Madshrimps Forum Madness

Madshrimps Forum Madness (https://www.madshrimps.be/vbulletin/)
-   Articles & Howto's (https://www.madshrimps.be/vbulletin/f6/)
-   -   Arctic Silver 5 (https://www.madshrimps.be/vbulletin/f6/arctic-silver-5-a-3514/)

jmke 27th November 2003 21:36

Arctic Silver 5
 
Thanks to Arctic Silver and ANO Int we are able to bring you an extensive review of AS5.

I will be comparing it to:
- Arctic Silver 3
- Arctic Alumina
- Arctic Céramique
- CoolerMaster Premium (ShinEtsu)

Arctic Silver recommends that the AS5 is "burned" in during a period of ~200 Hours. During which the system needs to be powered down once in a while for the system to cool down and the thermal compound to settle in. AS also states that using good CPU cooling will speed up this process.

What I will be testing:
a) Effect of the "settling" in, when AS5 applied the "old way" (spreading out with bankcard)
b) Effect of the "settling" in, when AS5 applied the "new way", dropped a dot in the center of the CPU and letting the Heatsink spread the thermal compound by the pressure the HS generates on the CPU's die when mounting it.

What material will be used:
- Asus P4C800
- P4 2.4 "C" CPU
- Zalman CNPS7000A-Cu

The P4's heat spreader is lapped to a mirror finish, the Zalman is set to max speed.

The setup:

The system is installed without a case; during testing I'm recording the room temp + CPU temp every other minute to a log file using MBM5. The difference between room temp <> CPU temp is the info we are tracking.

The room temperature can rise high, since the setup is placed on a spot where there is no airflow at all.


Since this is a long-term project I will update this thread of irregular intervals to keep you up to date of my progress. If I had 5-10 identical test systems this article could be finished in less then one week, unfortunately, I do not (yet ;)) have those resources.

If you have any suggestion on the testing methods or what other compounds you want to see compared please don't hesitate to post a comment. Thanks in advance for your input; I hope the end-result will be something of value to you, the reader.

jmke 27th November 2003 21:46

The test-setup, don't mind the box of that Asus ;)

jmke 27th November 2003 21:47

AS5
 
AS5

The Senile Doctor 27th November 2003 21:54

http://www.overclockers.com/articles885/

conclusion?

ceramique better then as5?

jmke 27th November 2003 22:04

he doesn't quite give concrete info, nor how long he "burned" them in. what was the difference between Ceramique and AS5?

if I can continue down the same road of "assumptions" I could now already state that AS5 performs 1°C better then AS Céramique... but that would be far from objective or even close to being "comparable"

look at this graph


check the HIGH fan zalman temp: 31.2°C
the room temp during that test was: 22°C
full article here.

My current room temp at the test setup is: 28°C
the current loaded CPU temp is: 36°C

if we add 6°C to the other room temp, we also get 28°C, if we also add it to the results: 37.2°C

But it would be a 5min hack&slash job to already make decisions and put the article online. it doesn't represent data acquired from the exactly the same surrounding, nor was it tested during several days to actually let the compound "settle in"

Quote:

I noted a reduction even further in my CPU temp with this compound over the AS-5. Arctic Silver could well have jumped on the "XP" (eXtreme Performance) bandwagon when naming this product, because that is precisely the results I've gotten with it. I'm very impressed with this product.
...

The Senile Doctor 28th November 2003 07:25

make you room 20°C please

jmke 28th November 2003 08:19

uhm.. why?

The Senile Doctor 28th November 2003 12:01

coz I think delta t is dependent of initial temp :)

but you're right in your conclusion, you cannot go and extrapolate results coz heatresistance can be different between both coolers

jmke 28th November 2003 12:13

My name is not Liquid3D ;) ? I'm not going to freeze in my room just to have the temperature at 20°C.

As stated the system is set up in a corner, where thereis no airflow, you can compare it to a very big case, the room temp at that spot rises up to 27-28°C during benching. the rest of the room however is at a cosy 21°C. If I want a room/case temp of 20°C, I'll have to open all the windows and sit in a room where the temperature is 15°C.

*brrrr* the thought alone make me all chilly ;)

A-star 28th November 2003 12:44

Or you can add a fan (a big one) blowing to that spot

jmke 28th November 2003 12:52

room is noisy enough with the 2 systems in it :)

The Senile Doctor 28th November 2003 15:12

okay, I get the point, you're not willing to sacrifice your abundant luxury for objective testing !

shame on you :)

jmke 28th November 2003 16:09

Quote:

Originally posted by calantak
for objective testing !
explain to me please, what you mean by that.

If I test different compounds, during an extended period of time, in the same test environment, what part of that testing method is not objective?

Bosw8er 28th November 2003 16:19

;)

The Senile Doctor 28th November 2003 16:32

I meant nothing, JMke,

I was just thinking that it is extremely difficult to test in equal conditions, and that led me to harassing you.

jmke 28th November 2003 17:00

that's why I intend to test for up to 100 hours which each paste and give and overall impression during that period

since I have both room temp & cpu temp during that period every other minute, I can get the overall difference between those 2 and use it to compare it among the results of the other compounds.

that's 6000 results for every test btw.. time-lined et al. :)

The Senile Doctor 28th November 2003 17:22

You are Hard Core, Dude!

jmke 28th November 2003 17:39

cool! can I get to drive to Porsche now? :p

FreeStyler 28th November 2003 20:29

Quote:

Originally posted by Bosw8er
;)
double that

jmke 29th November 2003 11:20

thanks for the support :*

jmke 30th November 2003 14:07

AS5 spreading out using bankcard, room/environment temp calibrated at 30°C, system was shut down twice every day to let the paste "cool" down.

Load Averages:
day1: 36.7°C -- Max temp: 38°C
day2: 36.4°C -- Max temp: 37°C
day3: 36°C -- Max temp: 37°C

Testing now with Arctic's methods, blob of AS5 in the middle and letting the heatsink spread it out.

The Senile Doctor 30th November 2003 18:04

interesting...!

jmke 1st December 2003 22:17

test over 24 hours, full stressed, 2 shutdowns of 10min, temps are no different from the other method at this time. and the margin of error is within 0.5°C.

On to the CoolerMaster Premium, that stuff is hard to spread out! ended up using the same "dot in the middle" method, much easier and same effect!

jmke 3rd December 2003 20:40

CM Premium:
DAY1 36.9°C -- Max temp: 38°C
DAY2 36.4°C -- Max temp: 37°C
DAY3 36.1°C -- Max temp: 37°C

must be said though, that this is a load average, with fluctuations, I will include max reached temps for each day also,

biCker 3rd December 2003 22:20

Quote:

Originally posted by jmke
On to the CoolerMaster Premium, that stuff is hard to spread out! ended up using the same "dot in the middle" method, much easier and same effect!
I spilled a full (!!) tube trying to apply that stuff and ended up shaking and shivering. After the shaking stopped I used the good ol' asIII...

jmke 3rd December 2003 22:40

use the blob in the middle method, fast & effective, as you can see by the results, it equals AS5 performance

jmke 4th December 2003 23:22

AS5:
day1: 36.7°C -- Max temp: 38°C
day2: 36.4°C -- Max temp: 37°C
day3: 36°C -- Max temp: 37°C

CM Premium:
DAY1 36.9°C -- Max temp: 38°C
DAY2 36.4°C -- Max temp: 37°C
DAY3 36.1°C -- Max temp: 37°C

The Senile Doctor 5th December 2003 06:46

how big should that blob be?


they're equal if we see the results...

jmke 5th December 2003 10:11

hard to put in words, so I have taken pictures of the blob, before & after, you can see it spread out very neatly all over the core. online soon

btw AS3 is performing far from bad either, hovering around 6-7°C over environment temperature.

jmke 6th December 2003 12:13

AS3:

DAY1: 36.8°C -- Max temp: 38°C

jmke 8th December 2003 18:20

AS5:
day1: 36.7°C -- Max temp: 38°C
day2: 36.4°C -- Max temp: 37°C
day3: 36°C -- Max temp: 37°C

CM Premium:
DAY1 36.9°C -- Max temp: 38°C
DAY2 36.4°C -- Max temp: 37°C
DAY3 36.1°C -- Max temp: 37°C

AS3:
DAY1: 36.8°C -- Max temp: 38°C
DAY2: 36.5°C -- Max temp: 38°C
DAY3: 36.1°C -- Max temp: 37°C

the maniak 8th December 2003 18:28

seems like all the compounds give the same results, so no need to get something new.

how long can u keep AS3? will it go bad after a while??

jmke 8th December 2003 18:34

I have never heard one "thermal compound gone bad" story when it comes down to AS3.

It will not evaporate like some other thermal compounds out there.

The Senile Doctor 8th December 2003 18:35

so it hasn't changed all that much...

they are all about the same

jmke 8th December 2003 18:36

9 days of testing.. wooo hoo! :/

the maniak 8th December 2003 18:47

one time my AS3 was almost liquid, but that was this summer.

It was over 30° C in my room and the AS3 was lying here somewhere.


All times are GMT +1. The time now is 16:54.

Powered by vBulletin® - Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO