E6540 Owners Thread

I just tried gaming on a perfored pad (without fan, just perfored, +-2cm above the table) and the CPU and GPU temps were MUCH lower, and the fan was even not turning at 100%. So, the problem could be the gap between the table and the computer.

 

This. I noticed that in my testing (I've been running daily thermal cycles to help break in my AS5, 30 minutes of Prime 95) that simply turning the notebook on it's side (fan edge facing up) further lowers my temperatures to about 78 degrees.

 

Is the laptop level on the "perforated pad", or is it slanted forward like most laptop pads, stands, and coolers?

The orientation of the laptop might affect the flow of liquid from the radiator to the heat source.

• Turning the unit on its side with the exhaust pointing up means liquid can flow back to the heat sources simply by gravity.
• Raising the back of the laptop might also provide some gravity assistance to help the liquid get back to the heat sources.

(See the "without the cover" image in the notebookcheck.net review)

[Curiosity: The notebookcheck picture shows the heat pipes are covered in black, maybe a heat insulating material, not normal exposed copper. I wonder if that raises the temperature of the capillary layer enough to prevent the liquid from getting back to the source. If it is solely for heat insulation (not electrical insulation), I wonder if it is there just for lap comfort, or to prevent melting nearby components. (I expect that altering it will void the warranty.)]

Heat pipes are designed to work as follows

1. At the heat source (cpu, gpu), the liquid evaporates, absorbing heat.
2. The hot gas fills the heat pipe, carrying heat to the radiator..
3. At the cooler radiators, heat is transferred to the blown air, and the hot gas condenses back to liquid, releasing more heat.
4. The liquid returns via capillary action along the walls of the heat pipe, to the heat source where the cycle continues.

Maybe problems can occur in several of these steps.

1. At the heat source, there may be (1a) poor heat conduction to the liquid (repasting helps), or (1b) it may run out of liquid (see 4 below).
2. The pipe could be blocked, preventing gas from flowing to the radiator. (Unlikely since hot air blows out.)
3. At the radiator, there could be (3a) poor heat conduction from the heat pipe to the radiator vanes, or (3b) there may be insufficient air flow.
   If the fan cannot blow enough air to carry away the heat as fast as the heat source produces it, the temperature has to rise.
4. The return flow of liquid to the heat source may be insufficient. Maybe the capillary flow is too slow because (4a) it's too thin or too dense (design), or (4b) the capillary layer was malformed during manufacturing. Maybe (4c) the sides become too hot, so the liquid evaporates before it gets back to the heat source. If no liquid is at the heat source, it no longer can function as a heat pipe.

Maybe some simple comparison experiments would help distinguish whether the larger problem is (3b) insufficient air intake, or (4) insufficient return flow in the heat pipe. Record CPU temperatures and clock rates, how long before throttling.

• (=Air, =Gravity): Normal: Run the computer on a level table (baseline problem case). (Airflow is limited by the height of the laptop feet. No gravity assist for heat pipe.)
• (+Air, =Gravity): Raised-only: Run the computer level but raised off the table so there is plenty of airflow in to the fan intake (maybe put some equally thick marker pens under all four corners). (Airflow improved. No gravity assist).
• (=Air, ++Gravity): Sideways-only: Run the computer on its side with radiator facing up, but with the laptop feet tight against a flat nonporous surface (like side of stiff box or bookcase) so the airflow available to the fan intake will be very similar to the airflow when the laptop is on a table. (Airflow limited by height of laptop feet. Large gravity assist.)

If B (Raised-only) runs cooler than C (Sideways-only), and C is similar to A (Normal), then the problem is mostly insufficient air intake. Maybe attaching larger feet would help if you don't want to carry around a laptop wedge.

If C (Sideways-only) runs cooler than B (Raised-only), and B is similar to A (Normal), then the problem is mostly insufficient capillary return flow. Maybe test if forward slant helps.​

 

The paint on the heatpipes should have very little effect, the main heat dissipation occurs at the heatsink fins, the pipe is just a way to transport the heat. Yeah, there is some dissipation along its length, but it shouldn't be enough to make a difference. Just raising the laptop might give it better airflow, thus lowering temps or access to cooler air (if the exhaust is close to the intake).

It's possible that the heatpipe orientation has an effect, but if Dell was even remotely smart when designing the laptop, the heatpipes should either give the best performance when the laptop is resting flat on a surface or heatpipes that perform very similarly regardless of orientation.

Dell uses a lot of thermal paste on the stock heatsinks (I know, I ordered a replacement for my Precision), so just a better paste job will work wonders. As an example, a 45 W heatsink in a M6700 with a good paste job, performs as well as a 55 W heatsink with the stock paste job. The 55 W heatsink gives even lower temps with a good paste job.

 

So it's probably a good idea to re-paste my e6540 then?

What do you guys suggest - some Arctic Silver and removing the existing paste with kitchen roll? Or do I need a cleaning fluid for removal?

 

isopropyl alcohol (90% if possible), Q-Tips and some Arctic Cooling MX-4 is what I'd recommend.

 

I repasted last night with similar results to vbman.. ArctiClean and MX-2..

Another thing to try is using ThrottleStop to set the multiplier lower while gaming. For the most part, the Gpu is going to be the bottleneck in this situation so it's ok to hold the cpu down a bit. Not ideal, but better than 90C+ temps. If I could get a hold of another bottom cover, I'd try modding it to add holes under the fan to see if that helped. That's my first guess any way..

This is a pretty solid laptop though. First business class and best laptop I've owned by far. It wasn't designed for gaming per-se, but I don't think it's reasonable for them to have assumed 90C+ was acceptable when it came up in their prototypes. All of us are having this issue so i bet if enough of us nag them they'll do something about it..

 

Great. Ordered cleaning agent and MX-4.

 

I'm having a very bad gaming experience with my laptop. I tough it was my CPU throttling (it goes to 0.8ghz after 15 min) but thanks to ThrottleStop I saw it's my GPU card who slow down. Using ATI System Monitor it shows that after a while the GPU goes down from 99% to 25-50% load, resulting in very bad FPS.

Is there a program similar to ThrottleStop but for ATI GPU cards ? Apparently the Mobile version of Catalyst is stripped down of anything useful for that.

Edit : Also I changed my dvd player for a HDD, does it affect the warranty ? (I can always swap back the dvd player when needed)

 

That won't void the warranty. Honestly, there isn't much (besides straight up breaking the thing) that will void the warranty. Ironically, Dell does mention that installing and running VLC will void your warranty (due to its volume overrides)...

Wanted to also comment on the screen issue. I installed Ubuntu 13.04 x64 tonight and everything installed perfectly. All the hardware works! Haswell graphics, audio, wifi, ethernet, backlight, webcam, two finger scrolling, even volume up/down/mute buttons and calculator button on keyboard. Literally, this hardware is 100% supported by ubuntu! Sweet Anyways, the screen issue is still present in ubuntu so its definitely a hardware defect with the screen. I'll leave it be for now... One things that nice about owning a Latitude is 3 year next business day support and I also have Complete Care for 3 years too Hello peace of mind! haha