Solutions for the M5N loud Fan Problem

This new topic is for constructive and useful suggestions on how to fix the fan problem. Like wise if you can also post failed experiments this would be also useful, as this will save others from going down a simlar path.
Thanks in advance and lets hope we find a fix !
a


ASUS M6Ne 15.4" WSXGA 1.7 PM ATI9700 80Gb HDD 1Gb RAM

 

My suggestion is that the only solution for the M5N fan noise problem is either:-

(a) the release of a new BIOS with less fan activity by Asus

or

(b) the return to and refund by Asus of all those M5N machines on which customers are suffering the unacceptable fan noise issue.

 

You raise a good point, as far as I am aware the ASUS retailers here are supplied through the EU channle..as they are probably part of the EMEA group (Europe Middle East Africa) ... so the one I bought here and returned, even after updating to the lastest and greatest BIOS, is prob. a redirect EU version...the power plug was UK style (WE use UK sockets here).

Back in a former life saftey and EMV testing made sure that "Parts" had to be changed for the EU and also TUV/Saftey had to be done in addition to any FCC testing done, this could mean that certain parts do differ across the markets, also I belive it was mentioned in another thread that ASUS supplies different parts in thier "Barebones kits" to the ones supplied in sealed units.

 

i already postet this in the other thread. here we go again
http://www2.badenstrasse.de/m5fan/

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by Underpantman

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by markus

 

nobackup, here it's like -100 degrees, so it works here
but i agree, that this "fix" is not quite satisfying.
with speedfan you can also set the "PWM Frequency" to a lower value wich has the effect of slowing down the fan. but: it does not get quieter, but the sound gets even more anoying! (not louder, just more annoying)
another "interesting" thing is, that i wasn't able to turn the fan off with linux. i made all the neccessary step to do that, and at the end, the kernel even said that the fan is off. but it wasn't. hmmm ...


<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by Nobackup

 

markus

hast du den auch den Vibration


Question is markus, do you also have the vibration in the keys ?

I really liked the form factor (more than the Acer Ti382 I am now using) and I am thinking of going back if we can find a solution..

Although Asus has just deployed thier latest an greatest here

I went down yesterday to look at the the new W5xx but they had just sold the last one , and expect stock next week again, and had a look at the w3x ...but thought the screen was week...nice design though..

 

did anybody try analyzing asus bios file? I vaguely remember that back in my college days I used to edit some hidden bios parameters of my desktop etc.
It seems that if one is able to change the fan shutdown temperature to something more sensible (like 40C) instead of current 20C it would be a solution.

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by dmanaenk

 

i tried the getting into the bios settings, but i didnt get anywhere. ASUS has kindly made it virtually impossible. They use a variation of AMI bios, but no program ive found supports it.
Furthest i got is nearly stuffing the bios totally, so i quit on that front for now. but ure welcome to try

 

did any of you guys try writing to the author of speedfan?
Given that his program is capable of turning the fan off, it should be a piece of cake for him to write a small upplet that turns it back on say when cpu temp reaches some level, and turn off when it drops below smth.

 

Yes. I've just done it.

Sam with M5

 

Well, I wrote him too.
He says that latest beta should be capable of some more action.
I'm gonna try it later on today.

 

If the current version of Speedfan is unable to control your fan I suggest you first write to Asus, asking them which fan control chipset is used in your notebook. That way, the author of Speedfan would be able to help you (and help us too, A2D notebook owners!) without too much effort. Otherwise, that will be too difficult.

 

So...errr...I think I got Speedfan to work?!

Kinda. At least it's a start. I have the fan reporting via SpeedFan *and* Asus Probe hovering at about 900 rpm, going down to 800.

Under "Advanced" tab:

Select "ADT7463 at $2E on Intel SMBus"

Then I changed PWMOUT1 CLOCK to 14.7Hz (you can play with these to make it higher/lower--e.g. 11Hz allows the speed to drop to 600-700RPMs)

It's not "off, faster then faster" according to CPU temp, but to be honest I was just trying random things...if anyone wants to pick up on this and go with it, please be my guest.

I also changed some other settings...Speed01 on main page is 61%, etc.

The CPU temp, incidently, appears to be stable about 39C and I have it on the Slow setting in Asus Power4 Gear.

Hope this is start (i'm using the beta 4.22 version BTW).

 

hehe, i also emailed him the other day. he probaqbly getting a mass of emails in about m5n's. unfotunately that PWMOUT1 frequency modulator makes the fan unstable and actually makes the sound worse when u use low settings, so i just set mine to around 35Hz. which is about 100rpm.
settings on the main page dont seem to do much, ive tried nearly everything to the fan sped and temp to change with no luck.
maybe he'll bring out another beta soon that will help us.

 

the pwm device is a ADT7463 isnt it?

 

well, I have some progress.
I got a latest beta from him.
With it, setting pwm1 to 'manually controlled' enables control of fan speed from the main tab via changing the % in speed1.

But in a REVERSE way - that is set 100% to stop it.
This is definitely some improvement, but not sefficient so far, since another function - making fan ignore settings and run full speed after some temperature is reached does not work.

 

Well, using speedfan4.22 beta 9 (not available via web site yet)
I was able to get the following:
1) limit the fan speed to whatever I want (even zero) if the cpu temp is below trigger point that I am also able to set.
2) if the cpu temp is above trigger point - fan runs at 100% speed

I am almost completely satisfied with this solution.

 

It isn't matter if you set the frequency or not.
I suggest this link to understand how this controller works.
www.analog.com/UploadedFiles/Application_Notes/ 3619620726148651766408599520AN612_0.pdf
I recommend the p.29 the "fan speed measurement" topic and the following parts. This will explain why doesn't want to change the sound of the vent as we could expect.
Aaand it is funny to set the offset of the temperature. You can achieve, let say, 18degrees of CPU temp [ ], but nothing to happen about vent [V]. (set this offset to - 30 [ )])
And the finish of that document is very interesting concerning our main problem.
Description of the "PWM1 mode" setting is at the end of p.69.
Who has enough time and skill to make a small program to adjust that bits?



Sam with M5

 

if somebody could tell me how the hell to access the smbus i would be happy to try write a program. ive been trying to find out how for ages, but cant find anything about it

 

oohhhh, can you put the new beta somewhere!?!?! please!
just put it up on some filesharing site, or give us a link to it.

 

duzigo,

you can just register for free at the speedfan website (takes 15 secs) and click on the "beta" tab and download the latest version (have the regular released version installed and just copy over the .exe. with the new one)

http://www.almico.com/speedfan.php

dmanaenk, could you post your exact settings for speedfan to the board so that others can mimic your limited success? thx.

Sam with M5, I really wanted to understand what you were saying, but it just didn't get though! lol [ ]

 

dmanaenk has beta 9, only beta 6 is available on website. thats y i asked him for a link.

 

Anarquista:
I've just tried to reflect to those posts where sy shares his experiences with adjustable parameters.
I mean for exmaple that frequency setting of pwmout1 doesn't have influence for fan speed,
and the visible Fan01 rpm on the main page isn't a real rpm, it is only the result of a time comparison between the base frequency (90kHz) and the adjusted one. But if you took that document you will understand what I say (- I hope so)

Sam with M5

 

the pwmout1 frequency setting does change the rpm of the fan. I can acutally hear it changing, as well as asus probe reporting the change.
But unfortunately, changing it doesnt change the amount of noise the fan makes much, cos its kinda unstable when set low and makes a thumping noise instead of the constant whirring.

 

Sorry, I promised not to distribute this beta. It will show up on author's web site withing 24 hour, or at least that's what he wrote me this morning.
As for settings:
In configure/advanced tab:
1) PWM1 mode - 'Manually controlled'
2) reverse pwm01 logic ON
After this speed control on the main tab works. After this you can play with it yourself.

I also did:
On configure/Speeds tab:
1) Select only speed01
2) set its maximum to something
3) Click on automatically controlled
On configure/Fan tab
1) selected only fan01
on configure/temperature tab
1) set desired temp (this is the temp below which fan speed is zero for me)
2) set warning temp - above this temp fan goes to 100% speed automatically
And finally on 'readings' tab I selected 'automatic fan speed'

 

well. 4.22 beta 11 is available.
Instructions I posted apply.
Please post you impressions.

 

new beta is awesome.
got my fan going between 10-17% ....just a nice very soft swish
only problem now is that dialup networking apparently uses quite alot of cpu....pushing it above the limit...lol....it keeps on turning oin and off now. strange....maybe need new drivers.
normally temp idles around 50-55 degrees celcius, but when connect using dialup, it goes up to 65

 

So are you guys saying that with the new beta, the m5n fan problem practically doesn't exist? Let me clear this in my head, so we'll be able to set the temperature to about 50C for the fan to go on, and it turns off automatically when it's below 40C??

s5n or m5n!??

 

well, not exactly.
You can set to temperatures:
desired temp, below which fan is at min speed which you set
warning temp, above which fan is at 100% speed
when the temp is between desired and warning the fan speed is either at speed set manually or automatically changes withing prespecified bounds.

So there are cycles: you start with fan off, load increases, fan starts, cools down below the trigger point, fan stops, temp increases again.
Or you can cycle around warning temp going between 100% speed and the max speed you set.

What would be cool to have: to be able to set 4 trigger points instead of 2.
That is for example to have fan go to 100% at 60C and then switch back to max (say 50%) not when it cools down to 60, but say 55C.

And the same for lower bound. I told that to the author. He said he'll think about smth.

 

Hey not bad!

I followed Dmanaenk's directions and lo and behold it works (thanks BTW)!

I have the "automatic fan speed" checked on the main page and as of right now the fan is "off".

My threshold temps are 40C for "desired" and 55C for "warning". Using the computer on a battery (how I use it most of the time) and on "mid" settings for Power use my CPU temp is 34C, which means "off".

I was curious as to what people were setting their "desired" and "warning" Temps at, e.g. what works best for comfort and CPU saftey?

Also, I "donated" $10 to the author of Speedfan with a note thanking him for his work on this issue for us...maybe if enough of us throw a couple bucks at him he'll continue to develop this the M5np owners...

 

Actually, one small question for the Speedfan users out there...

I notice when I use it I hear some weird noise, almost like a morse-code noise, very faint, sounds like something electrical--this used to happen with my Averatec too, but it only happens when running speedfan. I assume it has someting to do with the various aspects of the progam accessing fan speeds, systems temps and voltages, but I was curious if this was isolated and if there is a fix for it?

 

yeah, ive heard that before. didnt seem to be coming from the fan tho. seemed to be coming from the squarter of the L panel that would have made it a square panel. i havent heard it since then tho.

 

but basically the m5n fan problem is fixed, and when pluged in battery, the fan doesn't go crazy and it's silent now correct? Then is the laptop still pretty cool after about 2~3 hours of use without plugged in (because fan is off right?).

Man, this is tight. if all goes well, i'll prolly get me a m5n this summer!

s5n or m5n!??

 

no1pg,

Yeah, baiscally the "fan problem" (I never considered it really a problem) is fixed--at the very least we are headed in the right direction of controlling it. The assumption is that Speedfan will be conintued to be developed and hopefully we can be even more specific in threshold levels (e.g. 4 instead of 2).

even forgetting the "fan problem" the M5np is a great laptop otherwise.

 

"the fan doesn't go crazy and it's silent now correct? Then is the laptop still pretty cool after about 2~3 hours of use without plugged in ?"

This all depends on the thresholds you set for it--if you have the desired temp set at 40C the fan does come on for a few seconds, cools the CPU and turns off--but I notice the longer it is on, no surprise the overall system temp gets warmer and the fan comes on more frequently--and forget it if you're playing a game, it's gonna blast. It being "still pretty cool" is very subjective--hard to say (I also barely use mine for 2-3hrs straight).

The question is: What is a good "desired" temp? I believe the A2D users have their fan come on at 50C, or at least that seems a common threshold--but the M5n is a different beast, so I'm not sure that 50C is a good temp either--maybe 45C? I keep mine at 42C right now, not sure if I'm gonna up it or down it, but I'm hoping for some feedback for those who are using it....

 

Hey Guys.........

Not to rain on the parade in here or anything, but I found something out you all may want to take note of.

The fan in the S5 is not the same fan in the M5. It's the same thickness and physical dimensions..... but it has half as many fins.... this moves less air which is theorically a bad thing, but at the same rpms, it's quiter... I changed one over to my m5 this morning and the soft humm of the M5 fan is much less with the S5 fan.

I'll let you know if anything else comes of it.

Thanks,
Justin
PROPortable
www.proportable.com
[email protected]

 

anarquista, thanks for the comments! Sorry it may seem like a n00bish question, haha.

Justin, so what your saying is that the s5n fan is less powerful than the m5n fan, and thus quieter? But would that weaker fan on the s5n decrease its like span? Is there anyway we could buy a s5n fan and manually install it on a m5n if it doesn't decrease the life spam?

s5n or m5n!??

 

Justin
Interesting.
How about loading m5 with s5 to 100% load and checking how s5 fan temp compares to native fan?
It not worse, I might try getting an s5 fan just as a backup

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by anarquista

 

First... very important....... guys.. that morse-code you hear, those are beap codes... try to mark down the long, short beaps and look them up online...... there is either a ram, cpu, hard drive or fan problem... usually... it's important to know what's going on.


Ok...... I can tell you this .. I've been running the M5 for a few days with the S5 fan. So far I can't tell the system getting any hotter, but I can tell there is a little less of a humm and less air coming out the exhaust... I don't like the fact that less air is being moved though..... For instance, if anyone comes to me and says they want us to build them an M5 with an S5 fan, I wouldn't do it because we're responsible for repairing it if anything happens.... but on my system I'm not exactly worried about that. I'll see if I can work up a comparison.

Thanks,
Justin
PROPortable
www.proportable.com
[email protected]

 

Thanks Justin for your input on this and exploring the different fan issue and you're right on about selling them without modifying, though it is an interesting aftermarket modification for those interested...

For me at least those "morse codes" only occur when I run Speedfan--as soon as I exit the program they stop--they are very rapid though, not sure if I can count them, they are also very quiet... Kinda strange.

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by PROPortable

 

OK, sorry... I was exhausted when I started reading posts last night... this isn't in the bios... weird.... I'm lost for the time being.

Thanks,
Justin
PROPortable
www.proportable.com
[email protected]

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by PROPortable

 

No, I said it didn't whine...... I'm sure someone with some time can comb through for the exact wording... but what I'm saying is there is a hum..... a light humm. If you were in a perfectly quiet room, you could hear the fan, but being well within normal levels....

On the other hand, no one has ever said anything about the S5 other than it's damn quiet..... so I thought I'd give it a whirl and try it. Unfortunately, I've been trying out a handful of new notebooks this week and last weekend and early in the week I was on the road so I've been a little too busy to try to work on plotting the exact difference between the fans.... but I will get to it.

Anyway, we ship anywhere, what parts are you looking for? Just drop me an email.

J

Thanks,
Justin
PROPortable
www.proportable.com
[email protected]

 

I am using the M5678NWP (the white X-Mas special edition) with linux and finally found the solution to save the problem with linux.

All you need is lm_sensors and a recent kernel. Currently the SMBus is not enabled on the M5N, however there is an easy workaround for that, and the patch is already pending for inclusion in the vanilla kernel-tree (see http://bugme.osdl.org/show_bug.cgi?id=4391 and http://mhensler.de/swsusp/ with a patched kernel for Fedora Core 3).

When the SMBus is detected and initialized by lm_sensors all left to do is:

echo 60000 >/sys/devices/pci0000:00/0000:00:1f.3/i2c-0/0-002e/temp1_auto_temp_off
echo 65000 >/sys/devices/pci0000:00/0000:00:1f.3/i2c-0/0-002e/temp1_auto_temp_min

This will change the fan to start cooling when the CPU reached 65°C and stops when it reaches 60°C (the defaults are 20/15, which is broken I think). Adding that lines to my /etc/rc.local was enough to have a really nice, and now total quiet notebook.

I noticed no problems so far, when working the CPU will be usually around 55-60°C and the fan is permanently off. When doing high load the fan will kick in, cool the CPU down and go off again.

With digging a bit more in /sys/devices/pci0000:00/0000:00:1f.3/i2c-0/0-002e you will be able to change the frequency of the fan to make the noice a bit "smoother".

 

Hi. The fan noise of my M5N is very high pitched and a friend of mine has the same one (just less RAM, 256Mb instead of 512Mb) and the fan on his is not as high pitched. I was just wondering if there may be a fault with mine. What does the fan sound like for other M5N users? I was told that I should get it looked at because the fan might need some extra oil or something. Any suggestions?

 

<blockquote id='quote'> quote:<hr height='1' noshade id='quote'>Originally posted by anitam82