Jump to content


'The Test'...


97 replies to this topic

#1 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 27 November 2007 - 08:56 PM

Well...I grew some mini 'stones' and decided to see whether or not the eeePC can be ran safely at stock speeds without active cooling:

The Test

I had no desire to void the warranty, so I decided to insert a piece of thin, yet difficult to flex cardboard in the middle of the fan vent to stop the fan from being able to rotate. The cardboard was added while the eeePC was in standby.

After resuming from standby I heard a not-so-quiet 'clicking/clunk' sound every 5 seconds coming from the fan. Since the blades of the fan were not able to move, this sound was definitely from the motor's ball bearings which I now know is clearly responsible for the 'buzz/grind' noise we would hear when the fan is spinning (in a room with low ambient noise).

My first objective was to max out the CPU usage for roughly an hour, and observe the increase in CPU temperature. This was accomplished by doing the following (somewhat overkill):

- Three browser windows each displaying a flash based TV show from adultswim.com

- Console window running the 'updatedb' command two times

- Console window running the 'top' command to monitor CPU usage (set to update every second)

- Console window running a script that outputs '/proc/acpi/thermal_zone/TZ00/temperature' (CPU temperature) every second


The Meaningful Data

- The highest CPU temperature reached without any active cooling was only 66 degrees Celsius, which is no where near the 90-100 Celsius critical 'CPU power down' temperature.

- This temperature was reached within 15 minutes of maxing the CPU and did not increase at any point during the 1 hour max CPU usage test period.

- The top and bottom of the case were both warm, but nowhere near 'hot'. I was able to comfortably rest the device on the bare skin of my upper thigh for a short portion of the test.


Idle CPU Temperature 30 mintues after The Test:

- The CPU temperature hovered between 54 and 55 degrees Celsius.

- The case was mildly warm.


What does this all mean?

Irregardless of CPU load, active cooling on an eeePC running at stock speed is unnecessary


Pros of Fan Removal

- In places with low ambient noise you won't have to deal with the buzzy sounding fan noise caused by the bearings in the fan motor.

- Without the fan, you lower the eeePCs power consumption by roughly 1W.

As noted in one of my earlier posts once the fan is on, the fan is incapable of cooling the system down low enough to reach the fan 'off' temperature trip point, because of this the fan never turns off.

Cons of Fan Removal

In order to open the case and disconnect the fan you will void your warranty by damaging the warranty sticker.


My Dilemma

After having completed this test, I REALLY, REALLY want to disconnect the eeePC's fan, but I won't be doing so until I find out that someone who has opened their eeePC has successfully received Asus warranty repair service for some kind of defect/issue.


Notes:

I have removed the cpu_freq and p4_clockmod modules from my system as I deem them to be unnecessary. My doing so should have had no impact on the observed CPU temperatures.

The battery monitoring application 'powermonitor' appears to use 15%-25% CPU usage every 10 seconds while it checks the battery status. I plan to remove this app from '/usr/bin/startsimple.sh' and see if the eeePC battery life noticeably increases.

Edited by NoeeePC4MeWellMaybe..., 27 November 2007 - 09:11 PM.


#2 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 28 November 2007 - 04:25 AM

I expected some of the few 'Silent PC mod' fans (no pun intended) on the forum to be jubilant, but I guess the thought of voiding a 2yr warranty acts as sedative. Then again, the fan isn't really that loud to begin with... :/

#3 TacticalPenguin

    Senior Member

  • Members
  • 639 posts
  • LocationNear an EEE

Posted 28 November 2007 - 04:38 AM

I know i sound like an ass when I say this but it bugs me. Irregardless is not a word, replace it with regardless and you'll stop sounding as if you're trying to look smart.

Oh, yea, the topic. Sounds pretty cool but I think ASUS doesn't want to take any risks until they get a specifically low-voltage low-temp cpu made to be run fanless.

Edited by TacticalPenguin, 28 November 2007 - 04:39 AM.

EEE Fund: Who cares I finally got it 4/16/08
I got an eee after creating the "post here when you get your eee" thread, which had over 1450 posts in it, making it the forum's largest, before I could post in it. Also, XP SP3 + 8GB transcend SDHC card + 405-999mhz eeectl

#4 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 28 November 2007 - 05:24 AM

LOL, wait.. did use LOL as an acronym, or a widely accepted word among peers :D...I never knew my posts were going to be scrutinized, wait is it scrutinized or scrutinised :| for grammatical errors..and god, wait, God knows I've made a bunch and am..oops..I am likely to make more in the future so please bear with me :D

The processor in our eeePC is only 5W TPD, such a processor can be easily cooled passively with a decent heatsink. However in our eeePCs the heatsink, so to speak, appears to be the bottom of the keyboard tray :lol:. Regardless (happy ? :)), even with this type of heatsink the operating temperature is still very low.

Asus probably included the fan when they originally intended to run the processor at 900Mhz. However, I believe that after they decided to bump the processor speed down to 630Mhz they never re-tested whether or not the thermal design required a fan, When you factor in the CPU fan that comes on at a certain temp but then never shuts off, it's not far fetched to assume that Asus made a few oversights here and there.

#5 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 05:25 AM

I'm not comfortable with disconnecting the fan. Although the eee's cooling design seems to be very primitive, it is possible that components other than the cpu are helped by the fan. I've also noticed that sometimes when the cpu load is minimal for a while, the fan _will_ shut off.

With the eee 701's success, let's hope that next year's model will have a higher rez screen, a larger SSD, modern full-featured power management (like apple laptops, or even an 11 year old Toshiba portégé), no "warranty" sticker on the memory access panel, and perhaps heatsinks/heatpipes with a very low noise fan. :D

#6 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 28 November 2007 - 05:52 AM

Imabuddha (and/or others), can you do me a favour? The next time your fan comes on, please record the CPU temperature that triggers the fan in your eeePC to shut off. I have yet to reach or discover what that temperature is.

After the CPU fan has started in my eeePC, the lowest temperature that it has been able to bring my CPU temp down to is about 47 degrees,my ambient room temp is 24 degrees celsius. As other's have noted, I can get the CPU fan to shut off for a few minutes by placing the device in Standby and then waiting a few seconds before resuming, or by shutting it down, and turning it back on again.

Prior to the test, the CPU is the only component in the eeePC that would possibly require active cooling. After observing the CPU temperature when passively cooled, I am 100% positive the 50-66 degree celsius core temp range would not have an ill effect on the surrounding components.

On a side note, I for one have no plans on upgrading to the next major revision of the eeePC ;)

Edited by NoeeePC4MeWellMaybe..., 28 November 2007 - 06:05 AM.


#7 Daijoubu

    Advanced Member

  • Members
  • PipPipPip
  • 56 posts

Posted 28 November 2007 - 08:10 AM

How about running a 3D game? Was wifi on?
The chipset and other parts should also adds to the heat

#8 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 09:14 AM

Quote

Imabuddha (and/or others), can you do me a favour? The next time your fan comes on, please record the CPU temperature that triggers the fan in your eeePC to shut off.
Sure, I'll give it a try.

Quote

Prior to the test, the CPU is the only component in the eeePC that would possibly require active cooling. After observing the CPU temperature when passively cooled, I am 100% positive the 50-66 degree celsius core temp range would not have an ill effect on the surrounding components.
2 things:
-prior to the test (with the computer "cold") no components require active cooling by definition, right?

-I believe that the 50-66ºC temp range won't hurt the cpu, and perhaps its heat doesn't affect other nearby components much. However, the other 2 big chips on top of the motherboard (and parts underneath, like the memory) are also generating heat, and unless you insert some probes into the case (or use an expensive properly calibrated thermal camera) we won't know what their temperatures are, nor are we certain how well they will deal with even less cooling than the fan provides over long term usage.

It's interesting that Asus have used the thin metal plate under the keyboard area as the heatsink. This does reduce cost and size, and seems to work well enough.

Quote

On a side note, I for one have no plans on upgrading to the next major revision of the eeePC ;)
I hear ya, but one of the nice things about the eee's low price is that for the cost of a typical ultralight you can buy the latest eee for the next 3 years... :)

#9 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 28 November 2007 - 09:23 AM

Yes, WiFi was on during the test. A few minutes after the 1hr test, I played Penguin Racer for about 15-20 minutes, and the case did not get noticeably warmer during that period. Since the GMA 900 consumes very little power, and relies so heavily on the CPU and system memory, I did not bother including that information in the post.

#10 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 09:33 AM

The fan just stopped and the reported cpu temp was 46ºC. A few minutes earlier when the fan was running it was 48ºC.

Conditions:
The ambient temp is 21ºC.
wifi on
backlight full
only a shell console and Desktop Planetarium running

#11 Neil

    EEEmazing User

  • Moderators
  • 6,897 posts
  • LocationBerkshire, England

Posted 28 November 2007 - 09:38 AM

Quote

The fan just stopped and the reported cpu temp was 46ºC. A few minutes earlier when the fan was running it was 48ºC.
So the heat generated by the fan is greater than the cooling effect?!

#12 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 09:42 AM

Started playing a video, and the fan turned back on at 54ºC

#13 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 09:47 AM

Quote

Quote

The fan just stopped and the reported cpu temp was 46ºC. A few minutes earlier when the fan was running it was 48ºC.
So the heat generated by the fan is greater than the cooling effect?!
???

The fan was running until the temp dropped to a reported 46ºC. The last reading I observed when the fan was running was 48ºC. If the fan generated more heat than it cooled then the temperature would keep increasing and it wouldn't turn off...

#14 imabuddha

    Senior Member

  • Members
  • 192 posts
  • LocationNW Georgia

Posted 28 November 2007 - 10:29 AM

I stopped the video and the eee has cooled back down to 46ºC, but the fan is still running after 10 minutes at this temp. Strange!

#15 Neil

    EEEmazing User

  • Moderators
  • 6,897 posts
  • LocationBerkshire, England

Posted 28 November 2007 - 11:02 AM

Quote

???

The fan was running until the temp dropped to a reported 46ºC. The last reading I observed when the fan was running was 48ºC. If the fan generated more heat than it cooled then the temperature would keep increasing and it wouldn't turn off...
I wasn't being overly serious... the way it was written made it read as if the fan was making the machine hot, and, when the fan stopped, the machine cooled, rather than the fan cooling the machine sufficiently to allow itself to stop :)

#16 NoeeePC4MeWellMaybe...

    Senior Member

  • Members
  • 232 posts

Posted 28 November 2007 - 11:29 AM

LOL, I was just 2 degree shy of the fan shutting off...I should turn off my electric heater to drop the room temp down a few degrees :lol: . I guess my earlier assumption in the fan noise topic that the fan likely shuts off at ≤ 45 degrees was correct :)

The eeePC was on for about 30 minutes or so before I decided to take a risk and give passive cooling a try. I've been using it in the "cardboard fanstop" mode for about 5 hours in total yesterday.

In the prior to the test… paragraph I meant to say that prior to performing the test I believed that the CPU was the only component that would likely require active cooling, but after performing the test I realized that even it does not require active cooling.

I completely understand the "fear" surrounding passively cooling a device that was sold with active cooling, but the reality is that if Asus had decided to omit the fan it appears as though we (the non-overclocking silent pc seekers) would be using our eeePC's without issue, and without any complaint about it being fanless. There are many passively cooled PCs and terminals with 3W-10W TDP VIA processors on PCBs similar in size to that of our eeePC. I wouldn't be surprised if most people on this forum would call me insane for running a desktop A64 X2 based system with a fanless powersupply, fanless 7600gs, and only two fans running at 1100 RPM to cool the entire system, but it works fine.

Also, we should not forget that the 66 degree reported maximum temperature when passively cooled is the core temperature of the CPU, and not the temperature of the ambient air inside of the case. None of the other components on the PCB have to dissipate 5W of heat when pushed to their limit, so I am sure that the likely 5-10 degree increase of the ambient air when passively cooled is nothing to worry about. Another thing to consider is that the CPU fan's ability to drastically cool any of the Intel chips on the top of PCB is significantly reduced due to the fan being located on the back of the PCB, directly underneath the CPU socket. The difference in temp between being passively cooled when relatively idle and actively cooled when relatively idle is only 9 degrees. The difference in temp between being passively cooled when the CPU load is maxed and actively cooled when the CPU load is maxed is only 7 degrees.

Don't get me wrong, I'm not trying to persuade anyone to disconnect the fan in their eeePC. I'm just letting you guys know that if you've already voided your warranty, and/or you plan to do so, and you want your eeePC to be as silent as possible, you can go ahead and disconnect the fan without worrying about frying your processor or damaging the system. Yes my testing methodology did not include any probes or thermal cameras, but in non-complex situations like this logical deductions influenced by first-hand observations and prior knowledge/related experiences can also lead to a correct assessment :)

In all likelihood, after a few months have passed, and I have forgotten about the $400 price tag of the eeePC, I will probably end up voiding the warranty by disconnect the fan. In the meantime I'll just use the cardboard fanstop method at night.

*I apologize for the long posts. I just like to get all of my thoughts regarding an issue out at once….I think I should go on hiatus from eeeusers for a few days to give my fingers a rest :cool:

#17 TacticalPenguin

    Senior Member

  • Members
  • 639 posts
  • LocationNear an EEE

Posted 28 November 2007 - 02:29 PM

My vaio has a pentium M and it sits at about 70-80 celsius depending on how much Im doing. I agree that 60 is nothing in the eee pc but it IS a possibility that that much heat could POSSIBLY do damage to something else, if not shorten the battery's life.
EEE Fund: Who cares I finally got it 4/16/08
I got an eee after creating the "post here when you get your eee" thread, which had over 1450 posts in it, making it the forum's largest, before I could post in it. Also, XP SP3 + 8GB transcend SDHC card + 405-999mhz eeectl

#18 Chris

    Senior Member

  • Moderators
  • 526 posts

Posted 28 November 2007 - 03:01 PM

Quote

I wouldn't be surprised if most people on this forum would call me insane for running a desktop A64 X2 based system with a fanless powersupply, fanless 7600gs, and only two fans running at 1100 RPM to cool the entire system, but it works fine.
From someone who's run several single-fan desktop PCs, you're perfectly sane.

Now if you were running truly 100% fanless, then you'd have to be committed just for the sheer volume of cash you'd have spent. ;)
Eee PC 701 4G - HP 6510b - Desktops

#19 Daijoubu

    Advanced Member

  • Members
  • PipPipPip
  • 56 posts

Posted 29 November 2007 - 10:17 AM

I went ahead and disconnected the fan, you can do this without voiding your warranty
Just use a flat screw driver and pop off the keyboard (3 spring on the top)

Posted Image

And then, using the flat screw driver again, push the connector of the fan toward the left to disconnect it
It's the 2nd one from the left

Posted Image

#20 citivolus

    Advanced Member

  • Members
  • PipPipPip
  • 35 posts

Posted 29 November 2007 - 01:48 PM

is it just as easy to later reconnect the fan without voiding the warranty?





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users