superconducting heat pipes?

dan · Post by **dan** » Sat Nov 06, 2004 5:37 pm

hello,
i noticed that a couple of products, most notably the aerocool dp-103

claims to have a more advance version of a heat pipe called superconducting heat pipes.

by superconducing they are not referring to zero electrical resistence, but by much faster transfer of heat, facilated by lower interal air pressure.

is this true, does it improve performance, and should we watch out for heatsinks that have superconducing heat pipes?

thanks
dan

StealthGirl · Post by **StealthGirl** » Sat Nov 06, 2004 5:50 pm

Dan,

No, there's no such thing as "superconducting" heatpipes. This is just good marketing hype for using better quality parts. What they are referring to is using powdered metal (also known as "sintered") in the core instead of a spiral grove cut inside. It costs about 3 times as much to manufacture this type of heatpipe. The big advantage is the ability to use these types of heatpipes at any angle or configuration. The condensed working fluid (usually good 'ol H20) will flow back against gravity, literally "uphill".

Supposedly I have 2 dozen samples of these heading my way soon. We'll see if the sales guy actually comes through.

Kate

dan · Post by **dan** » Mon Nov 08, 2004 8:27 pm

thanks,

i guess i don't need superconducting heatpipes.

heatpipes are superior to normal heatsinks, right?

which is a better heatpipe design,
the aerocool dp-101 tower approach, or the
thermalright si-97 "startrek enterprise" approach?

i would think the tower approach would be better, as you can use the case fans to vent heat.

silvervarg · Post by **silvervarg** » Wed Nov 10, 2004 1:40 am

heatpipes are superior to normal heatsinks, right?

Well, that depends on exactly what you mean.
Heatpipes are great to transfer heat a short distance, so they do aproximately the same as the thick baseplate of a heatsink, except that heatpipes can transport heat a bit further.
To get ridd of the heat you typically transfer the heat to air, and this is the job for a large surface area. This is typically the job of the fins of the heatsink, and usually fins are used for the very same job in heatpipe heatsinks.

which is a better heatpipe design,
the aerocool dp-101 tower approach, or the
thermalright si-97 "startrek enterprise" approach?

i would think the tower approach would be better, as you can use the case fans to vent heat.

This depends on how you computer case looks like. If you have a hole on the side of the box you can duct air out from a Si-97 quite easilly. If you get a tower heatsink you can easilly duct air out through the back of your computer or dump the air straight into the PSU (is you use a PSU with a fan on the side).
The major benefit with tower heatpipe design is that it will allow heat to be transfered further away from the source by the heatpipes.
The disadvantage of a tower design is that the center point of gravity of the heatpipe gets further away from the CPU, so if computer is placed in a tower chassi you will get more leverage on the CPU and motherboard socket.

Personally I prefere the tower design, and there is no way I could have removed all my fans with an si-97.

dan · Post by **dan** » Wed Nov 10, 2004 11:25 pm

hi,
i would think that having a fluid, and one that changes phase, would transfer heat at a much faster RATE, per unit time, than immobile solid copper.

one question i have is how the thermalright product compares with the aerocool dp-101. it's not obvious to me which would perform better.

also, given its star trek enterprise design, i wonder if the heat pipes would jut enough past the socket to interfere with my PSU.

here's a pic of my motherboard
http://www.soyousa.com/images/products/800x600/tisu.jpg

thanks

silvervarg · Post by **silvervarg** » Thu Nov 11, 2004 12:23 am

Thats a PIII socket (370), right?
It is missing one lug on each side of the CPU socket compared to socket A, and the capacitors on both sides of the socket is awfully close, and you don't have any holes around the socket to mount any through-hole CPU cooler.
What CPU do you plan to run on that board?

i would think that having a fluid, and one that changes phase, would transfer heat at a much faster RATE, per unit time, than immobile solid copper.

Yes, this is correct. Heatpipes are better, but more comlex and more costly.
But there are so many more factors that are important for a heatsink. E.g. including a heatpipe usually means that you have to bridge between materials a few more times, and if you have poor heat connection in the bridges you can loose some of the gain that the heatpipe gave you.
To make a fair comparison you need to compare a heatsink with another heatsink with the same airflow (preferably the same fan). Unfortunately not that many reviews are made this way, but SPCR reviews are among the few that does really good comparisons.

one question i have is how the thermalright product compares with the aerocool dp-101.

This might depend on the airflow, so you need to decide on a few things before deciding on a heatsink and fan setup.
1. How many watts of heat you need to be able to dissipate?
2. How high CPU temp you believe is the max you are willing to tolerate?
3. What is the maximum intake air temperature to the CPU you will get?

Once you answered these you can calculate the C/W value you need.
Armed with this C/W value you can check the reviews on heatsinks that has this or a lower value at a very low airflow.

dan · Post by **dan** » Fri Nov 12, 2004 2:39 am

hi silvervag,

i'm sure we've chatted/spoken before. i remember the name.
well, to keep this post as short as possible

correct, it is a socket 370 pentium 3/via c3

1- i have a celeron tualatin 1100 overclocked to 1470mhz at 1.5 volts for an estimated heat output of 35-40 watts. i'd like to keep it as cool as possible, both for stability and long-term use. is it is as a via c3 solution, but not.

2- i'm willing and able to pay $40 for a good heatpipe solution, so as long as i can run a fan very quietly (inaudible outside the case). si-97 attracts my interest mostly.

3- you can say i want a "via-level silence" with a higher performing overclocked celeron tualatin. i'm willing to go heatpipe to do that. if si-97 can cool a 80 watt athlon xp overclocked well, 35 watts should be cake.
$40 is a small price to pay.

3- yes the capacitors might be a problem.

4- my case is very, very low airflow. only a very slowly rpm fan for PSU, and a high-pitched high rpm for cpu heatsink fan. that's it.

5- basically, in a nutshell, i want the best socket A/socket 370 cooler short of a water cooling kit, so i can cool silently. i do not know what this is, but i've heard good things of the si-97.

if you can cool a 12 watt via c3 with a heatsink, no fan, i hope you can cool a 35 watt 1470 watt celeron tualatin with a heatpipe like si-97, with a very quiet fan.

Post by **Ralf Hutter** » Fri Nov 12, 2004 7:05 am

dan wrote:
if you can cool a 12 watt via c3 with a heatsink, no fan, i hope you can cool a 35 watt 1470 watt celeron tualatin with a heatpipe like si-97, with a very quiet fan.

You can also cool an OCed Tualatin with a "plain old" Thermalright SLK-800 and an L1A Panaflo running at 6-7V. That's a very, very quiet setup and keeps the CPU temp at around the low-to-mid 40°C range unde full load. Most any of the decent high performance Socket A/462/370 heatsinks should do just as well.

dan · Post by **dan** » Fri Nov 12, 2004 12:14 pm

well i do have an all-copper heatsink, which i glued a second aluminum heatsink but i need the fan at a high rpm to keep it "stable."

i do worry the fins of the heatsink might touch the rather hot PSU, as the PSU is at the edge of the motherboard, and the socket is very close to the edge. - that and the capacities, and i'm concerned about "buyer's regret"