Laptop Update

[Antoine]

A cotton swab could not cause those scratches. Impossible.

Ok then I guess they were already there.. soo even better. I mean seeing small micro scratches on metal is normal right? After all even my cpu which hasnt been touched yet. If light hits it the right way i can make out fine scratches on even it

Btw you say a cotton swab wouldnt cause those scratches. But what about dry thermal paste thats been pushed around on it as its being cleaned? That could be it right?

 

[Antoine]

ok so I replaced both the CPU and GPU fans, and I cleaned the CPU and CPU heatsinks thermal pastes to the best of my ability then applied new thermal paste. I did nothing to the thermal paste of GPU or GPU heatsink (ill do that at a later date as trying to clean the CPU thermal paste look like 2 hours by itself... I coulda replaced 50 fans in that time lol). Then I started it up to see what happens.

No beeps, Windows failed to load, but a simple restart fixed that. The CPU fan revs up to loud and fast for about 5 secs before going back to normal speed... this happens maybe every 5-15 mins. So it sounds MUCH better, and temps are good. My only worry is I may not have applied enough thermal paste. The dot is used was like no bigger than or barely bigger than a thread or maybe close to a grain of rice. (but Im sure it wasnt even THAT big). I would think something that small even with spreading from the heat wouldnt spread very far. But the temps are fine so if I didnt use enough wouldnt I know right now, or is that something I wouldnt know until later on down the line when its been like that long enough for problems to develop?

Speaking of temps here are my current ones while idling at the desktop

 

[Antman]

Good work.

No, dry thermal compound will not scratch copper or aluminum.
Hardness (Mohs scale) Silver, 2.5 | aluminum 2.0 - 2.9 | copper 3.0
Even IF grains of silver the size of those scratches were in the old thermal compound (they were not), those grains would have to abrase the copper first.
It is more likely that an earlier exercise to clean that plate was performed with a metal tool, like a screwdriver.

Point of order:
GPU related heat stress causes substantially more dead laptops than CPU related heat.

Regarding the scratches on the cooling plate -
The plate will likely operate within it's design standard, but that condition is wholly unacceptable. If I found that on my laptop, somebody would be called to the carpet.

If your budget allows, replace the cooling plate.
If not - and you might enjoy this - NOW is the time to use the credit card smearing technique to apply thermal compound to the scratched cooling plate. Completely fill in those scratches.

You could lap it (grind to a smooth finish), but thermal compound filling the scratch is sufficient to the task.

https://www.google.com/search?q=lap...t-3XAhXhTN8KHRn5DiUQ_AUIDCgD&biw=1229&bih=617

 

[Antman]

...The CPU fan revs up to loud and fast for about 5 secs before going back to normal speed... this happens maybe every 5-15 mins. So it sounds MUCH better, and temps are good. My only worry is I may not have applied enough thermal paste. The dot is used was like no bigger than or barely bigger than a thread or maybe close to a grain of rice. (but Im sure it wasnt even THAT big). ...

The CPU fan revs up to loud and fast for about 5 secs before going back to normal speed
100% normal

this happens maybe every 5-15 mins
Not normal, especially if idle.

My only worry is I may not have applied enough thermal paste.
You did not apply sufficient compound. Confirm by removing the cooling plate to see spread.

 

[DOUGIE]

Well done, Antoine.
Like they say, "Practice makes perfect"

 

[Antoine]

The CPU fan revs up to loud and fast for about 5 secs before going back to normal speed
100% normal

this happens maybe every 5-15 mins
Not normal, especially if idle.

My only worry is I may not have applied enough thermal paste.
You did not apply sufficient compound. Confirm by removing the cooling plate to see spread.

well I left it on overnight just this once to see how things work and the revving every 5-15mins thing has stopped and only happens if I open something labor intensive (like the temp monitoring program or a movie). Also how do we know if I did or didnt use sufficient thermal paste? I mean much lower temps and leaving it on over night didnt make thins hotter. Besides if I remove the heatsink to check the spread and the spread is indeed sufficient wouldnt removing the heatsink break the current pastes seal meaning id have to do the cleaning process all over again?

 

[Antoine]

SO Ive also read that "idle temps" on a PC doesnt men much, what matters is the temps when its actually doing something. SO I ran Prime95 to stress test my CPU , and after 15 mins of it the temps stays between 79-88c soo I say things are looking good? >.>

 

[Antman]

if I remove the heatsink to check the spread and the spread is indeed sufficient wouldnt removing the heatsink break the current pastes seal meaning id have to do the cleaning process all over again?

No.

 

[Antman]

how do we know if I did or didnt use sufficient thermal paste?

"We" do not know anything. I only know what you tell.

You said -

My only worry is I may not have applied enough thermal paste. The dot is used was like no bigger than or barely bigger than a thread or maybe close to a grain of rice. (but Im sure it wasnt even THAT big)

A good tech always double checks.

Rule Of Life -
It isn't wrong until it leaves your desk.

 

[Antoine]

"We" do not know anything. I only know what you tell.

You said -


A good tech always double checks.

Rule Of Life -
It isn't wrong until it leaves your desk.

well if we're being honest since I noticed my CPU was rectangular instead of square or round like the ones in most the videos and its smaller (both in length and thickness) than the "Post Reply" button on this page. I determined that a dot wouldnt spread much on that shape (and was also told that unlike a PC CPU where 100% spread isnt a defacto requirement, with a laptop CPU 100% spread is a necessity) decided instead to use an extreme thin line method. SO because of that (and the cooler temps) I have reason to believe it must have spread sufficiently, also due to the small size of the cpu Im almost sure that if I applied anything thicker its 100% guaranteed to run off the silver part of the CPU and only the green parts an thats supposedly bad right?

 

[Antman]

SO Ive also read that "idle temps" on a PC doesnt men much, what matters is the temps when its actually doing something. SO I ran Prime95 to stress test my CPU , and after 15 mins of it the temps stays between 79-88c soo I say things are looking good? >.>

Temps that high are not unusual in a stressed laptop, and are nominally acceptable, but I would not be happy with that result. Sufficiently and correctly applied, thermal compound performance will improve over time.

For CPU stress and heat testing, Prime95 is best used with a BLEND test for two to four hours. Small or Large FFT tests for 12 to 24 hours.
In your situation, I would run a BLEND test for two hours.

FurMark testing of the GPU for 15 minutes is sufficient. I caution against a FurMak test until after you address the GPU cooling. Again, GPU heat kills many more laptops than CPU heat.

 

[Antman]

well if we're being honest since I noticed my CPU was rectangular instead of square or round like the ones in most the videos and its smaller (both in length and thickness) than the "Post Reply" button on this page. I determined that a dot wouldnt spread much on that shape (and was also told that unlike a PC CPU where 100% spread isnt a defacto requirement, with a laptop CPU 100% spread is a necessity) decided instead to use an extreme thin line method. SO because of that (and the cooler temps) I have reason to believe it must have spread sufficiently, also due to the small size of the cpu Im almost sure that if I applied anything thicker its 100% guaranteed to run off the silver part of the CPU and only the green parts an thats supposedly bad right?

You are probably in good enough shape. All doubt is addressed with visual inspection. Next time.

 

[Antoine]

Temps that high are not unusual in a stressed laptop, and are nominally acceptable, but I would not be happy with that result. Sufficiently and correctly applied, thermal compound performance will improve over time.

For CPU stress and heat testing, Prime95 is best used with a BLEND test for two to four hours. Small or Large FFT tests for 12 to 24 hours.
In your situation, I would run a BLEND test for two hours.

FurMark testing of the GPU for 15 minutes is sufficient. I caution against a FurMak test until after you address the GPU cooling. Again, GPU heat kills many more laptops than CPU heat.

well as for addressing the GPU cooling is we recall the cooling issue was with the CPU to begin with, not the GPU. You just said If Im gonna replace one, may as well replace both, so as far as we both know there is no GPU cooling issue that NEEDS to be addressed at the moment. It just highly recommended by you. Right?

Also "Temps that high are not unusual in a stressed laptop, and are nominally acceptable, but I would not be happy with that result." Now unless Im interpreting this wrong. That tells me those temps are perfectly fine, theyre just not good enough for YOU. Surely you notice the difference in those two things right? Some people are ok with "perfectly fine" or "within acceptable parameters"

 

[Antman]

You have received professional guidance. What you do with it is a choice.

 

[Antoine]

You have received professional guidance. What you do with it is a choice.

indeed I have, and I am thankful for it. But Im also taking in consideration the fact that the person giving said professional guidance also said "I am often guilty of detailing how to build a watch when the question was 'what time is it'." Which unless Im wrong means you also have a penchant for overdoing things or wanting to do far more than whats necessary, am I correct in my interpretation of those words?

 

[Antman]

Your perspective is not sufficiently broad. These threads appear on web searches. The audience of interest has a population greater than 1.

 

[Antoine]

so I just tested the most graphically intensive game I have (in fact its the only game I have) which used to run on my laptop at MAX PC (yes PC not laptop) settings just fine on day one. That over the years got so had that I have to play it on standard settings to that it isnt stuttering and dipping below 15FPS. Now with the CPU thermal paste application and net GPU and CPU fans, the game is running at MAX PC settings just fine again while the CPU temps stay between 40-50C and the GPU gets no higher than 93c the FPS (which goes above 60 on lower settings, stays between 32-28 on max... which 95% of the time stays at 30+) what would you say of those temps and performance?

 

[Antman]

At 93C, you are within 2C of the max operating temperature of the 780M chip design reference. Any improvement in GPU cooling will also improve CPU cooling, and vice versa. It is a system.

My goal is to assist in the diagnosis and repair of technical issues and to impart a thorough understanding of the relevant principles and dynamics involved. I am not here to proselytize, but I will argue that you may be pursuing the wrong goal. Frame rates and game settings do not matter - they are free and inconsequential and will not fail or fry as a result of heat stress and damage. But, your laptop will. This thread, and its related thread, is about heat related systems failure - not gameplay.

 

[Antoine]

At 93C, you are within 2C of the max operating temperature of the 780M chip design reference. Any improvement in GPU cooling will also improve CPU cooling, and vice versa. It is a system.

My goal is to assist in the diagnosis and repair of technical issues and to impart a thorough understanding of the relevant principles and dynamics involved. I am not here to proselytize, but I will argue that you may be pursuing the wrong goal. Frame rates and game settings do not matter - they are free and inconsequential and will not fail or fry as a result of heat stress and damage. But, your laptop will. This thread, and its related thread, is about heat related systems failure - not gameplay.

Well if highest temp is 95 and im hitting 93 when gaming. And the cpu paste plus cpu and gpu fan has been replaced. That means the gpu shouldnt get THAT hot regardless to what game is being played.

Which means its confirmed that the gpu paste NEEDS to be replaced. Its no longer a suggestion at this point. See how easy that was? now i actually have an indisputable reason to do so.

 

[Antoine]

Btw i just learned that "thermal pads" are a thing as well. Will i have to worry about/change those too or just the paste?