Jump to content

Thermal 'Compound' Experiment

Hey,

 

I have a old crappy laptop (Dell Studio 1558) and I got wondering the other day, how good is toothpaste as a thermal paste? And as I had this crappy laptop doing nothing I decided to find out.

 

First a few things about the laptop that I performed the tests on.

 

1) SERIOUSLY CRAP

2) Intel I7 720QM inside, 4GB 665MHz ram, Scoprio black HDD 

3) Nearly 5 years old, never had a thermal paste change.

4) Its crap.

 

Alrighty, so to perform the test I ran Piriform Speccy to keep an eye on my temps, and used burnintest to give my laptop a work out, all recordings of temps was made by me sitting there reading the temperature off Speccy every 30 seconds for 5 mins.

 

Initial Test - Stock Compound

 

Prepared to be horrified.

 

At idle my temperatures weren't good to start with. - 75DC 167FH

 

Under load I maxed out at 100DC 212FH (see figure 1) 

 

And my average load temp was around 97DC 207FH

 

post-56584-0-53986600-1422914347_thumb.j

Figure 1

(Note: Look at mobo temperature.....)

 

Test done.

 

Application of TP

 

For my new choice of  thermal compound Toothpaste I used the one and only: Colgate Max White - With whitening strips. 

 

When I came to apply the stuff I was frankly shocked at the amount of TC Dell had used, look at the photo below (Figure 2) showing the CPU straight after I removed the heat sink.

 

post-56584-0-06755300-1422914817_thumb.j

Figure 2

 

NOTE: I used Petrol mixed with 2 stroke to clean this, it worked fantastically!

 

 

post-56584-0-25642900-1422914952_thumb.j

Figure 3

 

The 'cleaned' product - I CBA to clean the bits off the contact area

 

NOTE: Pea method used to apply Toothpaste

 

Toothpaste Test

 

Be suprised! 

 

Max load temp recorded - 78DC 172FH

Average load temp -          76DC 169FH

Idle temp -                         58DC 136FH

 

The graph of temps

post-56584-0-62449100-1422916111_thumb.j

Figure 4

 

Let me be the first to say it, WOW! I expected something to be different but a change of 19DC average? Dam!

 

Removal

 

At removing the paste I found the stuff to have gone real sticky, it still smelled minty and fresh, but I wouldn't recommend scarping a toothbrush off it. 

 

post-56584-0-33382000-1423093462_thumb.j

Figure 5

 

 

NOTE: To remove TP I used WD40, cos why not? Worked well, but not quite as well as the petrol mix from above.

 

 

 

Steroid Eczema Cream Application.

 

So yeah, had a tube sitting next to me and a good hour of free time, so what better use of my time than to experiment with this stuff.

 

Applying the stuff was far harder than Toothpaste, using the pea method I accidentally applied too much due to the extra pressure required to get the stuff out of the tube, and then as I tried to remove the excess of the stuff everything just got messy and sticky....  

I mean look at this terrible job!

 

post-56584-0-12062300-1423093718.jpg

figure 6

 

SEC Test

 

NOTE: The toothpaste test was done a week after some other tests, these results are straight after and one or two test runs!

 

At idling the temperature averaged at 51DC 124FH

Max Temp under load came in at 76DC

Average temp under load = 73DC 163FH

 

 

 

post-56584-0-56635700-1423094330_thumb.j

Figure 7 - Sorry about flash messing image up, didn't think about it at the time

 

 

Pretty good if you ask me, however writing this a good few hours after the tests I can already start to smell the stuff, so onto the next test. 

 

 

TL;DR?

 

Me change Thermal paste to Toothpaste, then Steroid Eczema cream. Look at graph (Figure 7)

 

post-56584-0-49894100-1423094556.jpg

Figure 8

 

 

 

 

If this thread gets enough attention / if I get bored I may do other household goods to see what they can do, so anyone got any recommendations of thermal paste?

 

 

 

EDIT: Oops, forgot to add picture of thoothpaste temps! Doing that now!

EDIT: Added information regarding test proceedure

EDIT: Added TC experiment and data

Link to comment
https://linustechtips.com/topic/303235-thermal-compound-experiment/
Share on other sites

Link to post
Share on other sites

Nice! Going to try this on my 3570 /s 

CPU: Intel 3570 GPUs: Nvidia GTX 660Ti Case: Fractal design Define R4  Storage: 1TB WD Caviar Black & 240GB Hyper X 3k SSD Sound: Custom One Pros Keyboard: Ducky Shine 4 Mouse: Logitech G500

 

Link to post
Share on other sites

Technically, any pasty material will work.. It's all a matter of how long it will work.

 

Some site (I'll try to find a link) did a similar test using even more random stuff like mayonnaise and ketchup. It all works, though some stuff starts to smell really bad and then lose performance fairly quickly. 

Edit: damn, I really can't find the article in which all those random pastes were tested. You'll just have to take my word for it...

i7 not perfectly stable at 4.4.. #firstworldproblems

Link to post
Share on other sites

This is deffently interesting I might try this on the Toshiba Ultra book I have that sits around 80C idle and load of any type is 100C

|King Of The Lost|
Project Dark: i7 7820x 5.1GHz | X299 Dark | Trident Z 32GB 3200MHz | GTX 1080Ti Hybrid | Corsair 760t | 1TB Samsung 860 Pro | EVGA Supernova G2 850w | H110i GTX
Lava: i9 12900k 5.1GHz (Undervolted to 1.26v)| MSI z690 Pro DDR4| Dominator Platnium 32GB 3800MHz| Power Color Red Devil RX 6950 XT| Seasonic Focus Platnium 850w| NZXT Kraken Z53
Unholy Rampage: i7 5930k 4.7GHz 4.4 Ring| X99 
Rampage|Ripjaws IV 16GB 2800 CL13| GTX 1080 Strix(Custom XOC Signed BIOS) | Seasonic Focus Platinum 850w |H100i v2 
Revenge of 775: Pentium 641 | Biostar TPower i45| Crucial Tracer 1066 DDR2 | GTX 580 Classified Ultra | EVGA 650 BQ | Noctua NH D14

Link to post
Share on other sites

I'm still tempted to see what happens if you used JB weld.

 

Permanently affixed, stuff basically turns to metal, applied thin enough it could work I think.

Ketchup is better than mustard.

GUI is better than Command Line Interface.

Dubs are better than subs

Link to post
Share on other sites

Technically, any pasty material will work.. It's all a matter of how long it will work.

 

Some site (I'll try to find a link) did a similar test using even more random stuff like mayonnaise and ketchup. It all works, though some stuff starts to smell really bad and then lose performance fairly quickly. 

 

I did this over a week ago, got the laptop fired up next to me now and sofar it doesn't smell, I know a week isn't much but considering all I have been doing is benchmarking and tests it has been under some load. (Did the test today)

 

 

I have been planning on comparing MX-4 to toothpaste to mayonnaise on my Athlon X2 which gets plenty hot

 

Ha, let me know how it goes!

 

I'm still tempted to see what happens if you used JB weld.

 

Permanently affixed, stuff basically turns to metal, applied thin enough it could work I think.

 

Looked into it, gunna cost me a fiver which frankly I need for food right now (Unfortunately I'm a Hard up student...) But I am also tempted.....

 

 

Try lithium grease (automotive/hardware store)

 

I'll look into it, if its cheap enough I'll try to get some. 

Link to post
Share on other sites

I did this over a week ago, got the laptop fired up next to me now and sofar it doesn't smell, I know a week isn't much but considering all I have been doing is benchmarking and tests it has been under some load.

 

 

 

Ha, let me know how it goes!

 

 

Looked into it, gunna cost me a fiver which frankly I need for food right now (Unfortunately I'm a Hard up student...) But I am also tempted.....

 

 

 

I'll look into it, if its cheap enough I'll try to get some. 

Be warned, JB Weld is absolutely permanent.

 

You will likely not be able to get your motherboard back.

Ketchup is better than mustard.

GUI is better than Command Line Interface.

Dubs are better than subs

Link to post
Share on other sites

The laptop isn't really very useful in its current state, too many issues, so probably after my fun here I'll destroy it........

Idea:

 

Case made entirely from molded JB weld....

Ketchup is better than mustard.

GUI is better than Command Line Interface.

Dubs are better than subs

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×