Skylake CPU caution |
Post Reply | Page <123> |
Author | |
parsec
Moderator Group Joined: 04 May 2015 Location: USA Status: Offline Points: 4996 |
Post Options
Thanks(0)
|
The Corsair AIO CLC CPU coolers, which I have used in the past, all use a generally identical mounting system for the pump/water block that attaches to the CPU socket. There are four mother board standoffs that only allow so much pressure to be put on the CPU. Once the four "arms" on the water block contact the top of the standoffs, continuing to tighten the four thumb screws does not cause any more pressure on the CPU. I've broken the threaded part of the standoffs when I tightened them to much, and had to get replacements. I even created my own standoffs with screws and spacers that allowed more pressure to be put on the CPU. I never bent any pins in the socket doing this (but anyone doing something similar must know they are responsible for anything that happens when you don't use the standard parts.) If the CPU substrate (the green part that looks like a printed circuit board) was bent, you could see it. If you put the CPU on a flat surface, with the bottom of the CPU on that surface, you could see the edges of the substrate not touching the flat surface. Every time I get a new mother board, the first thing I do is inspect the pins in the CPU socket. Put the mother board in an area where nothing can fall onto it. Have a good light source so you can see the pins easily. Carefully open the the CPU socket. Use a flashlight but don't drop it! Since the pins are so small, it is easier to notice a difference in the appearance of the pattern of the pins than trying to look at each pin. Move the flashlight around to illuminate the pins in different ways. The CPU fits in the socket only one way, be sure you know the orientation when you put the CPU in the socket. When you tighten down the water block on any AIO or custom water cooling system, always do the following. This is only for mounting Intel processors:
|
|
FarmerVin
Newbie Joined: 04 Apr 2016 Location: CT Status: Offline Points: 6 |
Post Options
Thanks(0)
|
I had this issue with an Asus Maximus Impact VII itx and a Corsair Hydro Series H100i v2. Board wouldn't post so I went through everything and found a few bent contacts on the socket.
I have an ASRock replacement arriving today and I'm very nervous about installing this cooler again. I hope my CPU isn't damaged... It doesn't look warped. I'll post a picture of the socket. The way the contacts are damaged doesn't look like an over tightening issue, but this is the first time I've ever experienced something like this. I feel like I must be at fault. It's a $230.00 lesson...so far, but I don't feel like I've learned anything from it.
|
|
wardog
Moderator Group Joined: 15 Jul 2015 Status: Offline Points: 6447 |
Post Options
Thanks(0)
|
Coming back to this .......
Updated responses from cooler manufacturers: Original link in German http://www.pcgameshardware.de/Luftkuehlung-Hardware-217993/News/Skylakegate-Kuehler-zerstoeren-Sockel-1151-CPUs-1179237/ Translated via Google Translate https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.pcgameshardware.de%2FLuftkuehlung-Hardware-217993%2FNews%2FSkylakegate-Kuehler-zerstoeren-Sockel-1151-CPUs-1179237%2Fhttp%3A%2F%2Fwww.pcgameshardware.de%2FLuftkuehlung-Hardware-217993%2FNews%2FSkylakegate-Kuehler-zerstoeren-Sockel-1151-CPUs-1179237%2F&edit-text=&act=url |
|
parsec
Moderator Group Joined: 04 May 2015 Location: USA Status: Offline Points: 4996 |
Post Options
Thanks(0)
|
The latest take on this from Cryorig, maker of large CPU coolers.
It only happens on "prebuilt PCs that are shipped or transported by the owners", to paraphrase. Also, CPU coolers with spring mounting pressure screws won't suffer from this issue. Well DooRules, that really fits your situation... not. I'm sure that pre-builts like IBuyPower that ship PCs with two '980s in SLI on a $100 board with the stock Intel CPU cooler for Joey gamer boy are worried now. They just sell sooo many PCs with Noctua NH-D15s... Funny how it seems Intel is not taking much of a hit on this. You are wrong wardog, it's Bad aftermarket CPU cooler, and Bad UPS. I'll never forget the monitor I had shipped from Newegg that literally had a boot print in the middle of the box. I opened it a bit for fun, the screen was crunched and shattered. The best part was after contacting Newegg, whom sent me a new one, told me the ruined monitor now belonged to UPS. |
|
DooRules
Newbie Joined: 05 Nov 2015 Location: Newfoundland Status: Offline Points: 122 |
Post Options
Thanks(0)
|
It was i who had the bent pins on two mobo's.
I got lucky and was able to fix the second board with a bent pin. Up and running now and fully functional. I have no way to be sure but if I had to bet I would lay the problem at the feet of my previous chip. I got an rma for that chip and all good now. Built many many systems in my time, first time I ever saw bent mobo pins.
|
|
wardog
Moderator Group Joined: 15 Jul 2015 Status: Offline Points: 6447 |
Post Options
Thanks(0)
|
Twice in my OC'ing days do I wish AMD offered what Intel does.
https://click.intel.com/tuningplan/ To those with warping processors, buy a plan and then sit on it for a couple weeks before turning it over for replacement. That is if Intel themselves don't come around to owning up to this problem. hehe Bad Intel |
|
Xaltar
Moderator Group Joined: 16 May 2015 Location: Europe Status: Online Points: 25028 |
Post Options
Thanks(0)
|
This is a very interesting theory, the vast majority of the reported occurrences I have seen have been overclocked and while I did not see any indication of voltages used it is a fairly safe bet that at 4.5ghz+ they may have been pushing it. Heat may well be causing the substrate polymers to soften, resulting in bending. I have to admit, this makes a lot more sense than "heavy coolers broke it" and would also explain why not everyone using monster tower coolers are having this issue. That combined with the fact that the substrate is rated by intel at the same tensile strength as previous gen CPUs would lead me to believe you may be right.
Edited by Xaltar - 06 Dec 2015 at 2:49am |
|
wardog
Moderator Group Joined: 15 Jul 2015 Status: Offline Points: 6447 |
Post Options
Thanks(0)
|
Bad Intel
hehehe, that coming from an AMD Groupie I am. |
|
wardog
Moderator Group Joined: 15 Jul 2015 Status: Offline Points: 6447 |
Post Options
Thanks(0)
|
Yea. Hi Torque! Meh, most shoot for the biggest baddest HS money can buy. Then the OC the pis* out of it. Pushing insane voltages. Voltages which result in heat. Heat that is most likely, and over much time, now allowing/causing the substrate to deflect when hot and now "cure' in this position/deformity. |
|
parsec
Moderator Group Joined: 04 May 2015 Location: USA Status: Offline Points: 4996 |
Post Options
Thanks(0)
|
Well, I have a Noctua NH-D14 on my i5-6600K, in my ASRock Z170 Extreme7+ board. The specs list the NH-D14's weight with both fans as 1240 grams, or ~2.73 pounds.
But weight is not likely the main problem, it's pressure on the CPU from the CPU cooler mounting. My middle name is "torque", meaning I like things tight. I did not spare the screws during the mounting process. They are tight to a stopping point, but not much more. I have no problems with that PC at all. I just changed the memory to G.SKILL Trident Z 3200 Pinch which had me worried about using it at that speed. Clicked the XMP profile in the UEFI, booted right up the first time and every time. If the pins in the CPU socket were compromised, I would expect problems. I'm using a 4.4GHz OC. All this does NOT mean I think this issue does not exist. A friend of mine using the same board has apparently bent pins in the CPU socket on two of them, using a custom water cooling CPU block. He said he barely tightened down the water block on the second board, but had the same result, one bent pin in the socket. He used the same CPU on both boards, which I would bet has a warped/bent substrate. As Xaltar said, the substrate on Skylake processors is visibly thinner than previous Intel processors. IMO, it is a bit odd that just recently we are seeing this being reported. One of the main factors of compatibility with CPU coolers across different socket types (1155, 1150, 1151) is the distance between the surface of the mother board and the top of the processors IHS when mounted in the CPU socket. That must be identical or the contact pressure between the IHS and CPU cooler's contact plate will be to little or to much. My point is did something change with later production runs of the 1151 processor sockets. I guess it is still early in this mystery but hopefully we'll learn what is going on eventually. |
|
Post Reply | Page <123> |
Tweet
|
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |