Heat Pipe CAUTIONS.....

[LesKelvin]

Like water boiling in a pan...the heat goes UP. The same is true with heat pipe thermodynamics. Most heat pipe CPU coolers are designed with case constraints a priority. Unfortunately.... most CPU Cooling Devices are bench tested to optimize the devices performance but in the real world ..... tower and mini-tower configurations rotate the CPU 90 degrees making the Cooling Device much less effective. Some devices even form the Heat Pipes to create a "P" trap further reducing the effective flow of heat.

To be efficient, the heat pipes must move heat from the cpu "UP" and out of the case. I have yet to see such a design except in one prototype. Has anyone seen a commercial device designed for vertical heat transfer in a tower case?

 

[BassBomb]

heatpipe wicks expand at the base causing the fluid to flow whichever direction it can (i think this is how it works)

 

[Rubycon]

This is compensated for by adding heatpipe capacity - the number of heatpipes. Your observation is correct - turning the board simply reduces the capacity of the heatsink. If this reduction in capacity results in lower net capacity than demand, temperatures will soar to throttle or shut off.

 

[lopri]

I thought I have seen many? Not tower-type heatsinks, but those blow-down ones. You can usually mount them in such a way heatpipes go 'UP'. Actually I have one in my system (Scythe Samurai). Oh and the legendary XP-90.

 

[LOUISSSSS]

i've had heatpipe-based cooling solutions on MBs that were mounted upside down in my LLV1000 case. they were no different than any other people in "right side up" cases. my temps were well within spec.

 

[scruffypup]

Originally posted by: LesKelvin
Like water boiling in a pan...the heat goes UP. The same is true with heat pipe thermodynamics. Most heat pipe CPU coolers are designed with case constraints a priority. Unfortunately.... most CPU Cooling Devices are bench tested to optimize the devices performance but in the real world ..... tower and mini-tower configurations rotate the CPU 90 degrees making the Cooling Device much less effective. Some devices even form the Heat Pipes to create a "P" trap further reducing the effective flow of heat.

To be efficient, the heat pipes must move heat from the cpu "UP" and out of the case. I have yet to see such a design except in one prototype. Has anyone seen a commercial device designed for vertical heat transfer in a tower case?

If a wick is used along the "wall area" of a heatpipe,.. the positional direction of the heatpipe is not limited,... though vertical would assist with the condensation of the fluid returning and provide maximum efficiency,...

Most benchmarks I have seen of cpu coolers,.. are actually mounted in "real world" orientations,... so you actually are seeing "real world" results,..

I have also seen a review a while back (http://benchmarkreviews.com/in...&limit=1&limitstart=5) that shows the difference on one cpu cooler with mounting direction,... and it shows negligible "real world" differences,.. as you would see with most of the cpu coolers,... but if you are a fanatic about every last bit of possible cooling,.. you can do certain things,.. most cases can be laid on their sides and function just as well,... although many might actually be adversely effected by the same negligible amount or more so than the gain of the heat pipe orientation!!!

PS-I have also seen somewhat skeptical results as far as how direct touch heat pipe cpu coolers should be orientated in regards to quad chips since the pipes when they contact the chip may have more coverage on a quad chip if crossing one way vs another,.... it has the same negligible difference delta type as mentioned above,... but again,... if you are a fanatic,... then again, the case and overall internal cooling might be effected as above,... remember it is a "system" that you need to cool,.... as one effects another in the system,...

EDIT - A good general description of the difference a wick assisted heat pipe makes,...
http://www.answers.com/topic/heat-pipe

 

[Rubycon]

Wicking makes them more versatile for different mounting positions but there is an optimal position and outside of that range you lose capacity. If your temperatures do not change then you are still within that operational capacity.

 

[Zap]

Welcome to Anandtech Forums, LesKelvin.

I recall when motherboards first started using heatpipes to cool the chipsets that people started having problems when using "upside down" cases (ones that mount the motherboard with CPU towards the bottom of the case). That was caused by the heatpipes not having a wick. With a wick (well, I suppose with a properly designed and properly manufactured wick, with a proper internal pressure) heatpipes have near universal orientation. Any differences should be negligible.

 

[aigomorla]

Originally posted by: LesKelvin

To be efficient, the heat pipes must move heat from the cpu "UP" and out of the case. I have yet to see such a design except in one prototype. Has anyone seen a commercial device designed for vertical heat transfer in a tower case?

welcome to our forum.

this doesnt apply anymore because they are based on a wick design.

So orientation isnt important anymore.

Meh, everyone already told you about the wick.

 

[Rubycon]

Originally posted by: aigomorla

Originally posted by: LesKelvin

To be efficient, the heat pipes must move heat from the cpu "UP" and out of the case. I have yet to see such a design except in one prototype. Has anyone seen a commercial device designed for vertical heat transfer in a tower case?

welcome to our forum.

this doesnt apply anymore because they are based on a wick design.

So orientation isnt important anymore.

Meh, everyone already told you about the wick.

No it most definitely DOES apply. The wick allows use in any direction but there is a cooling capacity penalty. An unwicked pipe will work only in a designed position and deviating from such a position will result in very limited capacity or drop down to limits of physical conductance. (basically the pipe will conduct heat like it was empty)

This is why a heat pipe based heatsink will most definitely run hotter IF it's near its capacity when oriented in a non optimal direction. And for this purpose we have larger pipes and designs with more pipes to accommodate higher TDP's with over-volted and over clocked quads.

 

[aigomorla]

Originally posted by: Rubycon

No it most definitely DOES apply. The wick allows use in any direction but there is a cooling capacity penalty. An unwicked pipe will work only in a designed position and deviating from such a position will result in very limited capacity or drop down to limits of physical conductance. (basically the pipe will conduct heat like it was empty)

This is why a heat pipe based heatsink will most definitely run hotter IF it's near its capacity when oriented in a non optimal direction. And for this purpose we have larger pipes and designs with more pipes to accommodate higher TDP's with over-volted and over clocked quads.

and im sure the mobo company or hsf company already took that into consideration.

but what she said...

 

[Rubycon]

Well the reason why most people see no change is they are well within the capacity - even reduced capacity from less than ideal positioning. Use a heatsink running close to the limit on the table and mount in a case where its axis changes 90 degrees and BOOM! Temps go crazy. This is exactly what happens with the Tuniq and a quad.

 

[imported_wired247]

let's cut some open shall we?

http://www.phoronix.com/scan.p...article&item=892&num=1

 

[lavaheadache]

Originally posted by: Rubycon
Well the reason why most people see no change is they are well within the capacity - even reduced capacity from less than ideal positioning. Use a heatsink running close to the limit on the table and mount in a case where its axis changes 90 degrees and BOOM! Temps go crazy. This is exactly what happens with the Tuniq and a quad.

yeah, i just had a thread based on this very topic very recently. RUBY IS RIGHT!!

 

[LesKelvin]

Thanks all.... I spoke to two different heat pipe engineers and their opinion is that orientation make a big difference. Yes wicking helps but I see many heat pipe designs that actually form a "P" trap. In this configuration... even if wicked, which most manufacturers don't disclose, the evaporation and transfer of heat hits a Road-Block. In the lower portion of the "P" trap HEAT must overcome both gravity and the weight of the fluid in the trap. Yes, one can always add more pipes but there is a limit there too.

I keep looking for a cooling device with 6 wicked heat pipes that come straight off the top of the cpu copper block and go straight up and OUT of my case (a lot of cases have top fan openings). Add the heat sinks and fans to dissipate btu's outside of the case and maybe a T.E.C. at the processor too.. so I can run at a constat 30C. Anyone seen one of these ?

 

[zagood]

Originally posted by: LesKelvin
I keep looking for a cooling device with 6 wicked heat pipes that come straight off the top of the cpu copper block and go straight up and OUT of my case (a lot of cases have top fan openings). Add the heat sinks and fans to dissipate btu's outside of the case and maybe a T.E.C. at the processor too.. so I can run at a constat 30C. Anyone seen one of these ?

No, but I'm sure you'll tell us when your product launches.