No announcement yet.

Heatpipes: The Investigation Begins

  • Filter
  • Time
  • Show
Clear All
new posts

  • Heatpipes: The Investigation Begins

    Phoronix: Heatpipes: The Investigation Begins

    Since breaking open bottles of beer with heatpipes and other hardware last month at the Intel Developer Forum in San Francisco, we have been cutting open a number of different heatpipes. In this article we have some new details to shed on heatpipes from a numbers of manufactures, including Thermalright, Thermaltake, OCZ, and Abit. These cooling mechanisms are supposed to keep our beloved PCs from overheating, but how does their manufacturing quality differ? With this article, we have plenty of pictures and videos showing you the differing qualities in heatpipes.

  • #2
    That was cool, I never knew much about heatpipes, learned a bit from this. You should be careful with that saw though haha, I thought you were going to cut your finger off. Now blood is one liquid we don't want to find


    • #3

      Now I admit that I have not read the patents on the heatpipes from the various companies, but fluids include gases as well as plasmas. I think the definition of fluid was something like a substance that moves or flows under shear stress. Perhaps those heatpipes had some kind of gas at some pressure close to 1 atm. In any case, I suppose all is well if the heatpipes perform well. Perhaps a good experiment would be to test a heatsink with one or two severed heatpipes? If performance does not change, it's safe to say that the heatpipes are a scam. In any case, it's good that Phoronix is looking into how the heatpipes are made and whether or not they add value to the various cooling solutions available on the market.


      • #4
        Fine bit of journalism there. It's something that's always kind of bothered me about the hardware industry. They send out these engineering samples to the press who then runs them through the benchmarks and give glowing reviews, but does the average Joe get the same piece of hardware that the press got to review? How would the consumer possibly know if the heatpipe manufacturer took shortcuts to cut costs. The motherboard still overclocks well, just not as good as the press sample. That's only to be expected right? YMMV and all that crap they disclaim at the beginning of their overclocking results. I guess there's not alot you can do other than pick companies that consistently deliver. I have alot of OCZ parts in my systems for that reason so I was glad to see that there was some substance to what they claimed for their heatpipes, though the fact that one of them appeared to have failed isn't promising, but I can give them more of a benefit of doubt than the others discussed in the article.


        • #5
          You might have to look harder...


          A typical heat pipe contains a mixture of alcohol and water at less than atmospheric pressure. The pressure is to adjust the boiling point of the liquid and give it somewhere to go when it does boil. Maybe some people have figured out a mixture that can be put in at atmospheric pressure, but it is still important to exclude air from the inside of a heat pipe. Air doesn't boil and condense, so it shouldn't be there any more than it should be in your brake lines or air conditioner's freon piping.

          The walls are lined with a wicking material that holds the liquid and carries it back to the hot part, where it boils and enters the lumen of the pipe, travels to the cold end, condenses, and repeats the cycle.

          But the important thing is that, in a cold state, ALL of the liquid is held in the wick material. Any excess would be pointless. And it's held quite firmly; remember it's the capillary action that sucks the liquid the whole length of the heat pipe back to the hot end.

          If you want to test the build quality of heat pipes, find someone who can introduce a micro pressure sensor into a sealed pipe and check for a consistent temperature/pressure curve. But that's a little more complicated than hacking one open and shaking wet copper braid to see if any drops fall off.


          • #6
            Might Want to Heat the Heatpipe

            Heatpipes can also contain paraffin wax due to its low melting point and high thermal conductivity. Paraffin wax is solid at room temperature so it wouldn't simply flow out of a broken heatpipe unless the heatpipe was hot enough to melt the wax. This could lead to the false assumption that a heatpipe is empty when it really isn't.


            • #7
              Interesting stuff

              I think the braids are most telling, if it was bogus they would not put braids in the heatpipes because doing that just adds a lot to the manufacturing process.
              Of course it is possible that due to some mistake while soldering they accidentally lose liquid..
              But I think this needs a more scientific approach to find the truth, as previous posters said, there are many possible explanations and the liquid used is of very small quantity and absorbed into the wick to boot.

              It's nice that somebody actually tries though, about time too.


              • #8
                I'm sure those heatpipes are just fine. But if you have doubts, why not drill a few holes in one and see whether it makes any difference in actual use.


                • #9
                  no fluid visible is expected

                  Where to start? The heat pipes shown in the investigation use a sintered copper powder for the wick structure. This can be seen in the cross section of the pictures. A properly designed heat pipe that uses this type of wick structure will not have any excess water inside and there for no water will run out when opened. All of the water is held in the wick by the capillary action of the fluid. Very much like a wet paper towel.

                  There is evidence of fluid if you look closely. 1) the slight noise you heard when you cut them open is air rushing back into the heat pipe. The heat pipe is a vacuum device so when the seal is broke it will suck in air. 2) Now that the water is exposed to atmospheric pressure it will not vaporize until it reaches 100 degrees C. 3) If you carefully heat the opened heat pipe to over 100 degrees C you will see the water in the wick vaporize. 4) there is a very simple test for checking if a heat pipe is functioning. Stick one end in boiling water about 2 inches deep and hold the other end. See how long you can hold it? Should be a very short time, about 5 to 10 seconds. Try this with a heat pipe where you have cut the end off.

                  If indeed these parts functioned as cooling devices before they were cut open it is almost certain the heat pipes were functioning properly.

                  FYI, I have been in the heat pipe business since the 70's and have learned one or two things about them. Any questions, I would be happy to answer them..



                  • #10

                    I must say I DO have the impression it took the industry some time to tweak heatpipes, even with a liquid and a wick you still must find the optimum pressure and kind of liquid, and when heatpipes were new I read some reviews of products where heatpipes were not very functional, whereas now they seem to have gotten them to work quite effectively, most of the time.
                    And of course it's not unthinkable there are some chinese rip-off fake heatpipe products around.

                    And don't forget the general heatsink issues reviewers and consumers sometime find, I myself bought sticks of RAM from a reputable manufacturer where the heatsink didn't touch the pads that are suppose to transfer the heat to them, and I had the same once with a graphicscard, as did several reviewers of such products.
                    Oh and we've all seen the shots of enthusiast opening consoles and such and finding that they had strange issues, like way way WAY too liberal use of heattransfer paste.
                    Never trust these things if you can check them.
                    Last edited by Wwhat; 29 October 2007, 10:49 AM.