Investigating Reports Of Intel Skylake CPUs Damaged By CPU Coolers

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
Not sure what the torque would be for mounting screws, the 50lb spec by intel is overall not per mounting screw. Of course there are torque screwdrivers. This one ranges from 1.77 to 35.39 in-lbs for around $397 usd.
http://www.grainger.com/product/WESTWARD-Torque-Screwdriver-6PAF5?s_pp=false&picUrl=//static.grainger.com/rp/s/is/image/Grainger/6PAF5_AS01?$smthumb$

I've got the dark rock pro 3 mounted to my 4690k, the instructions explicitly state to tighten until they stop, no further. I'm not sure if they're tty though. I also don't believe they can be infinitely tightened so as to keep crushing the cpu or anything, they reach a certain predefined depth and stop.
 


I tried join cooler master forum to find out anything about evo 212 cooler and my i 6600k which i'm delaying to put together my build the activating url doesn't work. So still nothing from them?
 
Synphul, we were joking and being obtuse. I'd figure you'd be familiar with the old shad tree'ers joke about turning the bolt until it gets soft again, and then two more turns, just to be safe. In other words, turn it until it's stripped, and then two more turns. :)
 


Yea I got it lol. I was somewhat being serious, somewhat joking. Doubt anyone's going to run out and buy a ~$400 specialty driver to install a $35 cpu cooler. I knew what you meant, sadly because I've done it - tightened something just oneeee more turn and spun it free. Usually followed by an expletive and a search for a slightly larger tap and fastener lmao.

Not on a cpu cooler though (thought I'd point that out). I tend to try and be a bit more careful with smaller stuff with low grade soft metal and fine threads begging to strip.
 


I knew there were kidding so I'm waiting on cooler master answer or about modifying the mounts other cooler company's changing bolts or screws or bracket to be safe.
 


Like using wood screws on metal threads.
 


The Core 2 Duo heatsink is still in the top 10 of the quietest CPU coolers of all time and the only change from that cooler to the current one is the heatsink size. The fan is still the same.
 

The Hyper 212+'s screws are threaded for something like 3/8" so once those threads are in tight, you can't go any further and the springs on the clip provide the tension from there.

On many other CM coolers though, the threads come directly from behind the motherboard and are fully threaded, which means there is no limit to how much you can tighten other than when the screw, heatsink bracket, motherboard, CPU or whatever else fails. There are no springs either to buffer excess tension on the screws. Those are the types that you tighten slowly until the HSF is flat and snug against the CPU, and then tighten just a little more to remove remaining slack. I can easily imagine how people may over-tighten those.
 


That article has no credibility whatsoever. Their techs are using electric screwdrivers to mount the heatsink.
 


Which means those techs are just about as dumb as what we were joking about being up above. Not surprising really. PCgamer often does things that make you wonder if they are genetically challenged or what? Maybe missing a few chromosome pairs.
 

I hadn't looked at the original article's pictures before.

The way it is only the corners where the keying pins are that are bent up, I strongly suspect those damaged chips are actually from people closing the IHS latch with the CPU inserted the wrong way: since the substrate is thinner, the corners resting on socket keying pins due to incorrect orientation get bent or cracked when the IHS latch is closed.
 
I built 2 6700k systems with a ASRock Z170 Extreme7+ motherboard and Noctua NH-U14S cpu cooler in early November. One system is a server for our house and the other is for my son who is away in college. We put all his programs on it and got it setup for him over the thanksgiving break. He then took it back with him in the car to college and was using it in his dorm room. He can home for xmas break and brought it with him. I actually carried it from the car that time and into the house. It had been setup for around 13 days but yesterday he was in eclispe and the machine rebooted and windows say it had a problem and wouldn't start. I tried to use my acronis usb flash drive to make a backup of his ssd but kept getting like a small crash dump trying to load acronis. I swapped out the power supply and ram from the other system but still couldn't get it to load acronis. I thought I was going to have to reinstall windows but I took the ssd from my son's computer and put in the server machine. I was surprised but glad it booted just fine! I then booted off my acronis usb flash drive and made another backup. I can't say for sure but my sons system sure sounds like it has the skylake bendgate problem! I compared all bios settings(over an hour) between the 2 systems and can't find anything different. At this point I don't know if its the cpu and or the motherboard that has the problem. I haven't took the cpu cooler just off just yet but I don't know if I can tell visually if it has a problem when I do so. Does anyone have a different cooler recommendation? I am not over clocking so the cooler needs are not that great. I like the Noctua because it's so quiet but if I have to take the cooler off every time my son comes home from school then I need something else.
 


MSI was selling a reinforcing plate . If the processor is just bent and not broken then installing that might flatten it out enough that all the pins make contact again .
Might be worth a try

Probably a good idea to have the plate even if you have to buy a new processor since you would then be able to keep using the Noctua
 


Actually a decent electronic screwdriver will have torque settings that mean you can not over tighten a screw .
Since thats likely to be the case I think we have to reconsider who the dumb people are
 
I trust my sense of feel to not overtighten much more than I trust a powered device. The screwdriver doesn't care if it over tightens something. I do. I'm not saying there are not people out there that won't overdo it by hand, but those same people are probably even more likely to overdo it with a powered tool as well.
 
25+ years tinkering with PC, then running my own business doing fixing them, laptops, cellphones, consoles and always used a regular screwdriver. Sure you can set a torque setting, but once you know the systems, they do better with some things tighter or looser. PS3 ground screw from the PSU, tighten that thing down, XBox 360 screws, you can warp the board not tightening right.
 
6700k+212 EVO, removed the cooler and CPU to inspect for damage, nothing found. Mounted it again by bottoming out the screws as I did initially.

I dont think you have to worry abbout anything as long as you dont transport your system with cooler mounted. I suppose if you have the kind of cooler that has a mounting screw that doesnt bottom out you have to be careful not to go full neanderthal with the screwdiver.

I really dont see the problem here, it is all overhyped.
 


It is unlikely there is a computer assembly line anywhere where tools like electric screwdrivers are not used .
Its the fastest was so it will be cheaper and it will be SOP
 
I don't know of any assembly lines that install aftermarket coolers. It's not even recommended that they be installed prior to shipping, as clearly indicated in these and pretty much any related articles. And we're talking about people causing this to happen, not an automated assembly line. Equipment on an automated assembly line, or even one manned by individuals, is likely to be using equipment that is regularly calibrated, unlike the electric screwdriver some shmuck picked up at harbor freight. This is a lost cause, I don't know why you're even arguing the position really.
 
Status
Not open for further replies.