Clevo P150 and P170 cooling systems neglected by Clevo! (solutions here)

this is what the bottom of my machine looks like, i replaced the crap plastic grills with high airflow metal grills, i also added some fans that are actually attached to the bottom plate and they plug into either the gpu or cpu fan plug on the mobo, and then the cpu or gpu fan plug into it. if you guys are interested i could do this for you, i would buy a brand new bottom cover from sager, and install the metal grills in it, for extra i could add the additional fans, which drop the temps about 3c from my tests. the great thing about this would be that it wont void your warranty, you would just need to replace the original cover when you need to send it in.

 

All in all, what was your final temp drop after the modding?

Also, how much did it cost you? Is it easy to do?

Great job anyway!

 

You used the space that would occupy the second set of DDR3 RAM for the blower fans, right? I've already cut out the fan holes and ordered some grilles and have a fan splitter so might run some tests with blowers similar to how you placed them. But so far my CPU is less than 75C, typically less than 70C and GPU < 80C unless I OC the heck out of it. But if I just prop it up and add holes it drops temps like 10C on the GPU.

 

actually you could still use the bottom slot for ram, if you really needed to.

 

I want to order a cover! How much are they from Sager? Also, do you need to be a customer of theirs? I bought my P150EM from Malibal. I also have leftover fan grilles like that in my random parts drawer so I'm ready to mod

My CPU runs at 75C max when rendering video on high priority and all cores with the lid on... I think I got lucky with that installment but equally as unlucky with the GPU cooling, lol.

 

haha my dads old hp finally died, and so i took it apart, and it had this awesome cooling system all just for the cpu, two full sized blowers, one slightly smaller than the other, each had fins attached to the exaust, but the bigger fan ad fins that were attached directly ontop of the heatsink for the cpu as an intake, as well as a normal intake. man they must have had serious heat issues, espescially since the gpu was basically sitting under a fan and not even attached into the cooling system. maybe i can repurpose some of the cooling from this to cool the vrams on the bottom of my card.

 

I'm considering getting the U2 or u3 Cooling pad for my Sager. Is the u3 necessary? Where would the 3rd fan go? It seems like the U2 with the two fans being under the CPU and GPU intakes would suffice...Also, how much would you charge for your plate and modifications?

 

I would personally get the U2, as the U3 has slightly smaller fans and is too big anyways. Just stick with the U2 you can always buy fans and attach to it if it isn't strong enough

 

it would have to be around $65 plus shipping, also i was thinking it would have some kick up legs built in, and im actually working on a 3d printer right now, from which i could make that, but we are still getting parts for it and its going to take at least 1 and a half months. but i might be able to do a few without the legs for less just to see if people like them and to see what kind of demand there is for such a product. anyways if i get 4 or 5 people that are interested in getting something sooner i would be interested in making some for cost+shipping and seeing how you guys like them.

and perhaps when i have everything all finallized ill include enough copper tape to do the foil tape mode with.

 

Today my gpu got 100 degrees and laptop shut down ater only 1.5 hour of gaming, and not even on 100% usage, I really need some fixing here... So aduy if you can ship to europe I'm the first to buy!

 

What kind of surface was your notebook on when you were gaming?

If it was flat like its supposed to be you've got a problem with yours that just a cooling pad may not fully help you with. Over 100c is uncommonly high with these and indicates improper cooling on the inside. Just in using the thing for gaming there might be a risk of prematurely burning your GPU up even with the cooling pad.

 

This foil tape mod actually reduced the temp by 3C, the fan doesn't recycle the hot air anymore. I also have 3mm thick closed cell (up to 110C) foam on the way, I might just stick it onto the back panel as a permanent mod and put a foil tap on it to prevent direct contact with th heatsink surface, that way I can make sure the heatsink is absolutely flat.

 

It was just on a wooden table, so perfectly flat.. Should I try repasting or just go to the store where I bought it for help? Anyway this is the first time I'm not using it on a metal grill which lifts up the laptop 1cm above the table for good airflow, on which the overheating doesn't occur(but still 90-93 degrees though..) This shouldn't be a problem to just game on a flat table.. this makes no sense.. My previous laptop Msi gx660r with a 5870m had a max temp of 75-80 degrees.. I'm starting to miss that..

Btw my cpu has had a max temp of 79 when it shut off so that shouldn't be the problem

 

I just did a kombuster gpu burn in test wit the back of the laptop 2 cm above the table and I already got 98 degrees after 1min45 secs, please tell me this is not a normal behaviour? I guess I'll contact my seller..

 

Is the 98C with stock clock?

 

Yes this is on full stock settings..

 

whoa that's way too hot for a HD7970M at stock clock, have you done any repaste job before? If not, you might want to give it a go with higher w/m-K thermal pastes like Prolimatech PK-3, Gelid Extreme or Arctic silver 5. IC Diamond 7 is only good if the contact between the die and the copper plate is poor.

 

Has anyone tried lapping their heatsinks? Just wondering how much difference it might make, as I'm thinking of giving it a go.

 

will this work for w150hrm? i have opened the case and it seems similar

 

Yes
10char

 

You guys might wanna check the GPU heatsink to see if the pressure is uneven like in the tutorial... I wouldn't leave that likely possibility out especially since temps are so high.

 

i have lapped my heatsinks, and i think it works, but to be honest, i really wonder if adding extra heat sinks, is better than adding extra insulation to the heatpipes, the only reason i havent tested my theory is becuase you cant really re stick the heasinks.

 

So here are my findings after some investigation into my problem(gpu getting 100c after only 2 minutes of stress test):

1. With backplate off, after 10 mins of stresstest, gpu is only 88c..
2. While pressing gently with pencil remover on the 2 most outwards corners of the gpu heatsink, I get an instant -8c during stress test(which makes it only 80c max). Pressing the 2 corners more on the inside of the laptop has no effect(not even 1c)

And this is still with standard thermal paste, so if I do a repaste with decent compound, temps should be more than fine if I apply the fixes to my problems mentioned in this topic.(And I can maybe think of an overclock ^^)

Me happy, since I don't have the underutilisation bug, everything is perfect now

EDIT: I did the trick with the washers, my temp is now 83c max without cover, but now there is still some room for improvement. The corner where the ring tab is attached to(and the washer) is now in perfect alignment(read: same temp on pressure) with the 2 corners on the inside, but the other corner, when I press it gently, I still get +- 3c off the normal temp, how can I increase pressure on that side? using a different size screw or? since using a washer there isn't going to improve anything..

Thanks in advance

 

Today I use ic diamont 24 with foil which lower my temp by 4C in furmark,7C in kombustor and 10C in games.BF3 multi all ultra 4AA I have 73C max.Without 4AA is 71C max.

put additional springs (from used old laptop heatsink)
I will try this if I find cheap heatsink with springs

 

With the 680M, I hit 89-90C and this causes the turbo fans + card throttling. Would the tape mod alone lower temps enough to prevent throttling do you guys think? I have IC Diamond applied from the factory (possibly a poor job, CPU idles in the 50s and as stated, GPU maxes at 90C) and have not done any of the mods in this thread. If I could get down to even 84C, the GPU would never throttle and I could have sustained higher performance in my laptop. Which order do people suggest doing mods in? I was thinking of the aluminum tape and a repaste first, just to try out the simpler methods.

 

If you really want to get a good grasp of the contribution of each mod, do them one at a time.

1) Carefully doing the foil mod, with greatest care not to disturb the heatsinks and radiators will give you a good idea how important it is to maintain air pressure. If you apply pressure on the radiators and heat-pipes, you may disturb the paste layer and won't get a accurate measure of foil mod cooling contribution. Since you likely reach the highest temperatures in the current condition, the impact of foil mod may be more noticeable.

2) Manually apply slight pressure on GPU heatsink as proposed to see if there's an immediate decrease in temperature. This will confirm whether the contact pressure between heatsink and die is an issue. Maybe the washer mod is in order to remediate this. Try to see if your heaskink/heatpipe/radiator assembly is free-hanging or if it's bent in such a way that it prevents the spring-loaded heatsink fastners from applying equal pressure throughout the die. If the case cover contacts the radiator when you close the cover, then there is a chance you will ruin your paste job when your replace the cover. Ideally, the radiator/heat pipes shouldn't be free-hanging, yet they shouldn't be contacting the chassis in such a way that it affects the force distribution of the spring-loaded heatsink fastners (kind of tricky to strike a balance here). I.e. you should be able to rest the heatsink surface flat against the die with no pressure.

3) remove tape to free radiator(s). Do paste jobs. From this step, I would remove and clean-out the heatsink. Then I would try to line everything up and make sure things fit smoothly together according to step 2). Before applying paste, I would apply foil mod with everything in position. I would also unscrew the fan and seal the bottom gap as well. Now the fan and radiator are somewhat rigidly coupled, but not in a way that will apply counter force when heatsink is screwed down on the die. Test for a while and if temperatures don't improve significantly, I would then consider a pressure test as in step 2), and if there's a reaction to pressure, then consider adding washer to increase pressure.

4) Increase air intake grill apertures. (If you already intend drill holes in your cover, then I would do it between step 1) and 2).

If you're up for it, one thing that remains to be confirmed is how much thermal coupling between VRAM and GPU radiators and fan metal cover play a role. To test this, I would breakdown step 1, and test one at a time as follows:

1a) tape gap between VRAM and GPU radiators only. Even if gap is almost non-existant, copper tape (or Al foil tape) will allow the GPU radiator to sink some heat directly into the VRAM radiator.

1b) remove tape from step 1a) and apply tape only over gap between VRAM radiator and fan metal cover. This should sink some heat from VRAM radiator throughout the metal fan cover.

1c) re-apply 1a) such that all gaps are sealed and thermally coupled.

1d) plug the big gaps due to fan output and radiator fins being misaligned horizontally. This is most likely where the bulk of the air pressure is lost.

1e) Plugging any gaps that may allow hot exhaust air to recirculate inside the chassis is probably also a good idea also (e.g. open channel along side of GPU fan.

If there is a big difference between 1a), 1b) and 1c), then thermal coupling is key (thick copper tape may be beneficial).

 

Has anyone tried covering the gap at the buttom of the heatsink? Does it also make a drop in temp for like 2-3C? I covered only the upper part and it gave me a drop of 3C.

 

Read below for a sort of combined reply... the problem with these systems is that if the pipes are misaligned it affects/causes all the other problems:

Very much this! Getting the pipes shaped to evenly rest the heatsink is very important and can affect all the other mods completely. I barely needed any tape after bending my pipes to their correct position since it aligned the heatsink with the GPU and also aligned the fins with the fan. If you do try it, go very slowly since they are soft and it can be hard to detect that they are even bending. I made sure I had a fulcrum in the exact place it needed to be so the bend would be just right (I.E. I shoved a properly thick screwdriver underneath the point which needed to be bent for the correct bend angle)

 

Just a brief follow up...

75C max running BF3 for hours on end with max action 64 players etc. The laptop cover is on, I'm using a cooling pad, and the GPU fan never maxes out.

I'm satisfied considering without the lid it'd easily be under 70c running BF3

 

I'm thinking of making my own bottom cover made of this stuff:

 

if you can make it look good HT id pay you to make me one lol

 

Has anyone tried rubber washers to correct the pressure applied on the heatsink? I'm considering getting these.

 

You are better with fibre, ceramic or metal... rubber tends degrade with heat over time.

 

Saw this cool thread and decided to register.

Have read so many posts... People say nonsenses.


First, yeah games won't just go to as much temperature as furmark. BUT I think on the FUTURE, and I wanna see what will happen on some crazy games that will make full use of the hardware later on...

Second, I wanna see ur temps, when the computer get's full of dust and could easily reach 120Cº+ (If no heat protection is built in).

Third, lot's of people seem to not know that having a PC at 80Cº for everyday use, it means CPU\GPU life decrease a lot (and eventually leave to a black image on screen, and then needs either new GPU\CPU or reballing if it works that time, and pay a lot for that).

Fourth, I bought Clevo W110ER (still waiting for it), but I also heard it reaches to around 90Cº on furmark. After reading some reviews it seems the CPU never reaches the full turbo frequency, and they all claim HEAT might be the problem making the CPU to only throttle very little for what it is rated for that CPU...

Fifth, Later on, I would like to make little OC, and with this much temperatures, it's better to not even think about it.

Sixth, the more the heat = the more the performance decrease.



And now a thing that happened to me and that will be 100% worse on a laptop that normally runs at more than 80Cº:

I have an old ASUS. It started to shutdown from time to time. After some time, it started to shutdown in 10 minutes after cold boot and idle in the windows XP.

Formatted everything... same problem. Opened the laptop completely cleaned, and I realize it had that "hard dust" which killed completely the hair from the fan, the laptop was reaching around 100Cº and then it shutdown by overheating protection...

Cleaned everything, repasted with Artic MX-4, improved cooling with aluminium paper and injection of thermal paste on the cooler parts that the tube was not touching the metal that goes above CPU. I also even cut the exit plastic from the case where the air get's out of the laptop. (I'm thinking to do the same to my upcoming CLEVO).


Finally worked, and for more than a year already, it never shutdown again. Now it stays at like 56Cº idle and around 78Cº on furmark after 30 minutes and with 100mhz OC on GPU memmory and 50mhz on GPU core (Crap nvidia 7300go). It's a laptop from like 2006 and it's on 24/7 and with OC when playing old games (get around +10fps like on america's army online), and it's still alive in 2012.

Now imagine the same dust on a laptop that idles at 80Cº like on those hot clevo's, they would easily get to even 150Cº or more? (If no overheating protection shutdown is there, it would even explode). Also a laptop running at 80Cº for everyday 24/7, will probably never last like my crap ASUS which is lot more cooler.

ps: The more cool as possible = The more GPU+CPU will last = linear performance, with no power decrease and no half turbo frequency applied or anything like those limitations that HEAT creates.
Keep Cool[/SIZE]

 

Hi guilty6669 and welcome to NBR!

Funny thing, noticed you've said that you've bought the W110ER make sure to check out Prema's BIOS, it'll help you out

 

Sounds like you just need to do a better job of cleaning and maintaining your laptop to me. These components are rated to around 105C. 80C is nowhere near the danger zone

 

Idles at 80C? Don't know where that comes from. Normal idle temps are low to mid 40C.

 

By the way, I was considering grills with wider aperture for my p170em back cover but realized that it's the p150em that has restricted vents, not the p170em where the grills almost exactly measure as much as the fan fins. I don't think making the back cover grills too big or make the entire thing out of mesh is a good idea, because although it would give the fans more air to breathe, beyond a certain point you lose pressure and can't control for input/output of air.

 

Firstly, laptop's thermal protection kicks in at 130C. Secondly, idling at 80C means the users doesn't even know how to repaste/clean vent. Thirdly, for whatever reason, 80C is within the safe range. Last but not least, laptop doesn't explode just because of a overheated GPU. Only when the battery is overriden.

You don't need to worry about it if your laptop runs at 80C full load, but you surely will need to worry about it when your laptop idles at 80C.

 

That is very true:thumbsup:

I doubt anyone would optimize a game for an older system (we are talking about the future here, there will be new GPUs that the game developers will optimize for).

Software like furmark that is designed to heat up and stress a GPU is always going to be better than an optimized game.

You shouldn't let it get full of dust, how hard is it to blow out the dust every couple of weeks?

Only a faulty GPU (like the old nvidia 9800m/8800m/8600m/8400m and similar) would have it's life decrease a lot when running at 80C.

Any decent GPU can run at 80C 24hrs a day and last you until you need to get a new system because your current one is too weak.

 

Well Already had burned, ATI: 9200, 9600XT X1800GTO, Nvidia 7600GT and one more that I can't remember. And even older xbox360 and
older PS3 always going to 80 and 90Cº, after only few years, thermal compound is just totally burned, and then the solder start to lose it's resistance and eventually BREAK, leading to ROD, and Ylod.

GPU's just die on me, and having high temperature is always a problem and will always reduce the life span of any component. If you say it doesn't affect, you are wrong, since in everything HEAT is always a big life span reducer...

Have seen already like 100 laptops with dead GPU on it, my friend as two ACER, and both have burned GPU and needs reballing, none took OC, they just had high temperatures of 80C and so on, they never lasted more than 2 to 3 years...

Well, I buy a laptop for not buying one more for like 5 years, and no matter what I will try to run games as much high as I can even if it lags a little (I prefer having graphics higher than reducing to make it 60fps)...

The dust, I'm talking is the "HARD STATIC DUST" It's not easy to clean, And in my ASUS I had to full disassemble the whole Laptop until I could remove the cooler out and the dust was between the fan and the cooler INTERNAL, outsite everything looked cleaned, but no air was coming out (spraying trough the exterior, will make the whole laptop full of heavy static dust everywhere).

About the benchmakrs... Well I have ASUS GTX570 DCII which is a massive two huge fans 3 slots cooler, and sadly that GPU that runs like 30% colder or something, was reaching 80Cº on a crap game (Dead Island)... Don't wanna imagine in the upcoming games how much will this huge piece of crap card reach. (I know GTX600 series improved a lot and don't heat much, but still 80Cº is 80Cº no matter what).


And for last, in a huge forum of my country "Portugal", there's really LOT'S of people with their clevos near 80Cº on idle and reaching more than 100Cº on benchmarks (May be faulty temperature sensors, or bad repast), but they bought on very different well known high trust stores, and still there's so many people reporting same problem.


ps: I don't want to have a bloody OVEN, I want a Laptop wich could have full benchmarks at 80Cº and not 90Cº or 100Cº after 1 hour... It's so sad this high temperatures. (why don't they use direct pipes above CPU full of steam inside) .
Keep Cool

 

Thermal compound needs to be changed once you see it starts to degrade, this is also the advantage of Clevo, changing compound doesn't void warranty. But for Asus, they have a warranty sticker on all CPU heatsink's first screw. I have owned quite a few Asus before, lots of problems when it comes to thermal paste replace. Manufacturers that don't stick their warranty sticker on the heatsink is really the way to go.

Edit: By the way, those GPUs that you listed, their idle clock zone isn't very low. Unlike the GPUs after Mobility Radeon HD4000 and NVIDIA 400M series, they all have an ultra low clock profile when idling. You don't have to worry about they idling at 80C anymore. At recent generation 80C is now the load temp, not idle temp.

 

There's only one big problem... MX-4 which I use claim 8 years durability...

Clevo's are coming BRAND NEW with 100Cº prolems and from Highly Trusted stores. There's even a guy which reaches 108Cº on Furmark (Probably the sensors are reporting wrongly). Because I don't see how this specialized stores are making such a bad work on the thermal paste...
Keep Cool

 

With Clevo, repaste is still an option, and don't ever trust the durability that the manufacturer claims, once you see it degrades, time to change.

 

I know that, I will probably change thermal compound each two years.

Specially because going at 90Cº everyday, will make the thermal compound to burn and will not make it liquid for much time, not even MX-4 survive much in those high temperatures conditions (8 years, my ).

And then here is full of dust, and the cooler easily starts to catch "Heavy static dust" which kinda glues everywhere and it's hard to remove, it's not just normal dust. So I clean it and re-paste...

130Cº? At that temperature, the solder is completely WATER, and every xbox\ps3 on earth would have already given ROD and Ylod...

Very nice built in heat protection of 130Cº on components made for 105 Maximum Cº. And then fans stop, and with so much POOR metal to transfer heat, the temperatures will even rise more 10Cº or more, just like what happens to the 360 GPU when u shut it down and fans stop, temperatures rise...

I know laptops won't explode, it was just theoretically speaking (kinda a joke ) if there was no built in heat protection and there was so much dust as my ASUS that no air could get out of the laptop, it would even reach 200Cº (with such inappropriate coolers and ventilation), and GPU's and CPU's do explode if you remove the heatsink, they can go above 200Cº .

Thank You so much!

Is that bios a 3'rd party Bios for the W110ER? Does it finally fix the CPU throttling?

ps: If that's the case, I was waiting for it. Will research about it. Thanks.
Keep Cool

 

Well it's 130C on the die, not the board or anything near the solder, it's also why there's no solder parts on the die, the surrounding temperature won't reach that because the peak temp is always on the die itself. Once the protection kicks in and laptop shuts down, the heat will be transferred through the copper plate into the heatpipe and then to the fins and be dissipated quickly. Even though there's no fresh air among the fins after the fan has stopped, the heat reduction from 130C to 70C can be reached in a very short time.

 

130C is the thermal trip point, really?

Which system is that?

 

Was my old ASUS A8

 

Wow AMD can really take some heat. That's impressive.

 

Yeah... Impressive until 2 years after they just keep dying on me... with no OC.

Of course it's on the die... But I always heard around 100Cº the solder can easily break leading to very different kind of failures, and most likely the gpu just looks dead (black screen). However most of the times in fact they don't need a new GPU, but to reball (de-solder, clean and solder GPU again).

Then, they don't use the most strong solder type, as I have heard that ones are not environmental friendly. But a reballing service from my country, claim they use an even stronger solder that won't also oxide which have lead mixed too (though, not Eco-friendly).

And about the temps going down as soon as shutdown... That's not what I have seen in so many things. Xbox360 for example has lot more metal though it's not copper (but still is a big-one piece above GPU die), and with a laser thermometer, we can see temperatures rising after fans stop, and it climbs like 8Cº then it starts to go down.

ps: And looking at something like W110ER cooler, it's just a little base with one single pipe, which is what have to transfer everything for the main cooler, that's the big FAIL, cooler may be enough, but the heat transfer is just a JOKE! . That's why I don't know why most of people come here saying they are good as they are, and don't need cooling mods... Well I think I will even cut the case if needed.

EDIT: Does anyone know if Clevo W110ER have extra fan pins, to put some more fans in it? I can solder, but surely don't wanna screw warranty in my upcoming brand new clevo...
Keep Cool