DAN HSLP-48: A powerful sub 50mm heatsink

No one knows, we can theory all we want, but we [Dan] cannot be certain until prototypes & results are released. This heatsink might not even be worth making if there is a 1 deg difference between the l9i or the c7.

Everyone is hoping something miraculous comes out, but tbh I think max change would be 5 Deg C -- which is pretty good, something I might be willing to drop $100 USD on.

And there is the question, even if the CPU has lower temps, what are the other mobo components temps at? IF it lowers CPU by 5 deg, and heightens everything else...

Also, remember if this does comes out, it is a while out for now.
If it is as good as the Nexus I am happy.
 
The aim is 15-20°C better than the L9i so nearly the same level like the NH-L12.
Great! Close to Nexus Low-7000 temps or better should definitely be the goal.

Any ETA on your cooler dondan?
This may seem like a rushed question, but are we a year away or do you think it will be more?
 
Great! Close to Nexus Low-7000 temps or better should definitely be the goal.

Any ETA on your cooler dondan?
This may seem like a rushed question, but are we a year away or do you think it will be more?
I think it would be easily longer than a year considering what has to fit into the timeline.

Imagine dondan wants to use Kickstarter for this as well. Before the campaign can even start he has to find a manufacturer who is willing to make some prototypes for him, then he has to test the prototypes for performance/feasibility. In this stage there may be many iterations on a design before a final design is picked. It's probably at this stage, once a final design is picked and a manufacturer is willing to make it, that dondan would want to launch a Kickstarter campaign to raise funds for mass production. Then you have the duration of the Kickstarter campaign, duration of manufacturing and packaging, shipping logistics, etc.

I would personally think this project won't see a completion (all backers have the heatsink) until sometime in late 2018 at the earliest.
 
Before the campaign can even start he has to find a manufacturer who is willing to make some prototypes for him

Nah bro, you don't understand how the Kickstarter game works. He can either mod a AXP100 somehow and call it a prototype or just build one out of cardboard or something while still searching for a manufacturer ;)
 
Nah bro, you don't understand how the Kickstarter game works. He can either mod a AXP100 somehow and call it a prototype or just build one out of cardboard or something while still searching for a manufacturer ;)
He could literally doodle the cooler with chalk on the back of a closet door, upload a picture to kickstarter, and still have a successfull run on kickstarter at this point.
 
I think a shorter version of the Thermalright AXP 100H Muscle would do, I'm not sure. (With 180w of max TPD so a i7-7700K won't throttle with high overclock on the A4-SFX):nailbiting:
With a Noctua fan, so it can be quiet.
If it's only copper, it would look AWESOME!!!!!!!!!!!!! :cool::artist: I wouldn't bother if it's all black :D
Like the A4-SFX, I wouldn't care about price. It would be "enthusiast" stuff!
Keep up the good work, #dondan
 
With 180w of max TPD so a i7-7700K won't throttle with high overclock on the A4-SFX

Yeaaaah.... As awesome as that would be, that's just not going to happen.

It would be an impressive feat for it to handle keeping a 95w part under 75c. For it to cool a significantly overclocked chip is, well... Asking it to break physics.
 
Do you think this motherboard should work with your cooler ?? Asrock Fatal1ty Z270 Gaming-ITX/ac



Or do think the heatzink back against the io are a problem ??
Fatal1ty%20Z270%20Gaming-ITXac(L3).png
 
Here is the next big update:


What do you see is the copper version. The fin surface is now width 130mm and depth 120mm. The fin thickness is 0,5mm and the spacing 1,5mm.

Mounting Solution:
I add a black front shield for a nice look. If you unscrew the shield you will find 6 clip holes. After you installed the heatsink select the best fitting position for the fan.
Now select the suitable holes and the fan is secured. This solution will only clamp the fan between clip and heatsink so you can modify the position of the fan after mounting with pressure.
You will be able to mount 120mm, 100mm, 92mm or 80mm fans with a thickness of 14-15mm.


Air bypass blades
If you use a small 100mm or 92mm fan, the fan will not reach the complete surface with fresh air. To get more air to the sides, I add a row of bypass blades on each side that shift a part of the air to the sides.
If you use a 120mm fan you can unscrew the bypass blades.


Here are two renderings:

key2.88zcjho.jpg

key2.89jdk8f.jpg
 
Could you explain the mounting system a little bit more? I'm intrigued by the ability to move it after mounting, but I don't really get how it works.

Same thing for the bypass blades - from the renders, it looks like they're just sitting next to the fan? Clearly I've got to be wrong about that, because they'd be pointless if that were the case.
 


Air bypass blades
If you use a small 100mm or 92mm fan, the fan will not reach the complete surface with fresh air. To get more air to the sides, I add a row of bypass blades on each side that shift a part of the air to the sides.
If you use a 120mm fan you can unscrew the bypass blades.

I have tried this before.

Please do test it, but I have found that the "edge" of the blade seems to creates wind turbulence and so adds noise. Not a lot, but what it does is to increase the pitch of the sound.
 
Based on the renders, I would make the following interpretations

The fan clips appear as though its a "C" shaped spring steel wire that hooks into the holes in the ends of the fins. By having multiple holes for the wire to hook into, it allows some flexibility with where the wire will be placed. It sort of acts as a sling holding the fan up against the heatsink fins. The wire would be inserted by first inserting one end of the wire into the hole on the heatsink, stretching/bending the wire gently with the other end and carefully lining up the other hook so it goes into the hole on the other side of the heatsink. All this whilst holding the fan steady and in place. Should be doable if done on a desk. After its installed, you can simply move the fan around by pushing it where you want it to go. Assuming the fan isnt too firmly secured, and it doesnt have too much vibration damping rubber on it, it should slide.

The bypass blades I presume are a pressed metal sheet that is installed from the top side of the heatsink and has metal fingers that fit between the heatsink fins. They divert the air that would be blown by the outer diameter of a smaller 92mm fan outward so that there is some airflow over the last heatpipe. This is to share airflow from a smaller fan over the outer heatpipes with the goal of increasing the surface area of fins/heatpipes blown by a smaller diameter fan.


dondan
I am a little concerned at the difficulty of installing the wire clips. Its a fairly common thing to hear about people have difficulty with the wire based fan retention clips, ones like the Noctua on the U9B heatsink. Your design looks much more fiddly with a chance of bending fins. Hopefully not as bad in practice if the spring tension isnt too great, but I can forsee many users bending fins or at least scratching up the end fin or bending the spring wire too far. A bent wire if not corrected could then rub on the fan blades.

The other concern is the retentive force such a wire could apply with its current design. Maybe introducing a Z bend or a curve on vertical wire between the hook and the fan in the same plane as the fins might permit some degree of springiness and adjustability. This would allow for variances in fan thickness or vibration damping materials, might help give you a couple mm of leeway where you can ensure there will be sufficient retentive force on the fan so it stays put. Doesnt' make it any easier to install though if we are trying to achieve a degree of tension or pre-load.

I personally like the design as it will give a nice clean result, has great flexibility as it doesn't bother with the mounting holes of the fans, and should work with practically anything. I have minimal concerns over the fiddly nature, but it looks like it would be best installed with a steady hand and maybe the help of pliers if one didnt want to scratch up the fins too much before covering it up with the end plate.
 
Last edited:
Here is the next big update:


What do you see is the copper version. The fin surface is now width 130mm and depth 120mm. The fin thickness is 0,5mm and the spacing 1,5mm.

Mounting Solution:
I add a black front shield for a nice look. If you unscrew the shield you will find 6 clip holes. After you installed the heatsink select the best fitting position for the fan.
Now select the suitable holes and the fan is secured. This solution will only clamp the fan between clip and heatsink so you can modify the position of the fan after mounting with pressure.
You will be able to mount 120mm, 100mm, 92mm or 80mm fans with a thickness of 14-15mm.


Air bypass blades
If you use a small 100mm or 92mm fan, the fan will not reach the complete surface with fresh air. To get more air to the sides, I add a row of bypass blades on each side that shift a part of the air to the sides.
If you use a 120mm fan you can unscrew the bypass blades.

I must admit I lack the more technical insights to the cooler design, but that sounds quite nice with the placement and fan options. Also it's not bad to look at.
I completely agree with okwchin no concerns over fiddly when you gain so much.
 
A piece of beauty! If this works out like the case, I'll take one!

Would the choice of an 120mm fan require the use of VLP-RAM, as the heatsink is planned to cover the RAM bays?
 
Last edited:
Nice work Dan! So Heatpipe Direct Touch is out of the question?
if the whole block is copper it doesn't make that much of a difference in most cases since copper has a really high thermal conductivity per volume.

dondan
Can you give us an approximate weight?
For me that is basically the only way I can know how well this heatsink could perform xD

Edit:
My guess, just from looking at the picture, is about 700 grams. Is that in the ballpark?
 
Last edited:
if the whole block is copper it doesn't make that much of a difference in most cases since copper has a really high thermal conductivity per volume.

The problem isn't the heat transfer in the copper, the problem is the heat transfer between the heat spreader, the copper, and the heatpipes. The heat transfer from one body to the other is significantly slower than inside that body. With direct touch you eliminate one of those boundaries.
 
This is ID-COOLING IS-50 with 5 Direct Touch heatpipes:

IS-50-3.jpg


They also have IS-60 with 6 heatpipes, without Direct Touch though.
IS-60-6.jpg



Dan, maybe you could take some inspirations from them?
 
You asked for more technical details:



Outer Heatpipes:


To stabilizes the outer fins next to the screwdriver holes, I shifted the outer heatpipe more to the outside and made the small fins wider so the heatpipe goes through it.


1c5j1u.jpg



Furthermore I add a tiny bend to the sides on every fin to improve the stability.


4xgkzo.jpg




FAN Clip mounting:


You see there are six mounting points for the clip on the front side and backside of the fin surface. You can choose the pair you need for mounting the included or a different fan.

1. Step: Mount the heatsink on the socket.
2. Step: Move the fan under the heatsink in a position you like. (Hold the fan with one hand in position)
3. Step: With your second hand take one clip and put one end in a hole on the backside. Stretch the clip and put the other end in the hole on the front side.
4. Step: Redo the third step for the second clip.


3gek69.jpg




Bypass blades:

Here is a picture that show you how the bypass blades will work. If you take a look on my renderings some post above you will see there are two bigger gabs. Theoretical we need only one bigger gab for the screwdriver.
But I need this part for both sides, so if you rotate the bypass blades for the other side you will see that the gab is now on the wrong position. This is why I add two bigger gabs.


239jy9.jpg


Note: All pictures are with removed front plate to show the details on the fins.
 
Last edited:
I don't know much about aerodynamics, but wouldn't this be a better solution from what it looks like? Just an idea.

1.jpg


By the way, would a small 120mm fit under the cooler under certain conditions?
 
Last edited:
I think the "air-diversion-blades" are not gonna add any cooling performance and may add noise, but probably it won't make a difference either way. I think not adding them would be better and less confusing. Those of use who will put a smaller fan will just have to live with less cooling capacity. But I think that this is exactly the beauty of this cooler, you have the freedom to put different fans in different places such that everybody can find a combination that works for their specific case.
 
I think, just forget about the air divertion blades, it's better to not have a seperation in the fins to allow for them to fit because its going to compromise the heat dispertion and the two copper pipes at the end will be hotter and wont be able to cool even with the blades installed. thats just my opinion, but you can make two prototypes and use a heat camera to verify if you want :)
 
thermalright might be interested now, considering the sudden interest in their copper axp-100. check out the other dan a4-sfx thread pg. 241-242-243
they are clearing their engineering stock as we speak , 100$ each too!
 
Today I god a feedback from Noctua. Currently they are validating if a heatsink is possible with their current rentention kit. They send me a drawing based on a NH-L12 with 4x8mm heatpipes and the dimensions of my finlayout.

Furthermore I made a little change on my current design. I moved the complete surface 4mm more to the back. So the surface will be the same but the total dimensions of the heatsink is smaller.

571jng.jpg
 
Back
Top