Warning: Some i7-6820HKs and i7-6700HQ have Uneven Core Temps due to Uneven Heatsink

Oh.. Do not be afraid that I will post what I find about AW, LOL
One more thing.. Do not underestimate the number of sold Clevo laptops for gaming purposes. Clevo is sold under many branding names. Not only Sager Recent figures I have about AW units sold. Is approximately around 100,000 units. You can figure the source and numbers on the forum threads or on internet. I have posted this earlier

 

I was never afraid LOL...just to be clear. You know me by now...I'll call BS like many of us on here when we see it.

Yes, that's true, Clevo has a bigger global presence than just the DTR's that most only know of, but included in those figures are also BGA Clevo's ... Just saying....

Regardless, the sheer fire power and strength of Dell squashes Clevo. They both have faults of their own.

Looking forward to that Tweet...it was hilarious like someone from AW just copy pasted the same answer over and over again. It really showed a lack of professionalism and a big hint of used car salesman in there....such a shame. Denial is not the answer to lessen a real problem....only the customers lose out, while the company makes a fool out of themselves. Don't take my word for it...let the Tweet post speak for itself. I didn't say it LOL... :X ..Proof is in the pudding.

 

Alienware "engineering"

I am not paying (especially that big price tag) for a machine with a design flaw out of the factory

 

can you give the link to where can i buy this pressure paper?

 

IIRC

https://www.amazon.com/gp/aw/d/B004...280_SY350_FMwebp_QL65&keywords=pressure+paper


https://www.amazon.com/gp/aw/d/B004...AC_SX280_SY350_FMwebp_QL65&keywords=pressurex

 

Ok thank a lot for this dramatic news.
I will receive mine today...
Just a curious question, in fact maybe in the alienware design it must be necessary to screw more the single screw than the other two in order to rebalance the pressure ? And not just to screw hardly all the screw ?
What do you think about that ?
Keep in mind that the heatsink is extremly rigid therefore it will necessarily go to the high pressure side especially as screws are push too hardly

 

I am wondering about whether this could pose a solution, i.e. not completely tightening the two screws, to take away some pressure.
A different approach could be to get slightly longer screws and use those.
I'm assuming that there actually are screws widely available that can be used.
(According to Dell's service manual, the default screws are M2x3 so maybe if we were to exchange the screws on the too tight side with M2x3.5 ones that are shortened to M2x3.2?)

Has anyone tried this?

 

Well good to know AW15 is out of the equasion in terms of buying a gaming laptop... Thank you iUnlock as I seriously was about to order one for myself.

Pretty disappointing. I like many things about it but temps are my #1 concern.

 

I am confused.. You own the 15R3 with 120Hz screen right? Means you use the 6700HQ.. and you had such dramatic problems still? Wtf AW...

 

Yes, that's correct. The uneven core temps and high temperatures in general were not limited to the 6820HK. It's pretty much entirely down to the heatsink and bad paste (I think mostly the heatsink though).

 

There is a Mod request in to change the title and redirects to update the thread based on what has been found by owners, and found by @iunlock as fixes for the problem.

Thanks to everyone's efforts in participating to find a resolution to the problem so quickly

 

Hello all!

I have a 6700HQ that seems to have Core 0 and 3 which go hot where the other 2 cores stay "fine"..

Did this forum actually come up with a solution, or is it just wait and hope to see if Alienware fixes this issue? I am going to try to repaste in the meantime.

Thanks!

 

Hang in there...I'll be posting a solution and results soon...

 

The cause of the problem is in the first post.

Also i created a thread on reddit, linking back to here and crediting iunlock (hope that's okay). Hope to bring some awareness to others and Alienware regarding the issue.

 

Thanks! I look forward to your post. This is already my replacement laptop so I hope to not have to return this as well LOL. Thanks for all your help!

 

Sorry but I forgot to mention that since I was in contact with Alienware's Premium Resolution Expert, I brought the overheating issue up (and my 4K backlight bleed) and was told they are aware of the backlight problem but not the overheating. They also asked if I wanted to see if they could send me a new heatsink to try. I might inquire about it further to see if its different.

 

Very nice. We have raised this issue as well through two channels that are pretty deep within the company so lets hope that they hear us....I think few people even sent them a link to this thread. I know for sure that someone at AW (won't mention any names) has seen the pics posted here...

If they are willing to send you a new heat sink, if it's from the same batch from the one you have, it may not do any justice. However, it is worth a try and I'd say ...go for it.

Keep us posted!

Cheers

 

hi, do you guys think the i7 6700HQ AW15 R3 is safe to buy now? i plan to buy with specs i7 6700HQ/1060/8GB/128GB + 1TB HDD/1080P and i live in the Philippines, i would order from a third party which would ship the unit from US to PH. i am concerned about having issues that wont be fixed instantly since this is just a new product and the replacement parts would probably be ordered from US in case any parts of my unit needs to be replaced. Thanks!

 

Thanks! Yes I will definitely update this post for sure. It has been very helpful.

As a side note, do you think I should stick to conventional silicon paste instead of the liquid metal for now?

 

Thanks, glad to hear.

If you are going to be experimenting with the heat sink, use traditional paste for now until you get it right, then you can always repaste with liquid metal later.

 

Okay sounds good. I will go with one of the ones you indicated in your repaste post unless would you recommend one to use?

 

After a repaste I have a differential < 3-4°C.

But prior to the repasting it was more 10-12°C !

I think SOME units have this problem, other it's just bad paste.

 

I wondered about this when others posted they *didn't* see a core differential change from re-pasting, but got an overall improvement across all cores.

It's likely down to how you screw it back together

Did you use any kind of guide for how "tight" to make the screws?

Did you count "turns" as you tightened to get balanced pressure from all sides?

It's likely going to be an "art" to get these "bad" heatsinks affixed down with even pressure to balance the temps between cores.

Any tips?

 

Maybe yes, I used a high torque screwdriver and turned at max (just like I do for any heatsink), I stop turning when I'm about to damage the screw ^^. I used a mroe than usual amount of ICD7 that has a high viscosity, High viscosity can sometimes dampen the effect of a bad pressure heatsink.

after TitanFall 2 yesterday it was 73°, 71°,71°,72°c I think, everything was between 70 and 73°C that's for sure.

 

That's great, glad that worked for you. So maximum turns, at maximum torque, until the blade slips grip - or just before

ICD is also a good idea.

@iunlock - briefly, to what do you attribute your success in getting the heatsink to mate successfully with even pressure applied?

 

Glad to know what solution we can have.
I just receive my laptop.
It's 1070/ 6700HQ/FHD IPS Produce at 1nov.
I'm sure that the purpose is to get balanced pressure from all sides with just 3 screw.
That's my test idle for less than 5min. so bad temp.

 

I guess I should expect the same result from any incoming replacement. Hopefully we get an official response soon. I am making it my mission over the next weeks to see something happen.

 

Forgot to post the results we had back on 10-31 that confirm the findings of the pressure paper tests.. essentially if you have a core temp difference even after repasting then your heatsink isnt making uniform contact. we essentially had the laptop upside down running on the table pulled apart hooked up to a keyboard, mouse, and monitor running prime 95 as I made adjustments to the heatsink in real time. While the stress test was running I was able to close the temp gap to where all cores were at 70-72c under full prime95 load with simple heatsink pressure changes. We were able to achieve these uniform temperatures by loosening the bottom two cpu heatsink screws about 1/2 to 3/4th turn and by slightly bending the top solo retention bracket away from the board to allow the bracket to apply more pressure on the north side of the CPU once the screw is all the way down.

This trick isnt a "fix" by any means. If you want a permanent fix for right now (until Dell/Alienware provides a solution). I would recommend finding some very thin plastic washers to act as a buffer region for the bottom two screws. because if you leave them loose they will eventually back them selves out from vibration and moving the unit.

as I said before this isn't a bad cpu or "hot core" issue. This is bad quality control and poor thermal paste material selection.

 

The possible fix solution is in the OP. I could repost it here?

 

When you look at the actual layout of this CPU. and compare it to the pressure paper. you can see how 2 cores are uneven.
https://www.techpowerup.com/img/15-...ps://www.techpowerup.com/img/15-08-18/77a.jpg

 

trying to steal credit eh? where's your proof?


maybe you should ingest some liquid metal to clear your mind up first and cool down your mind.

 

Have you read the OP? This sounds like a regurgitation 1:1 of the info. already contained in this thread. *confused*

Do you have any documentation, screenshots or pictures of your findings? If so, please post them and make them available for everyone to see.

FYI, loosening the two bottom screws will only cause a more uniform gap, which is not good for most pastes as it'll pump out due to the viscosity, let alone liquid metal....that is a disaster waiting to happen. I do not recommend creating more of a gap by your suggestion.

 

in our case the pressure on the bottom side of the heatsink was to high causing the other side of the heatsink to rise up.

 

Yes, we've discussed this already, but thanks for sharing.

 

Right as TheSandman2236 already discussed on page 9 that we discovered our issue was a heatsink mount issue back on 10/29.

 

Do you have pictures or anything to show such findings? There are things that are just not adding up here. First you claim to have forgotten to post results and then claim that in your case that, "the pressure on the bottom side of the heatsink was too high causing the other side of the heatsink to rise up?"

If what you're saying and claiming is to be true, then you should be familiar with the fact that the pressure that I've posted up is not on the top side, but the bottom, therefore, by you getting something you should be very familiar with having done all these, "tests," then there is no way that you would have gotten that mixed up.

So again, could you please provide pictures of your findings so that we can all see?

 

Have at it: http://imgur.com/a/GlIU5

(See the deadzone areas we mentioned)

 

And your point is? The possibilities of what it could have been is no mystery. It's not rocket science as the choices were: Thermal Paste, Heat Sink, Mobo and I did mention heat sink screw holes back on post #92.

I think you're missing the point here. So again, do you have any pictures, screen shots or anything that you could share with us all?

 

Excellent. Thanks for sharing. So from your tests, where you able to gather the approximate pressure that needs to be applied for that given side?

The pictures of the CPU die is a common trend as it looks similar to what's posted on the OP. What cracks me up is the little magical air bubbles that the Dell toothpaste creates on its own. Now that's magic! LOL

 

http://imgur.com/a/GlIU5

 

how do you read "pressure on the bottom side of the heatsink was to high" and conclude that I am saying the pressure is on the top?

 

I don't own a scale, but I would say it was about 1-2lbs of pressure. I'll let @dadealus chime in

 

I take it that you two are friends?

 

Nah, I just let a stranger come out to my house with McDonald's and take apart my laptop and lay it upside down on my kitchen table

 

Due to your wording perhaps? That's how.

Post #233

You said, "in our case the pressure on the bottom side of the heatsink was to high causing the other side of the heatsink to rise up"

It would be normal to assume that you were referring to the south side of the heat sink where there are two screws, which would cause the north side to rise up.

This is exactly what I was referring to....

 

It's clear that you missed the humor in that as I've asked him to provide something and you were the one that provided it instead. LOL...

Anyhow, good info....let's keep this on topic and not flood the thread.

 

Hmmm that's not much so your heat sink wasn't as bent as mine. I had to apply a pretty good amount of pressure to compensate for the unevenness.
Not surprised as it'll vary ...

 

I placed an order 2 days ago for an i7-6700hq, 32gb memory, 512gb ssd, gtx1070 alienware 15 with the new 120hz gsync 1080p display and am now worried after seeing this thread. Should I cancel the order?

 

If it is on its way, perhaps just wait it out and check your temps upon receiving it. If they are really bad with a core differential of ~30C or something bizarre like that, then definitely send it back to Dell and complain about the issue.

They are fully aware of this issue, hence why they've recently rolled out a band-aid (firmware) in hopes to address the heat issue, but that isn't helping much. This heat sink issue will be something that they will need to address at the factory level and the only way to get this into gear is by our voices and actions of not putting up with a flawed heat sink.

Granted I was able to fix several 15R3's of this issue, but the fact of the matter is that no one should have to deal with the extra work to have a working system.

 

Honestly -- I would cancel and buy the RAM and SSD yourself -- much cheaper. As for this issue, it's hard to say, though we haven't heard of a solution yet.