as it should. Oh, and I cleaned all the dust & gunk off the fan blades, too, then with a can of air I blasted the whole thing out.

My question is: Is my "fix" here putting uneven mechanical stress on the actual CPU die itself?

The one "popped" heat sink leg & foot (stainless steel - very rigid) that I secured to the board is pitched down from the top of the CPU to the system board at an angle noticeably steeper than the other 3 (since it no longer sits 2 mm higher on the nut on the top side of the board, as before)...the "popped" foot pad is actually screwed flush down flat on the motherboard, right over the (empty) 3mm socket hole.

The other 3 stainless steel feet sit about 2 mm on their sockets above the board; that's how high the socket nuts poke up through their holes and above the system board.

Anyway, all 4 screws are nice and evenly torqued & tight (the 4th one a couple turns tighter), but now i'm a little worried about stressing and cracking the CPU die on that corner where the heat sink leg and foot pad sit lower.

i'm wondering if i put a 3 mm thick washer topside of the hole where the socket nut used to go, then get a longer machine screw, toss in another washer bottomside, between the nut and the bottom of the board, then put the longer screw through it all, from the top...maybe that would reduce any offset strain on the die.

I've included a .jpg of these little 3mm knurled "socket nuts" (as I call them) - "socket_nuts.jpg"

Phew....apologies for the long, repetitive narrative -

Any feedback / ideas welcome...& thanks in advance

- Trev

ps - again, I haven't a clue as to how, or why the socket nut popped out of the system board on its own. Crazy.