[Guide]: How Heatpipes Work
Ok, so I've noticed that there has been alot of discussion about heatpipes here on OCN. Ranging from what is inside of them and how they work, to which design is better and how many you should have. Because of this, I've decided to write this guide.
So, while not an "expert" on heatpiped coolers, I do have a pretty good handle on how and why they work so much better than their non- heatpiped counterparts. So I'm going to do my best to explain things in a very simple, and hopefully easy to understand fashion, that should help dispel some of the "myths" that surround these copper beauties.
First off I should start with the basics (really basic). We all know that we need a heatsink on our CPU. The reason a heatsink works, is because they increase the surface area where the CPU can dissipate its heat. So in general, the bigger the heatsink, the better it will perform, just because its got the law of physics on its side.
ok, so a "traditional" (aka non-heatpiped) heatsink fulfills that requirement just as well as any other heatsink does (Sure, they've got alot of surface area) but they run into one major problem that limits their effectiveness, especially as they get really large. The problem is simple, there is only one place where all of the heat comes from, and therefore only one point of the cooler that is really being effective. In other words, there is only really one "hot spot" from which all of the heat is radiated. Here is an example of what I mean.
![]()
Ok, so now that we can see why the traditional heatsink design has its limits, we'll get to why the heatpipe cooler beats it so badly.
A heatpipe is actually something very simple. Its a hollow copper tube, that is filled with just a tiny amount of liquid (usually water, or a alcohol/argon mix).
How they work is pretty simple too. The liquid is all pooled at the bottom, and then its heated up (in our case, by the OC'd CPU). The liquid vaporizes, and absorbes a huge amount of heat by doing that (that gets into chemistry, and I wont go there right now... just accept it). As the vapor rises it is cooled, and eventually condenses (releasing alot of energy) on the inside of the heatpipe and drips its way back to the bottom. Other times, they have a "wick" on the inside that allows the liquid to move even if the unit isn't positioned vertically. (Which is why benchmark reviews tests the coolers both vertically as well as horizontally, because some actually perform better positioned a certain way).
[EDIT: Benchmark reviews no longer performs both the vertical and horizontal positons, as of the latest round of cooler testing {Best CPU cooler: Q3 2008} Most likely because most people that will purchase one of these, will use it in their tower PC, and it not doing it cuts his work in half, since one of the testing steps is eliminated]
Here is an example
![]()
And here is a animated GIF of how this works (compliments of theCanadian, he found this for me)
![]()
Notice that the heat goes to the end as a gas, then condenses onto the walls of the heatpipe, and flows back to the bottom, to repeat the cycle again. Basically, a heatpipe is air powered phase change cooling. The fan cools off the gas and condenses it instead of a compressor like in traditional phase change cooling.
So if I lost any of you in those paragraphs, or with that picture, just remember this. A heatpipe is basically a superhighway for heat. There is hardly any resistance to the thermal transfer inside of the pipe, and so that allows the heat to be transported somewhere else to be taken care of.
![]()
As you can (hopefully) tell from this picture, you can see how heatpipes allowed the heat to be effectively transported to another location, so that there is now 6 "hotspots" instead of just one. This means that the heat can be dispersed much faster, because its not concentrated in just one spot like it was before.
The last topic that I want to mention is that of HDT coolers. The reason that HDT technology makes coolers more effective, is that there isn't any excess thermal transfer going on, through material that really doesn't need to be there. In case you didn't know, each time heat is transferred from one substance to another, some of the efficiency is lost (which is why we use the least amount of thermal paste that we can, because to much of it will hurt the thermal transfer... basically making a wall of paste). So in a HDT cooler, the heatpipes contact the CPU directly, making it that much more efficient, and allowing them to use less heatpipes (this is how a OCZ vendetta 2 or a Xigmatec HDT S1283 can compete with a TRUE).
For a nice little comparison, look at this HERE and HERE. Notice that the HDT coolers consistently outperform their traditional rivals with stock fans, but after a High CFM fan is added, then they are outperformed by few of the more tradition styles, although they still do quite well.
Well, I think this pretty much wraps up my "guide" about heatpipe coolers. If you find things that are wrong, or you you'd like me to provide more detailed info just let me know, and I'll do my best to deal with it in a timely manner.
Hope this helped clear up some confusion, or maybe just enlightened a few of you out there.
EDIT:
Recently I have been thinking about expanding this guide, or creating another to cover the other topics. However, I was thinking about this, and then I saw retrospekts guide, that covers much of what I was going to cover. So I would advise all of you to at least go check his out if you're wanting to read more.
And thanks again for everyones support, I'm glad that you all like it
![]()
.
So, while not an "expert" on heatpiped coolers, I do have a pretty good handle on how and why they work so much better than their non- heatpiped counterparts. So I'm going to do my best to explain things in a very simple, and hopefully easy to understand fashion, that should help dispel some of the "myths" that surround these copper beauties.
First off I should start with the basics (really basic). We all know that we need a heatsink on our CPU. The reason a heatsink works, is because they increase the surface area where the CPU can dissipate its heat. So in general, the bigger the heatsink, the better it will perform, just because its got the law of physics on its side.
ok, so a "traditional" (aka non-heatpiped) heatsink fulfills that requirement just as well as any other heatsink does (Sure, they've got alot of surface area) but they run into one major problem that limits their effectiveness, especially as they get really large. The problem is simple, there is only one place where all of the heat comes from, and therefore only one point of the cooler that is really being effective. In other words, there is only really one "hot spot" from which all of the heat is radiated. Here is an example of what I mean.
Ok, so now that we can see why the traditional heatsink design has its limits, we'll get to why the heatpipe cooler beats it so badly.
A heatpipe is actually something very simple. Its a hollow copper tube, that is filled with just a tiny amount of liquid (usually water, or a alcohol/argon mix).
How they work is pretty simple too. The liquid is all pooled at the bottom, and then its heated up (in our case, by the OC'd CPU). The liquid vaporizes, and absorbes a huge amount of heat by doing that (that gets into chemistry, and I wont go there right now... just accept it). As the vapor rises it is cooled, and eventually condenses (releasing alot of energy) on the inside of the heatpipe and drips its way back to the bottom. Other times, they have a "wick" on the inside that allows the liquid to move even if the unit isn't positioned vertically. (Which is why benchmark reviews tests the coolers both vertically as well as horizontally, because some actually perform better positioned a certain way).
[EDIT: Benchmark reviews no longer performs both the vertical and horizontal positons, as of the latest round of cooler testing {Best CPU cooler: Q3 2008} Most likely because most people that will purchase one of these, will use it in their tower PC, and it not doing it cuts his work in half, since one of the testing steps is eliminated]
Here is an example
And here is a animated GIF of how this works (compliments of theCanadian, he found this for me)
Notice that the heat goes to the end as a gas, then condenses onto the walls of the heatpipe, and flows back to the bottom, to repeat the cycle again. Basically, a heatpipe is air powered phase change cooling. The fan cools off the gas and condenses it instead of a compressor like in traditional phase change cooling.
So if I lost any of you in those paragraphs, or with that picture, just remember this. A heatpipe is basically a superhighway for heat. There is hardly any resistance to the thermal transfer inside of the pipe, and so that allows the heat to be transported somewhere else to be taken care of.
As you can (hopefully) tell from this picture, you can see how heatpipes allowed the heat to be effectively transported to another location, so that there is now 6 "hotspots" instead of just one. This means that the heat can be dispersed much faster, because its not concentrated in just one spot like it was before.
The last topic that I want to mention is that of HDT coolers. The reason that HDT technology makes coolers more effective, is that there isn't any excess thermal transfer going on, through material that really doesn't need to be there. In case you didn't know, each time heat is transferred from one substance to another, some of the efficiency is lost (which is why we use the least amount of thermal paste that we can, because to much of it will hurt the thermal transfer... basically making a wall of paste). So in a HDT cooler, the heatpipes contact the CPU directly, making it that much more efficient, and allowing them to use less heatpipes (this is how a OCZ vendetta 2 or a Xigmatec HDT S1283 can compete with a TRUE).
For a nice little comparison, look at this HERE and HERE. Notice that the HDT coolers consistently outperform their traditional rivals with stock fans, but after a High CFM fan is added, then they are outperformed by few of the more tradition styles, although they still do quite well.
Well, I think this pretty much wraps up my "guide" about heatpipe coolers. If you find things that are wrong, or you you'd like me to provide more detailed info just let me know, and I'll do my best to deal with it in a timely manner.
Hope this helped clear up some confusion, or maybe just enlightened a few of you out there.
EDIT:
Recently I have been thinking about expanding this guide, or creating another to cover the other topics. However, I was thinking about this, and then I saw retrospekts guide, that covers much of what I was going to cover. So I would advise all of you to at least go check his out if you're wanting to read more.
And thanks again for everyones support, I'm glad that you all like it
1 - 20 of 35 Posts
Joined
·
18,131 Posts
Quote:
Maybe some unknown knock-offs have fake heatpipes, but AFAIK, all the common stuff from places like Newegg have real ones with liquid inside
Originally Posted by IEATFISH  Wow, nice explanation. So my question is, Do all heat pipes have the liquid in them or are there some known impostors? |
Joined
·
3,371 Posts
I agree, I think that all heatpipes have something in them, but my guess is that some of them don't have the right amount (or type) of liquid in them. Its actually kinda tricky to get all of that right, because they are so sensitive to the interior pressure.
There is alot more that goes into all of this than what I explained, but I tried to keep it pretty simple.
Thanks for reading, hopefully this is helpful.
There is alot more that goes into all of this than what I explained, but I tried to keep it pretty simple.
Thanks for reading, hopefully this is helpful.
Joined
·
24,896 Posts
Joined
·
3,371 Posts
Thanks... although I'm surprised that it went there already.
Well honestly, that's why I wrote it, so that it would get referenced to. Otherwise, if this just vanishes into the old threads, then it really does no one any good.
EDIT:
I didn't realize that there was already a "FAQ" for this in there... sorry about that. Although I think that mine is different enough that it shouldn't hurt.
Well honestly, that's why I wrote it, so that it would get referenced to. Otherwise, if this just vanishes into the old threads, then it really does no one any good.
EDIT:
I didn't realize that there was already a "FAQ" for this in there... sorry about that. Although I think that mine is different enough that it shouldn't hurt.
Joined
·
2,514 Posts
Great guide!
Joined
·
5,723 Posts
This would might explain the 8C difference across my cores..... then again, that problem existed both vert and horiz
Joined
·
5,841 Posts
Quote:
All heatpipes in the available heatsinks will have wicks. If they didn't, then they would only work with gravity, so they'd only work in one orientation. Heatpipes are usually vacuum sealed, so the only material inside them is the working fluid itself. It's the heat of vaporization that makes them so effective. To vaporize the fluid requires enough heat energy to match the heat of vaporization, and for water that's 40 kJ/mol at standard conditions. I believe you'd have to adjust that for pressure, since the working pressure of a heatpipe depends on whether or not there's a gas mixture, or it's at the vapor pressure of the fluid inside.
Originally Posted by theCanadian This would might explain the 8C difference across my cores..... then again, that problem existed both vert and horiz |
Joined
·
3,371 Posts
Correct on all counts there, I just tried to keep the guide simple, so I didn't bring up any of the thermodynamics/chemistry into it, but yes, you would have to account for the pressure in those calculations, because like you said, the pressure inside is variable (depending on if its "hot" or not).
That's why the pressure inside, and amount of liquid inside is so important, because it has to be enough to transfer heat, but not enough that it will burst the tube, and also have a low enough vaporization temp to be useful. (if it vaporized at 200c, then it would be useless).
Also, the only HS that I could think of that didn't appear to have a wick, was This one by vapochill.
![]()
As you can see, it actually does have a specific orientation that is required (as the top of the unit is offset, so that it can be installed in a tower case, and still have gravity work for it), suggesting that it doesn't have a wick.
You can also see in other coolers that it is often recommended to mount it a certain way, to help the wicks (and/or the flow of the vapor) out. For instance, the noctua C12p.
see, check this out.
![]()
As you can see, they recommend mounting it with the tips of the heatpipes up, think about it... if it is mounted in with the tips down, then all of the gas will get "stuck" at one end (for the most part anyway), and it will be way less effective.
Oh, and theCanadian... that difference is probably caused by the seating of the IHS on the actual CPU die. I hear that a good lapping of the HS, (and for best results, the IHS as well) will help even up the temps quite a bit.
Thanks for the feedback guys, I'm glad that this is helping out.
That's why the pressure inside, and amount of liquid inside is so important, because it has to be enough to transfer heat, but not enough that it will burst the tube, and also have a low enough vaporization temp to be useful. (if it vaporized at 200c, then it would be useless).
Also, the only HS that I could think of that didn't appear to have a wick, was This one by vapochill.
As you can see, it actually does have a specific orientation that is required (as the top of the unit is offset, so that it can be installed in a tower case, and still have gravity work for it), suggesting that it doesn't have a wick.
You can also see in other coolers that it is often recommended to mount it a certain way, to help the wicks (and/or the flow of the vapor) out. For instance, the noctua C12p.
see, check this out.
As you can see, they recommend mounting it with the tips of the heatpipes up, think about it... if it is mounted in with the tips down, then all of the gas will get "stuck" at one end (for the most part anyway), and it will be way less effective.
Oh, and theCanadian... that difference is probably caused by the seating of the IHS on the actual CPU die. I hear that a good lapping of the HS, (and for best results, the IHS as well) will help even up the temps quite a bit.
Thanks for the feedback guys, I'm glad that this is helping out.
Concerning "fake" heatpipes, I doubt anyone would dare to use them.
Quite simply, a "heatpipe" that does not have a working fluid at the right pressure is not a heatpipe and will not work. A solid metal rod of such thickness, or worse a plain hollow tube, would have such poor heat conducting characteristics as to be unable to keep most CPU from overheating at stock.
Quite simply, a "heatpipe" that does not have a working fluid at the right pressure is not a heatpipe and will not work. A solid metal rod of such thickness, or worse a plain hollow tube, would have such poor heat conducting characteristics as to be unable to keep most CPU from overheating at stock.
Joined
·
3,371 Posts
Quote:
Not a problem at all, that's why we have guides... so people can see them, and be helped by them... right? So the more this gets seen, the better. So go ahead.
Good luck with your new responsibility, its looking great so far though.
Originally Posted by N2Gaming Great explanation, I’m going to add it to the m2n & m2n32 club seen here. If it’s a problem let me know and I’ll take it down. Rep + for U sir |
Good luck with your new responsibility, its looking great so far though.
Joined
·
10,971 Posts
Thanks, I'm still doing a lot of research, I myself will have a lot of questions about my system. I felt this thread you posted will definately help newbs understand cooling a littel more. I'm going to wadd up on links and try to remember what I have in each link so when I get those newb questions I can refer them to a guide that will be able to explain it much better then I ever could and save me some time.
![]()
1 - 20 of 35 Posts
