[Final Question] About Repaste - AW15R3

so guys, after waiting for a while and reading the topics, i decided to finally repaste my AW15R3. Yesterday i was playing Albion, and all my cores reached 99-100 degress, which is very very high for this kind of game. I immediatly turn of notebook.

Since im very newbie at this, and i'll aks to a friend to do the process, and i live in Brazil, i would like to know which materials should i buy to, that will be the most effective.

I saw the topic of @iunlock and at the images there are a lot of material such as conductonaut, kryonaut, artic mx-4, etc. Also saw some post of @Mobius 1 i guess, and he posted:

.

Since i really dont undertand everything, im posting here in hope of someone to make me a little checklist of the things i need to buy for the repaste.

Sorry for creating a new topic for this guys, and sorry for the newbie

 

If this is your first time doing a repaste I'd advise to stick traditional paste. Also, refer this post for recommended paste Thermal Grizzly Conductonaut vs Grizzly Kyronaut vs Arctic Silver 5

 

1 tube of non conductive paste [GC Extreme / Kryonaut] this is easy to apply because you do not need to worry about safety, the paste is not conductive
if you have an HQ chip, you do not need the liquid metal paste, ever

I believe unlock has the thermal pad guide somewhere for the 15R3 thermal pad thickness

 

thanks for the link

so 1 tube of grizzly kryonaut do the job right?

i find here on a store next to my place, thermal gryzzly HYDRONAUT....what is the difference?

Also, how do i know if i have a HQ chip or not? i really dont understand those thecnique terms, sorry.

this thermal pad guide of unlock is necessary, to know the thermal pad thickness? Also, i was reading again his topic ( http://forum.notebookreview.com/thr...5r3-disassembly-repaste-guide-results.797373/ ) and there is a section that he recommend replacing the stock thermal pads with Fujipoly. Do you think this is necessary, or only the repaste and im okay?

 

If stock thermal pads are in good condition then no need to bother with it. In the future, you might want to replace thermal pads to get decrease in temps.

 

you can see the cpu type on task manager

the fujipoly is better than arctic, but it's also much more expensive, choose it if you don't mind spending more money

as always, be careful when handling thermal pads to not touch them directly with fingers


Kryonaut is better than Hydronaut, you will get slightly lower temps.

I only recommend that you replace thermal pads on the black mosfet/VRM chips alongside the VRAM chip, anything else is kinda pointless

 

thanks for you both guys. Here is my SS:



Just to resume, so i can make sure if i understood everything hehe:

1) I only need to change thermal pads IF they are in bad shape right? if not, leave it as it is, right?
If they are bad, ill trade them. Think will choose fujipo, as i dont mind spending a little more money .

With this part :

Do you recommend replacing this thermal pads, even if they are OK? Could you post here some screen shot so i can identify this "black mosfet/VRM and VRAM" ?

2) TG Kryonaut right? ill have to look more or import to Brazil, since i only found the hydronaut.


Last question (i promise): do you guys know if Amazon or ebay or some security website export those products to Brazil (and also if i can find them at Amazon)? Thanks

 

With 6820HK I recommend Liquid Metal, such as Thermal Grizzly Conductonaut



I take apart my 15R3 again because of you.

Red = important thermal pad
Yellow = remove the thermal pad and put nothing











Red = important thermal pad
Blue = remove the thermal pad and put nothing

The red ones are GPU memory and system power delivery chips.

 

woaaa mobius big thanks mate !! you shouldnt have all this trouble for me. Omg thanks mate. Blue and Yellow pads i can remove and dont have to replace it, and red ones i should replace?

Also, as you said, since im 6820 hk i should use the liquid one? isnt it too risky?

thanks mate ! big big thanks

 

Well the liquid metal is preferable for the HK CPU if you want to overclock.

When applied properly there should be no risk.


Red = replace

blue/yellow should be removed

 

TG Kryonaut is no slouch either. Use standard paste and overclock/tweak the CPU and GPU clocks. If temps are worse during OC, you might want to try LM. If you sure about LM, read the article I linked above because there are some prerequisites.

 

ok thanks for all the help you 2 guys provided me. Also: after repasting, how do i do overclock, in order to obtain maximum experience?
I mean, is there any risk by overclocking? Do you guys did or recommend for playing games? thanks

 

ok thanks for all the help you 2 guys provided me. Also: after repasting, how do i do overclock, in order to obtain maximum experience?
I mean, is there any risk by overclocking? Do you guys did or recommend for playing games? thanks

P.S> Do you guys used Liquid metal or standar paste?

 

I am working on an overclock guide for the HK-CPU Alienware. Will post soon I hope.

 

thanks mate

 

is there a big difference between liquid metal and kryonaut? because since im newbiw, i think normal paste is more "easy " and less "risky" to apply right?

i mean, if the difference is huge, ill go for the liquid metal.... if not (like 2-5 degress), ill stick to the normal one.


p.s: how do i overclock and is there any risk to my computer?

 

Metal to metal contact dissipates heat way faster than traditional paste would do. I'm sticking with traditional paste because my PC can't OC its locked. My advice for you is apply Kyronaut and see/post the results at stock clocks. After that, tweak your CPU and OC to 4Ghz by setting multipliers to 40x and also increase the core voltage so that enough power is supplied to CPU.
If you see CPU temps in 80C @4GHZ then LM is your best bet.

 

good info, so should one go all out and just do LM, that was the reason I started the other thread ? I do have an HQ and despite that there won't be any overclocking, why should you go with regular instead of LM? Is the risk just greater ?

 

Here you go this might help you

 

Well iunlock has shown Kryonaut gets temps near-equivalent to conductonaut. I'm not sure if that is due to just doing it perfectly, or maybe because first he used conductonaut and it stained into the copper (potentially filled in any imperfections, like lapping).

 

thanks for the vid

so the traditional paste (kryo) is almost the same result as LM??? i guess ill stick to the regular one so, since there are risks with LM.,

 

iunlock has A LOT of experience with laptops and repasting. A lot might even be a understatement. I repasted my 15R3 from conductonaut to Kryonaut and I plan to compare the temps on both of them. I'll try to do it in the next 1-2 hours, not only for myself but for anyone else that's interested

 

Interesting just bought a tube of conductonaut, plan on recasting wish me luck

 

Would be interesting to see Kryonaut to conductonaut and then back to Kryonaut to see if applying Kryonaut after liquid metal makes a difference compared to before.

I assume gloves are fine and the issue you are tying to avoid is getting oil from your hands on the pads, is this correct?

 

Yep that's exactly why gloves are recommended.

Alright I did a test going from stock, to conductonaut to Kryonaut but I am pretty sure I need to go back to 1.0.8 from 1.0.12 and do the test, here are the results -

Tests are Intel XTU CPU stress test + 2 runs of Heaven back to back, no closing or pausing either program :

7700HQ + 1070 (Dell Paste):
idle gpu - 50
idle cpu - 48, 49, 48, 48
load cpu+ gpu - 90, 89, 84, 84 gpu: 73

7700HQ + 1070 (Conductonaut): no copper shim
idle gpu - 40
idle cpu - 40, 41, 40, 40
load cpu+ gpu - 81, 79, 75, 76 gpu: 69

7700HQ + 1070 (Kryonaut): 0.1mm copper shim on CPU + GPU
idle gpu - 37
idle cpu - 36, 37, 37, 37
load cpu+ gpu - 74, 71, 67, 67 gpu: 58

I know there's a big difference in the Conductonaut and Kryonaught... it may be due to the conductonaut being BIOS 1.0.8 and the Kryonaut being 1.0.12 . 1.0.12 seems to stutter a lot during the test, but checking Afterburner shown it was still being 100% stressed. Also I am unsure of if the Conductonaut had a undervolted GPU at the time of doing the benchmark but I am 100% positive the CPU undervolts are consistent.

 

Conductonaut will make biggest difference when you start overclocking an HK chip, HQ chip see minimal difference.

 

Yeah I thought I did my conductonaut application wrong and planned on redoing it, but when removing it it seemed very much still liquid after 2-3 months (so it was likely fine). Ended up putting on kryonaut since I was often worried about putting the laptop in my bag and taking it places.

 

So it's not worth doing conductonaut on us chip? Should I go the Kryonaut route?

 

The copper areas on the AW heatsinks are nothing like 80+$ closed loop cooler heatsinks, where they're nice and polished. In my opinion, the conductonaut likely helps even if you apply it and used it for a while, and then remove it. I mean I have no proof, but there has to be SOME reason that stain is impossible to get off without sanding it off right? You could tag alcohol to it for hours and it'll still be shiny silver.

tl;dr, if you plan to take your laptop to places often, just use Kryonaut. If you don't, use Conductonaut + the 33M Tape. The tape idea is genius and easy to remove but effective at keeping the LM from shorting anything. If your laptop stays on a desk, the LM won't go anywhere

 

I do move it a lot so I guess I better return it and get kyro then

 

Well the whole point of the tape is to protect it from LM if it does leak/spread when moving. But idk, even with the tape, I often thought about 'what if' my 2000$ laptop got damaged just from using LM lol. I like the temps + peace of mind of Kryonaut

 

thanks all for the debate guys. If i overclock, am i able to revert it, in future, if i want?

 

The comments on the effects of the liquid metal on the heatsink after being "removed" make me wonder if applying liquid metal in place of lapping would be a good idea for users who for whatever reason do not or can not use liquid metal as the actual TIM.

Basically you just apply some on the heat sink and then wipe it off (leaving the stain behind and hopefully a smoother surface with it) and then applying Kryonaut as the TIM.

 

Yes you can.

 

Well this is what happened when I went from conductonaut to kryonaut. It wasn't the reason I removed conductonaut, but if you search around the internet you will see people have been talking about LM for many years. The LM is supposed to be pretty much fine with copper, definitely can't be used on Aluminum. There's others that can provide more info but there is still a reaction between copper and conductonaut. And besides just think about it - if you poured liquid metal in holes and wiped off the top layer, the LM that filled the holes will remain. It's too much work to test if it has any positive effects but I can't see there being any negative. I also doubt you can just paint it on and then wipe it right off, I would think it has to kinda cook/seep/melt inside a bit

 

Hey guys,
quick question here...
I just re-pasted my 15R3 for the second time and did the little mod to heat sink where you slightly bend up the arm thingy to increase the pressure and ensure proper contact with the CPU.
Now, my ambient temps are between 27-33 Celsius and the mod also helped reduce the temp diff from sometimes 10-15 degrees to around 2 degrees Celsius.
This is all good. The issue I'm having is that the CPU temps tend to get pretty high (high 80's to low-mid 90's) when gaming. These are spikes and are not constant values.
Is this to be expected since these laptops are built to be slimmer now?

They dont build them like my old M15X I tells ya!

 

What are you temps before and how did you bend the heatsink?

 

Your temps seem a bit high for the 15. Mid 80s should be the most I'd tolerate for a 15R3. Of course, ambient temp plays a factor too, so your conditions may be warmer than mine. Though the cooling is very good on these. Seems like you don't lift the rear end up when gaming, that should drop it about 5-10C. You can also disable turbo with throttlestop and undervolt. I haven't ran into a game that needs more than the 2.8Ghz and that'll drop temps a lot, also by about 5-10C. You can undervolt the GPU which I tested recently and it lowers the GPU temp by 5C for me. The CPU and GPU share the same heatsink so reducing wattage/load on one, reduces the temp of the other by nearly the same amount

 

Quick question: besides repaste, you recommend to undervolt? How do i do that? Which program?

Question 2: do you guys know the paste from GC Extreme Gelid Solutions?
Thats the only one ive found here on BR....also found the artic silver 5

 

Throttlestop to undervolt the CPU and MSI Afterburner to undervolt the GPU.

http://www.eteknix.com/thermal-grizzly-thermal-paste-vs-6-major-brands-review/3/
GC Extreme Gelid should be just as good. I say should be, cuz this is a laptop, but it has about 8.5 thermal conductivity VS the 12.5 for Kryonaut.

 

I did undervolt the GPU, for games FPS is consistent @60 but during benchmarks the UV settings cause my scores to decrease. I've 15 R2. I downlocked the core and memory clocks by 100MHz and 200MHz respectively.

 

Oh yeah the score definitely drops a bit, even some games as well. I tested 3-4 games at 1080p and 4K. Basically the minimum, average and max framerates dropped by 1-2 @ 4K. And at 1080 the minimum and average also dropped by 1-2, except the max dropped by as much as 20 sometimes. Keep in mind @1080 the framerate would be around 180.. so a loss of 20 would make it 160, which is still enough for my 144Hz screen and anybody elses.

I prefer to lose some points and fps for longevity and quiet gaming. However if I run this undervolt with +500memory clock, I think it will come close or do better than stock, while being just as stable and likely still generating less heat

 

Fresh test done 20min ago, since admittedly all of my Heaven benchmarks were done with Intel XTU stressing the CPU.
Stock -


0.800mV, 1518Core, +500Mem -


I can guarantee without checking that adding +500Mem doesn't increase the GPU wattage by 30W too haha. What I can't gurantee, is that it is safe for the memory. Many people overclock desktop GTX 1070 memory around +700, but this is a laptop with higher temps. iunlock is going to do thermal tests sometime, which will be awesome to see the difference between stock mem and overclocked memory chips.

Also, its a pain to undervolt using afterburner, I am sure the core clock can be raised a little bit, this was just something quick I went down to, when bored enough I plan to play with it more

 

@illuMinniti: Did you run Heaven and XTU at the same time? If it was 15 r2, my temps would be around 80 or 90C.

 

Oh no, those two test were without XTU and that was with turbo disabled. With XTU and turbo enabled these are my temps -
7700HQ + 1070 (Kryonaut): 0.1mm copper shim on CPU + GPU
idle gpu - 43
idle cpu - 36, 37, 37, 37
load cpu+ gpu - 74, 71, 67, 67 gpu: 58

The idle can actually be in the high 20s for CPU and GPU if I start up my PC and just let it idle lol

 

Temps before repaste were mid 30's to low 40's at idle.
the bending part is from a post here where you slightly bend the bracket/arm that all by itself (for the CPU there only 3).

 

Using a cooler master pad and I stuck on small rubber feel to the bottom to lift it up by an extra inch
I am currently undervolting the CPU by -150 (stable)

GPU temps never go above 65-68 while gaming from what I've seen.

I’ve never tried to undervolt my GPU. How is this done and what are the possible implications of doing so?

 

I see, wondering what Bios you are running?

 

What Bios are you running?

 

Honestly you might see more of a drop with the laptop simply being raised, no special fans underneath or anything. I haven't actually tried a cooling pad, but I did a lot of searching around and came to the conclusion that they added nothing or minimal temp drop, meanwhile being bulky and possibly noisy (depending on which you get).

There's about the same implications as with undervolting a CPU, none if you do it right, or a unstable system. However with the GPU, you will simply get artifacts and driver crashes if its unstable. The screenshot I posted pretty much has all the info, there's lots of guides on YouTube but for whatever reason 99% are non-English lol. Pretty much just drag and move those dots. You can also use the keyboard up and down arrows to move the dot slowly.

1.0.12 I think, whatever the newest one is