Overclock.net banner

Better than Direct Contact Heatpipes

2 reading
7.7K views 32 replies 15 participants last post by  Pantonious  
#1 ·
Heatpipes mounted on base:
Image


Direct contact heatpipes:
Image


The issue with direct heatpipes is there are terrible gaps. So how about base with heatpipes solder into the base? The base will have holes drilled into it horizontally. Then the heatpipes are bonded to the base. The inefficiency of the bond doesn't matter since the heat is transferred via the gas INSIDE the tube/base. This will give best of both designs. More efficient heat transfer to the heatpipe (one material instead of two materials and gap) while being about to maintain a flat base.

Image

 
#2 ·
I'm not getting what you are trying to say. Sorry.

In my mind the heat pipes do make more sense as you are not heating up a huge chunk of metal prior to the heat being transfered to the heat pipes. Which then stays hotter. But then again I have never tried the TRUE I just have a Xigmatek.
 
#3 ·
Quote:


The issue with direct heatpipes is there are terrible gaps. So how about base with heatpipes solder into the base? The base will have holes drilled into it horizontally. Then the heatpipes are bonded to the base. The inefficiency of the bond doesn't matter since the heat is transferred via the gas INSIDE the tube/base. This will give best of both designs. More efficient heat transfer to the heatpipe (one material instead of two materials and gap) while being about to maintain a flat base.

Sorry, I can't "bold" or "italic" enough of that statement to try and emphasise the point you are trying to make.
 
#4 ·
Quote:


Originally Posted by Tufelhunden
View Post

I'm not getting what you are trying to say. Sorry.

In my mind the heat pipes do make more sense as you are not heating up a huge chunk of metal prior to the heat being transfered to the heat pipes. Which then stays hotter. But then again I have never tried the TRUE I just have a Xigmatek.

Heating up the chunk of metal? Thermodynamics says this doesn't matter since the heat will move to the heatpipes since they are cooler. The issue is that there are two additional thermal inefficient layers (the base and pipe-base interface). My design removes this.
 
#6 ·
Quote:


Originally Posted by noldevin
View Post

I think he means that they should just solder the heatpipes into coolers such as a Xigmatek instead of just letting them rest or using some kind of epoxy.

Yup, solder open heatpipe ends into the base. Make the base PART of the heatpipes.
 
#7 ·
Actually the direct touch is a pretty sufficient design compared to a regular plate-pipe-plate design.

They just need to work on it a little. My Xigmatek is great but I was considering taking some solder and filling in the little gaps between the pipes and then doing some sanding with a dremel and then lapping.

Now that would be a better design.

Since the solder's melting point is around 90c? it would heat up faster and maybe help even enhance the heat dissipation. Of course, I have no scientific data to back that thought up but... whatever
Image
 
#8 ·
Quote:


Originally Posted by DuckieHo
View Post

Yup, solder open heatpipe ends into the base. Make the base PART of the heatpipes.

Now I get it sorry. The pain medication has kicked in and its about bed time.

I had heard about this and was going to get the little brother to braze this for me. He's a plumber with all the cool tools and good brazing rods.
 
#10 ·
Quote:

Originally Posted by porschedrifter View Post
Actually the direct touch is a pretty sufficient design compared to a regular plate-pipe-plate design.

They just need to work on it a little. My Xigmatek is great but I was considering taking some solder and filling in the little gaps between the pipes and then doing some sanding with a dremel and then lapping.

Now that would be a better design.
My design already takes care of that by NOT having any gaps in the first place. Plus it would provide better heatpipe to CPU contact
 
#12 ·
Quote:

Originally Posted by dr4gon View Post
yeah that's what bothers me about the xigmatek cooler.... looks interesting, might be time consuming to make? (expensive?)
Charge $15 more and call it "Beyond Advanced Direct Contact Super-Duper Heatpipe Cooling". You know we will buy it.
Image
 
#15 ·
Quote:

Originally Posted by DuckieHo View Post
Charge $15 more and call it "Beyond Advanced Direct Contact Super-Duper Heatpipe Cooling". You know we will buy it.
Image

That's a catchy name right there, Quick get it patented
Image
 
#19 ·
How are these gaps a huge drawback if the Direct touch heatpipes have some of the lowest temps out of the cpu coolers?

First off the gaps are minimal... Secondly use a little extra thermal paste and they fill right up
Image


I actually think i am gonna solder and to some tweaking to fill in the gaps anyway just to see if I yield a -2c difference
Image
 
#20 ·
Quote:

Originally Posted by porschedrifter View Post
How are these gaps a huge drawback if the Direct touch heatpipes have some of the lowest temps out of the cpu coolers?
I'll be honest I'm not sure if I have the coolest temps with this thing or not. According to Core Temp 0.97.1 I idle at 40C and Max load to 65C on OCCT after 5 minutes. I have a fairly good air flowing case, CM690, with so many damn fans it sounds like a jet engine.

Could just be the chip though I have read the QX6700 just isn't really meant for aircooling.
 
#21 ·
what if there were just more heatpipes and nothing in the middle no gaps then, polish them all together to a smooth surface. or just the existing ones were either bigger or flatter so they can be uniformly polished .

I also was under the impression that the heatsinks with the pipes on the bottom as opossed as the ones sandwhiched worked better already.
 
#22 ·
Actually what I would like to see is the heat pipes staggered as they go through the fins. I would imagine this would lead to better cooling than we currently see with either the True or Heat pipe coolers. Then again I could be way off. :Shrugs:
 
#23 ·
on my 5000+ i'm idling at 26c

why cant the only plate on the bottom be a thin hollow metal plate with pipes welded to the sides. They could prob forge the whole thing. Oh wait that's what duckieho thought of
Image


And are the pipes filled with water or just air?

Personally I feel that the future of cpu coolers are going to be variants of the TEC coolers

http://en.wikipedia.org/wiki/Thermoelectric_cooling
 
#24 ·
Quote:

Originally Posted by porschedrifter View Post
How are these gaps a huge drawback if the Direct touch heatpipes have some of the lowest temps out of the cpu coolers?

First off the gaps are minimal... Secondly use a little extra thermal paste and they fill right up
Image


I actually think i am gonna solder and to some tweaking to fill in the gaps anyway just to see if I yield a -2c difference
Image

Gaps are mininum... but they still exist. Why do we lap? Because there is a gap... even though most ppl would call it minimum. Thermal paste is few hundred times less conductive than metal.

Also, adding solder wouldn't be as efficient since the heat has to go through CPU-solder interface, the solder, and then through the heatpipe-solder interface. A base acting as a heatpipe would only have the CPU-base interface.

Quote:

Originally Posted by porschedrifter View Post
on my 5000+ i'm idling at 26c

why cant the only plate on the bottom be a thin hollow metal plate with pipes welded to the sides. They could prob forge the whole thing. Oh wait that's what duckieho thought of
Image


And are the pipes filled with water or just air?

Personally I feel that the future of cpu coolers are going to be variants of the TEC coolers

http://en.wikipedia.org/wiki/Thermoelectric_cooling
Heat pipes are filled with a working fluid that changes to gas.

TEC won't ever be mainstream since they have to be tweaked to a specific system (as in the load, ambient, and cooling have to be known within a few degrees or it will be bad). Also, TEC draw much more power than other more common cooling.
 
  • Rep+
Reactions: porschedrifter
#26 ·
Any other thoughts?