PDA

View Full Version : Anyone uprade their fans in the G75 VW(best buy) or cooling upgrades?



gingles
03-03-2013, 07:16 AM
As the title says. Anyone upgrade their fans? if so what fans you using? Anyone do anything to help increase cooling capability?

TokoDude
03-03-2013, 10:54 AM
It's not necessary your fan can cool down your GPU to 69 degrees (overclocked)

as for MSI it will get 85 degrees (same grapich card only other cooling)

thats a huge diffrence of the cooling they use. you have a good cooling now so why change it?

You grapich card will damage at 90 degrees.

mrwolf
03-03-2013, 11:34 AM
Yea i agree, the cooling is exceptional in this laptop compared to others with the same graphics card. However since i tend to overclock very very high, all the time.. I have modified my casing to allow better airflow of cold air directly to the GPU - Has helped me quite a bit to maintain under about 75degrees on full OC 750-780MHz.

I do not reccomend doing this unless you are sure you want to and know what you are doing. Also you would need to add filters inside to prevent dust..

http://dl.dropbox.com/u/57333195/IMAG0055.jpg

gingles
03-03-2013, 06:05 PM
To the gentleman actually posting pictures of his setup.....thank you. I said I had the best buy version but i guess i should have been more specific in sayin it is no longer oem......if anyone has a legit answer please let me know......im seeing temps around 84-87 right now under full load but have some plans to overclock some more..therefore i will need to upgrade cooling and the only way i see doing this is finding some fans with a higher cfm flow.

MarshallR Edit: Removed flames to others in response to deleted posts above

nirO
03-03-2013, 06:44 PM
i get same temps like u.... we have a bad paste job on the heat sink ... many get 70-75max under full load others even less then 70 so we got 10C+ then them... now either u send it to asus if u are under warranty or do it your self of find someone who can do it for u but u lose then warranty...

i presonal keep it like this ... did not OC it either ... bcuz if i send it to asus to fix it... they will probbly fail doing it right and keep it for a month+ or even destroy something else on it as many other reported.

xperiaplayerpro
03-03-2013, 07:24 PM
hi there i found that keeping my g75vw on 4 small cup costers one each corner does the job for me as i also oc very high CC760 SC1520 MC1950
this keeps my temps 80c and below, oh and if u are getting higher temps then that i would get it repasted, b4 i as geting temps going up to 97c so i rma'd it for repaste , now as i said much better only thing wrong now is that my keyboard light suddenly not working at all. i think while it was being fixed they have dislodged my cable as they very rearly flicker on randomly for 1 second!!!

Kawazu
03-03-2013, 09:28 PM
I did the same thing with mine.

http://img42.imageshack.us/img42/4214/img20130228192720large.jpg

Did only take a pic on one of the sides but the other one looks just the same.

nirO
03-03-2013, 09:50 PM
question for u guys that drilled holes in case do u lose your warranty by drilling the holes since there was a threat where i mention to someone he can do this for better temps and some flamed not to do it cuz he loses the warranty

bignazpwns
03-03-2013, 10:10 PM
I seen the temps of the guys who drilled hols and they are the same as mine with the same clocks without holes or in some cases mines cooler. What it all boils down to is if the heatsinks were mounted right and they did not put 3lbs of thermal past on it.

My gpu load temps never go over 74c with furmark running at GTX 670MX at 810 on the core and 2510 on the memory.

Kawazu
03-03-2013, 10:14 PM
I seen the temps of the guys who drilled hols and they are the same as mine with the same clocks without holes or in some cases mines cooler. What it all boils down to is if the heatsinks were mounted right and they did not put 3lbs of thermal past on it.

My gpu load temps never go over 74c with furmark running at GTX 670MX at 810 on the core and 2510 on the memory.

The thing is this, some of us have GTX670M that is based on Fermi And you have an GTX 670MX that is based on Kepler so you card will run cooler just becouse of that.

But do like this if you want proof, remove the big plastic lid under your computer and run the same tests like that and see how much the temp drops, then you will get a idea of how much you would gain to drill.

bignazpwns
03-03-2013, 10:34 PM
The thing is this, some of us have GTX670M that is based on Fermi And you have an GTX 670MX that is based on Kepler so you card will run cooler just becouse of that.

But do like this if you want proof, remove the big plastic lid under your computer and run the same tests like that and see how much the temp drops, then you will get a idea of how much you would gain to drill.

Ok forgot to say that it was the same VW with the same GTX 670m. I was using my GTX 670mx as a example of the temps i get with this system now. I just forgot to say the same card in that system. Keep in mind the TDP of the GTX 670m and MX is both 75W so heat output will be pretty close. Its not like comparing a a 45w 660m to a 75w 670m.

Now process size 40nm vs 28nm will be the same like the whole deal with sandy vs ivy. the smaller process should run cooler but it did not due to the smaller die having less surface area to displace the heat. Same with kepler vs fermi. Kepler temps are slightly better then femi stick but overclocks femi can hold the temps better then kepler, kepler does start to get pretty warm with higher clocks and its a few c warmer then my old GTX 670m at 750. The big factor here is not the core or process its the fact they are 75w cards so you are talking +/- 5c on the high side if the heatsinks were mounted right and the thermal paste was put on right. This is why heatsinks are specs for displacing so many watts of heats. The stock heatsinks on the gpu for the VW should be close to about 90w or so i don't know the official specs but in the same VW i tested the GTX 670m in mine and the temps were pretty much the same clock for clock because of how the system is set up. the fan intake is not on the bottom of the case so anything touching it will starve it for air.

Now we can go into talk about positive vs negative air pressures for cooling computers but i did run the numbers with a GTX 670m vs people that drilled hols and the temps were the same if not better in some cases. And i was hopeing to see my temps about 5c hotter because i wanted to fire up the CNC are work and mill in some air vents but from what i seen it was not worth it and it was not worth adding some holes to suck up dust. This case may differ if your heatsink was not mounted right, base is not 100% flat, thermal paste is just gobbed on i have not tested those but i know i did replace my thermal paste with AS5 and put it on the right way not the Asus 2/4 tube method.

Kawazu
03-03-2013, 10:42 PM
Ok forgot to say that it was the same VW with the same GTX 670m. I was using my GTX 670mx as a example of the temps i get with this system now. I just forgot to say the same card in that system. Keep in mind the TDP of the GTX 670m and MX is both 75W so heat output will be pretty close. Its not like comparing a a 45w 660m to a 75w 670m.

Now process size 40nm vs 28nm will be the same like the whole deal with sandy vs ivy. the smaller process should run cooler but it did not due to the smaller die having less surface area to displace the heat. Same with kepler vs fermi. Kepler temps are slightly better then femi stick but overclocks femi can hold the temps better then kepler, kepler does start to get pretty warm with higher clocks and its a few c warmer then my old GTX 670m at 750. The big factor here is not the core or process its the fact they are 75w cards so you are talking +/- 5c on the high side if the heatsinks were mounted right and the thermal paste was put on right. This is why heatsinks are specs for displacing so many watts of heats. The stock heatsinks on the gpu for the VW should be close to about 90w or so i don't know the official specs but in the same VW i tested the GTX 670m in mine and the temps were pretty much the same clock for clock because of how the system is set up. the fan intake is not on the bottom of the case so anything touching it will starve it for air.

Now we can go into talk about positive vs negative air pressures for cooling computers but i did run the numbers with a GTX 670m vs people that drilled hols and the temps were the same if not better in some cases. And i was hopeing to see my temps about 5c hotter because i wanted to fire up the CNC are work and mill in some air vents but from what i seen it was not worth it and it was not worth adding some holes to suck up dust. This case may differ if your heatsink was not mounted right, base is not 100% flat, thermal paste is just gobbed on i have not tested those but i know i did replace my thermal paste with AS5 and put it on the right way not the Asus 2/4 tube method.

Please share the numbers with us that have not seen them.
Room temperature: ?
Clock: ?
GPU temperature: ?
What you stresstest with: ?
How long you run it: ?

nirO
03-03-2013, 11:02 PM
fact is some got a god paste job some dont.... its nothing new.. as i said some get max 70-75C under full load others go to 85.... thats all about the paste job done at the heatsink

bignazpwns
03-03-2013, 11:38 PM
fact is some got a god paste job some dont.... its nothing new.. as i said some get max 70-75C under full load others go to 85.... thats all about the paste job done at the heatsink

Yess but people will still say speed holes give you a -10c drop in temps and 40fps gain.

Fact is this. Asus put some engineering into the cooling system on G75. Its the selling point of the system not the specs not the features the big selling point is the cooling system. So much so that the fans used are the match for the fin count and density of the heatsinks in terms for flow, static pressure and blade count. Now we all know they did test covers with slots for air intakes and their is a reason why they did not use it. They even went so far as to setting up the rear air vent grills to optimize for cooling and air flow.

-little to no improvement in temps on their test systems. Remember test systems are the best of the best. The heatsinks were flat when in production heatsinks can be flat or convex or concave. This is why people lap heatsinks.

-Increase in dust build up in the system. Holes int he pc will get more dust faster then how it is set up now. And the filters block some dust they do not keep your heatsink from clogging or it building on the fan blades. Remember dus is small those holse in the filters are big to allow airflow so it does not choke off the fan. Even small amounts of dust on the fan can cause a rpm drop that will raise your temps and some dust in your heatsink can also raise your temps pretty good since dust is really nice at holding heat in.

Can holes improve cooling? Yes depending on some factors. But if your thermal paste is applied right. Your heatsink is making good contact with the core of the cpu and gpu the gain will be little or nothing. It will not help with giving you a higher overclock or keeping your G75 from overheating if it has overheating issues.

Fact is Asus put more thought into the cooling system then they did to the rest of the system and they went with no vents. The Fans are far enough away from the bottom of the case where they can keep air and its not on the bottom of the case like other laptops where it either needs a vent or gets starved.


End of the day it all comes down to if the paste job was good or not. And some cpus and gpus run hotter then others. This is why people will cherry pick batch numbers on intel cpus to get cpus from the same wafter that overclock good, run on lower volts or just run cooler. And the whole 40nm vs 22nm the 22nm is cooler tell that to my i7 3770k that runs hotter clock for clock then my i7 2700k with the same heatsink and volts. Smaller die = less surface area to displace the heat and thats why people were mad because ivy was hotter then sandy. End of the day the GTX 670m and 670mx's temps are pretty close and from what i been seeing with this GTX 670mx is that it runs hotter after 750mhz then my old GTX 670m ran.


But yes drilling holes in those panels will void the warranty. You are altering the cooling system and how it was designed to work. This is a hardware modification what they do frown on. Not saying they would not honor it if you sent it to them with speed holes but legally they do not and can just send it back with a return bill for return shipping.

Kawazu
03-03-2013, 11:54 PM
Then if you have a badly applied thermalpast the connection betwen the GPU and the copper plate the heatpipes are applied to would have a worse contact right?

If it had worse contant you would not gain anything by drilling the holes just becouse the heat cant be transported from the GPU in the first place.

So ppl that have good thermalpast applied would benefit the most from this mod just becouse the temperature is transported from the GPU and into the heatpipes.

So before you ppl bash the ide of drilling the holes, try it yourself, remove the big plastic part in the bottom and see the resoults yourself.
I lost my warrenty when i opend the computer up to replace those crappy ram sticks that came with the computer so for my part it dosent matter if i drill or not.

tkolarik
03-04-2013, 03:49 AM
I just don't see it. Cooling systems were designed to take fresh air in at one place and push it buy something hot then push it out. Putting holes in at the wrong location can cause it to get hotter. And as far as post references "best buy" , the cases are the same and as far as hardware the "best buy" version has a lower quality camera, a wifi card that doesn't do 40mhz bandwidth, only 20, and a 1TB 5400rpm hard drive.

KiwiG75
03-04-2013, 04:11 AM
By drilling hole right over the fans, you may lower the video card and cpu temperatures but you will reduce the airflow over the ram, hdds and wifi card. Running hdds at high temps can be very damaging to them.

bignazpwns
03-04-2013, 05:58 AM
Weird....It seems other people are wondering the same question I asked. To everyone else....again.....thank you for he tips. I will tear it apart tomorrow and check out the heatsinks while taking a peek at the fans to see if I can find some higher cfm ones now.

Higher CFM fans are not going to help.

The computer uses a fan control that is based off temps. Really i have not even been able to get temps high enough to get the fans to kick in at 100%. Higher CFM fans are usualy just going to high a higher rpm in these were not talking more blades or dif. blade profiles like in desktops the blades on the stock fans are pretty much as high as you are going to go.

Now i have not looked into the fan specs in a while since i had the VW but if you want the fans at full power this is your only option and honestly i dont even know if it can be done. However it will void your warranty.

Lets say for a example the fans are 5v "Don't know if they are 5 or 12 or whatever using 5 for a example" and they are lets say 1w each. You are going to need to find a poin that can supply the fan with that power and can handle the extra 1w load on it. Also they are going to suck up lets say 1a. So its a bit more complex then just finding 5 volts and soldering it. If you do manage to find a spot off the top of my head im thinking if you only run one hdd the second drives power can be tapped into. Were amusing the system will boot up without any fans running. If Asus is not totaly useless this will not be possible. Now few ways they can detect this. Power draw of the fans on the circuit . More advanced then what Asus wants to put in the g series so were justing going to say its a rpm based. So you will need ti wire that that sensor to that spot "Not really hard".

I did this on the VW i had for a guy who got it after me. It was not to hard for a advanced computer person with advanced soldering skills. Also built a circuit and turned the rog button into a a dedacated fan speed control button. default low. one press med, two is high, and three is reset to low. Basic circuit then i just cut the traces to that button and wired up the unit to it. I'll dig up some pic of it if i can find them and i have a multisim file that shows it as well as the pcb layout if you want to make the pcb. If not it can be easly done on a project board. Total cost was about $15 all the parts you need can be found at radioshack.

_
03-04-2013, 06:04 AM
I've removed the negative posts from this thread. If the OP wants to discuss cooling improvements and NB modding, that's a legitimate discussion and not for ridicule.

OP: please report posts, don't respond aggressively.


To All: This is a forum designed for CONSTRUCTIVE FEEDBACK. If you can't express that, please keep quiet. Thank you!