Concept Sun Cool 150W TDP slim heatsink
- Thread starter XCreator
- Start date
-
- Tags
- cpu cooler heat sink sun cool
It is a little bit difficult to explain so much importing information to you without me writing a few sentences...
Why even come to this forum if you don't want to read the feedback?
The Dynatron R25 is able to cool 160W due to the insane radial fan that is on it. Once again you only consider what you want to see = 22.28 CFM. Look at its static pressure of 74 mm-h20. That is 45 times higher than the Noctua A9x14 (but it also spins at 7000 rpm with 59 dBa). That high static pressure is needed to actually get around 22 cfm of air through that cooler.
Also the short fins does decrease it's heat dissipation property. But not as much as one would think. Remember that the longer away from the heat source, the less heat is dissipated from that area of the fins:
And the heat source can be heat pipes or vapor chambers (which also have similar loss of heat transfer over a distance).
So the R25 not really a good cooler to show as an argument. You don't think Dynatron would have used the Noctua A9x14 if it could supply ~30 cfm of air through the cooler at only 2200 rpm and 20 dBa? So again, if Dynatron need such an extreme fan (on a similar design cooler as yours) for just 22.28 cfm, how do you plan on getting 33 cfm (50% more) through your cooler without a 12000 rpm fan?
Also the video Ljspke linked is not an argument in your favor. It shows a similar design cooler to yours, with vapor chamber and the A9x14, that is just able to cool the i7 6700k@stock (91W TDP cpu). So again, I don't see how "just" a copper block is able to be better than these types of coolers. Copper doesn't transport heat that well internally compared to vapor chambers and heat pipes (as written in previous post).
That is why the Copper Dynatron K129 is only rated to 95W (even with its very high surface area)
"We got many unnecessary arguments when show concept without prototype, so some important designs we still keeping until make the real one."
Why do you think you get these questions? Because you boldly state that you have designed a new slim cooler that can cool 150W with a A9x14 without any really evidence.
As caniplaymayo said: "I think you're setting people up for disappointment."
And your answers to our criticism doesn't make any sense.
Our feedback and criticism (may come off as harsh) is in the end an attempt to help you not to spend money and time on a cooler design that likely won't be able to live up to your expectations and statements (the 150W cooling property).
Last edited:
Thanks for your information!It is a little bit difficult to explain so much importing information to you without me writing a few sentences...
Why even come to this forum if you don't want to read the feedback?
The Dynatron R25 is able to cool 160W due to the insane radial fan that is on it. Once again you only consider what you want to see = 22.28 CFM. Look at its static pressure of 74 mm-h20. That is 45 times higher than the Noctua A9x14 (but it also spins at 7000 rpm with 59 dBa). That high static pressure is needed to actually get around 22 cfm of air through that cooler.
Also the short fins does decrease it's heat dissipation property. But not as much as one would think. Remember that the longer away from the heat source, the less heat is dissipated from that area of the fins:
And the heat source can be heat pipes or vapor chambers (which also have similar loss of heat transfer over a distance).
So the R25 not really a good cooler to show as an argument. You don't think Dynatron would have used the Noctua A9x14 if it could supply ~30 cfm of air through the cooler at only 2200 rpm and 20 dBa? So again, if Dynatron need such an extreme fan (on a similar design cooler as yours) for just 22.28 cfm, how do you plan on getting 33 cfm (50% more) through your cooler without a 12000 rpm fan?
Also the video Ljspke linked is not an argument in your favor. It shows a similar design cooler to yours, with vapor chamber and the A9x14, that is just able to cool the i7 6700k@stock (91W TDP cpu). So again, I don't see how "just" a copper block is able to be better than these types of coolers. Copper doesn't transport heat that well internally compared to vapor chambers and heat pipes (as written in previous post).
That is why the Copper Dynatron K129 is only rated to 95W (even with its very high surface area)
"We got many unnecessary arguments when show concept without prototype, so some important designs we still keeping until make the real one."
Why do you think you get these questions? Because you boldly state that you have designed a new slim cooler that can cool 150W with a A9x14 without any really evidence.
As caniplaymayo said: "I think you're setting people up for disappointment."
And your answers to our criticism doesn't make any sense.
Our feedback and criticism (may come off as harsh) is in the end an attempt to help you not to spend money and time on a cooler design that likely won't be able to live up to your expectations and statements (the 150W cooling property).
Your say many many just for us to stop this project to save our money, thanks! But now, we are earning money ourselves to do it, you don't provide us money to do it so please stop do it.
We fabricate many many, some success, that good, some will be fail, that's normal, we gain some thing.
I'm not native English but TDP is Thermal design power, that mean I want to design a 150W heat sink, if the prototype not reach our desire, I will change the design.
Fruergaard is right. The heatsinks in the video are superficially similar to your one, however, they have a fan included, they aren't designed to be passive. They both have really rather thin fins, maybe no the absolute thinnest, but still incredibly thin, in comparison to yours, both have heat pipes, one using a vapour-chamber with vapour heat pipes.
This is a comparison of heat pipes with and without vapour chambers. If you have a "3D" vapour chamber, then at least, supposedly, the effects are further improved, however if you strip out even heat pipes, then you get even worse hot spots as you have nothing aiding the spread of heat across the heatsink.
If you use heat pipes or a 3D vapour chamber, you could cut down on the width of the thins, increasing the surface area, and at the same time increase the ability for the heat sink to carry, heat. You wouldn't need to increase the vertical height of the heat sink much if at all either.
Right now though, after looking at your design properly, I'm questioning if it would even provide average performance in a build with already good airflow. You need to run some proper simulations, not makeshift ones which might or might not work, and look at the temperature on different bits of the die. Look at a few different alternative setups, then test them in popular enclosures on a few different CPU's to see how they perform, again, taking temperatures on the die. Advisably testing them under slightly unrealistically high loads to give a margin of error and cover yourself.
Dude, you have to relax lol, nobody is trying to make you quit your project. Actually, it's the opposite, most people here have interest on low-profile cpu coolers, as many of us own a sff case that supports it. They are providing you very useful stuff, not just a random big text.
Also, I would like to add something: the extra piece on the side might be a bad idea. Mb chipsets don't need as much cooling as the cpu, and that feature will cause mb incompatibilities. If the buyer wants to cool their chipset, it's just easier to install a separate mini heatsink on it.
So, please hear people's feedback before spending tons of money on prototyping. Good luck anyway
Dude, you have to relax lol, nobody is trying to make you quit your project. Actually, it's the opposite, most people here have interest on low-profile cpu coolers, as many of us own a sff case that supports it. They are providing you very useful stuff, not just a random big text.
Also, I would like to add something: the extra piece on the side might be a bad idea. Mb chipsets don't need as much cooling as the cpu, and that feature will cause mb incompatibilities. If the buyer wants to cool their chipset, it's just easier to install a separate mini heatsink on it.
So, please hear people's feedback before spending tons of money on prototyping. Good luck anyway
Yeah, next time these guys post about a project, I just won't say anything. Apparently they know everything and don't need any input
Update 8-May-2018 Design change.
Estimate performance will be same water cooler if equip with 92mm x 25mm fan or bigger.
Estimate performance will be same water cooler if equip with 92mm x 25mm fan or bigger.
What are the dimensions on the heatsink? Is it solid copper or does it include a vapor chamber?
Thank, everything is delay a bit due it takes a few months to make Certificate of Eligibility for my boss business trip.I look forward to pre-production samples & benchmark (in case & open-air).
Keep up the good work!
Dim: 97x127x24mm
We thought about a phase change solution which is more expensive, but in this case, it's really not better as you thought.
Base thickness ~ 6mm solid copper provide same performance in this HS footprint.
See this graph to compare:
Last edited:
I'm sorry, I need to comment on this.
Performance like a water cooler? What water cooler would that be?
Can you link to where you found that performance ratio for heat transfer information?
To me a few informations is missing; heat source load (watt), fin structure and airflow (if any).
And what is the thickness of the copper at these three points:
This one has a big 22mm extruded in middle and fins thickness in bottom side is 1mm allow it to spread as well, the exchanged thickness will be about 6mm, result will be confirm by test the prototype.I'm sorry, I need to comment on this.
Performance like a water cooler? What water cooler would that be?
Can you link to where you found that performance ratio for heat transfer information?
To me a few informations is missing; heat source load (watt), fin structure and airflow (if any).
And what is the thickness of the copper at these three points:
Last edited:
Out of curiosity, what software have you used to model the heat flow of the cooler? Having done this somewhat recently myself I do find it hard to believe that this cooler is going to manage to dissipate 150W of heat reliably (that being said I wish you luck in doing so as it would be nice to have a low profile option like this).
You still haven't answered my questions?
1) What water cooler is it you believe that your cooler will be able to perform similarly to?
2) Link to the article or site that have the phase change vs copper data that you show.
3) What are the copper thickness at the three sites showed in earlier post?
A quick dig around and I'm pretty sure the diagram came from here: https://www.qats.com/cms/2017/07/26/vapor-chambers-solid-material-base-high-power-devices/
Seems to be a pretty reliable source. I'm also curious about what water cooler perform similarly to this cooler.
Did find that as well.
It refers to an article from 2002:
https://ieeexplore.ieee.org/document/1178760/
I would think that a thing or two have happened on the vapor chamber front in 16 years? I may be wrong.
If the chambers have become better the curves are shifted downwards = thicker copper base is needed for same ratio.
Second thing is if the 19x19 mm heat source in the predictions are uniform load distribution, which our CPU's aren't (way more hot spot from each core).
And I still don't see how their cooler dimensions hold up if there should be 6 mm thick copper at all the areas that is meant for spreading the heat (the three points I have asked about).
I'm suspecting now it's unlikely to provide any notable performance and will just about be on par with other stock heatsinks, with the nice feature of it being copper, but which comes at the drawback of cost.
It might, if they get incredibly lucky, be able to be roughly comparable to an old 92mm fan water cooler, however water coolers use case fans to blow air over radiators, so are typically at the very minimum 120mm and have been for a long time. A 120mm or 140mm case fan is entirely feasible in most if not all non-passive SFF cases and tbh there probably wouldn't be a mount for a 92mm fan because they aren't used anymore.
Plus, your right, understanding and precision of vapour chambers has come on a lot. There is even those funky 3D vapour chambers now too.
A more recent investigation into them which shows actual heat dissipation capabilities - https://www.thermacore.com/news/vapor-chambers-in-blade-server.aspx and also http://www.thermalnews.com/main/art...s-integrating-vapor-chambers-into-heat-sinks/ and http://www.electronicdesign.com/components/what-s-difference-between-heat-pipes-and-vapor-chambers
As I said VC is good for thin base.
No direct air flow in two side in this HS, higher side cool chipset as well, you can use 120mm fan but you will get compatibility and dust problems.
We also found a solution that bring this one a unbelievable performance, please wait!
No direct air flow in two side in this HS, higher side cool chipset as well, you can use 120mm fan but you will get compatibility and dust problems.
We also found a solution that bring this one a unbelievable performance, please wait!
