AIR COOLED WATERBLOCK







    Finally, the solution to do it all has arrived, and the simplicity of it all is amazing!  I have been pondering how to combine the two technologies for almost a year, and then finally found the right materials (and tools) to finally accomplish the task.


    What better way to cool a CPU then with water right?  Well what happens when your pump goes, or your line gets kinked, or any other myriad of things that can go wrong with an unmaintained watercooling setup?  Your CPU fries.  However, imagine a system where you can go to sleep at night, and not worry if any other component fails.  Imagine a safeguard that allows you to use this block just by itself without a drop of water going through it.  This is the solution to safeguard water cooling. 

     The performance of this block is uncontested by conventional waterblocks.  I did my own personal comparison with the Heat Killer LT block.  The comparison was done as unbiased as I could possibly conduct.  These are the components used to test:

Swiftech 120x2 radiator
Swiftech MCP 655 pump
3/8 ID tube
X58 Micro with i7 920 (lapped)
Ambient temps cycled between 70-74F
Artic Silver Alumina & all fans 100%

     The test consisted of putting the same amount of thermal paste on both waterblocks, installing, running Prime95 (blend) for 6.5 hours straight, & leaving realtemp on to record max temps.

     First, here is the picture of the setup, with the heatkiller block installed in the Antec 300 case:

As you can see, just a very basic setup.  I didn't do anything extravagant like use bigger radiators, or reservoirs... Just the basic components.  The fans are low CFM Coolermasters. 

The results of the 6 hour test:

 

As you can see the max temp recorded was 71C.  All settings were stock bios settings (less the 1.65 Ram setting) 

NOW, to move on to my block...

The same setup was used:

No change, just change the waterblock with the same quantity of the same thermal paste, and repeat said test.


Here it shows good spread of the thermal paste, and the overall size comparison:

And here is that part you have ALL been waiting for.... the RESULTS:

Clearly a marked improvement over the Heat Killer block.  Nine degrees of improvement using the same setup. 

     As with everything there is always room for improvement.  I can refine my methods, use different equipment etc, but overall I am quite pleased.  I use the block as we speak for just air operation.  The fan will get quite noisy when the smart fan kicks the fan up to 100%, but thats what it is supposed to do.  U manually changed the CPU fan to 100% in the smart fan because when your benching you want it as cool as possible.  Also because the Smartfan worked a little too well and allowed the CPU to reach 59C before it kicked in.  Kudos to the board & smart fan for doing its job.






    

Attachment(s)

Attachments are not available: Download requirements not met

Finally nate you show some pics! Very nice work. 10*C over the HK is insane! How is the compatibility though with other boards for the block?

That's fantastic Nate!!  9C is HUGH!  It would be interesting to see the temps on air too....with a fan and passive.

the air block is looking damn nice...how many different types of boards does it fit on?

Let us not forget that this block performs just as well as any of the air cooled blocks on the market as well. so if you are the type that likes to try to go for the air cooled and the water cooled records then this is the block for you or maybe you are a air cooled guy and are thinking about making the change over to water cooled but you are not quite there yet this is the perfect block to use and you can transition at your leisure and never worry about heat issues from your cpu.
 
this is personally my favorite of all the blocks he has engineered. it’s so simple in design and yet performs with the best of anything on the market not to mention the ability to act as a cross over between the air and water cooled is simply amazing.
 
this is blue ribbon material all the way !!!!

great job, any future plans with these things?

Very nice! Great job Nate, that's really awesome! You gonna market this too?

That some awesome performance you're getting out of that thing.  Whats up with the huge difference in vreg temps?  Is it due to the block being larger and covering the vreg chips?

did u make the block, or buy it, the (water/air ) one.

cudenver


did u make the block, or buy it, the (water/air ) one.

I made it in my basement.  The picture of the flat plate next to the blue can is the bottom plate which I milled in my traditional ghetto fashion. 

As for air performance, it is on par with most of the major manufacturers.  I did some comparisons with THIS article, but my thermal paste was different.  The only thing that was similar with his test and mine was the 920 chip idled at 37°C with the Intel Factory Heatsink.  I would have to do a controlled experiment with one of those 1366 coolers, and I don't have one.

BTW, the heatkiller block was lapped all the way down to 2000 grit (along with the chip) vs my block was lapped to only 1000.  Also, the test of my block had the additional 9800 GTX+ pumping additional heat, so I the test was slightly skewed in the HK's favor, but what the hell right? 

mcarey6


That some awesome performance you're getting out of that thing.  Whats up with the huge difference in vreg temps?  Is it due to the block being larger and covering the vreg chips?

My block has the fan blowing at the heatsink (which performed better then drawing the heat off)  so it blew directly on the vregs.  The Heatkiller block has no moving air.

how soon are you taking orders 

Tribbs


how soon are you taking orders 

*cough* PM me *cough*

I know it will fit the 759 X58 board, but the biggest hurdle is caps (if I even call it a hurdle).  This block can be made to fit any 1366 socket and I can just simply extrude any area that comes near a cap. 

In the current configuration that this version sits, it hangs off the board about .5" (i think, i have to measure)  The next version can orient the nozzles closer to the ram slots over that wasted space.  It will be VERY close to the first ram slot, but the benefit is that it will also blow air over your ram.  The fan that I use moves a TREMENDOUS amount of air so it cools other components dramatically (almost 4 thousand RPM, as shown in the eleet)

nateman_doo

cudenver


did u make the block, or buy it, the (water/air ) one.

I made it in my basement.  The picture of the flat plate next to the blue can is the bottom plate which I milled in my traditional ghetto fashion. 

As for air performance, it is on par with most of the major manufacturers.  I did some comparisons with THIS article, but my thermal paste was different.  The only thing that was similar with his test and mine was the 920 chip idled at 37°C with the Intel Factory Heatsink.  I would have to do a controlled experiment with one of those 1366 coolers, and I don't have one.

BTW, the heatkiller block was lapped all the way down to 2000 grit (along with the chip) vs my block was lapped to only 1000.  Also, the test of my block had the additional 9800 GTX+ pumping additional heat, so I the test was slightly skewed in the HK's favor, but what the hell right? 

that is some good work you did there. I'm just getting into this water cooled world and you have a some very unique ideas.

 

 ImageShack.us

Totally awesome Nateman and congrats on the results!

Milled in the basement, just amazing.

PS, whatcha got against the shore ?

Thanks!  nothing against the shore, just that awful show :P

Nice job.  Definitely works well.

utnorris


Nice job.  Definitely works well.

side note: 
dont take this the wrong way...but ummmm.
*smack upside the back of the head - like the original 3 stooges show* 


pctools had to point me to your frys thread. and i got over there like a mad man!...lol

Nice work Nate.

Just wondering if you are going to make a waterblock with the same design.  I always wanted to know how well a waterblock performs with a fan attached to it..hehe

atfrico


Nice work Nate.

Just wondering if you are going to make a waterblock with the same design.  I always wanted to know how well a waterblock performs with a fan attached to it..hehe

umm... that is a waterblock with a fan attached to it. 

Just some different spacers and springs I could find.  Not a fan of the blue, but the bronze spacers match the block. 

Attachment(s)

Attachments are not available: Download requirements not met

not a problem.  i know the feeling!
you need more coffee? or you drank too much of it.

nateman_doo


not a problem.  i know the feeling!
you need more coffee? or you drank too much of it.

....need more coffee  Brain is on vacation, but body is in auto-pilot at work..

Your going to want standoffs between inside the fan, your mounting pressuring should be able to go way up. You could also put bolts at the bottom of the fan, that way you dont have to deal with the fan flexing, and dont need standoffs.

Looks beautiful nate. Im really surprised on how "Non-Ghetto" it looks. I really want to know how you manged to get those fins on. Did you just buy a copper plate with fins pre attached?

You should send the block to Vapor or Skinee for an in depth analysis :)

heh... I am sure their results would be similar to mine, as far as the difference in temps.  But I am sure they have much better equipment and analysis methods to better control variables then I do. 

Standoffs "in" the fan aren't necessary, you don't have to torque the heck out of the nuts to get it to stay.  As you can see it seats quite well with a vertical mobo.  Since its not very tall (the block at least) it has a low center of gravity.  I will look into such improvements such as more standoffs from the board to the block etc, and those in the fan.

this is an awesome idea! looks great keep up the good work 

One more congrads from me, Nateman_doo, great job.

Just so everyone know how this idea works is because all coolers absorb some heat (which is why temperatures are raise above ambient to begin with) and dump the rest.  It's not that the typical waterblock isn't doing it's job, 100% heat removal is simply impossible without a heatpump.  Whatever heat the block absorbs, in this case, the fins try to take away.  They don't removal all of it, again, 100% is impossible. 

It's mainly the thermal resistence (TR) factor that tells you, which is the percentage of absorbsion.  Waterblock by themselves can be as low as 0.05, or 5%.  Nate's block is less than that, maybe 3-4%. 

Calculating the TR = (full load temp - ambient local to source)/wattage of source.

Note: Ambient isn't just one reading in a room or general case temps.  It is the coldest reference to affect the source; for an air cooler that temp entering the cooler, for water it is air temps entering the rads.  Depending on system configuration, these two might be different by a few degrees. 
 
His heatfins are what I'd regard as the perfect synergy of a prexisting cooler, a Dynatron G129 if memory serves.  I still think you should have tried out the G118 or G218 with the vapor base, it would have made one hellva cooler.

That's one huge block lots of mass, very nice. Looks like it would hold chiller temps really well!

ShockTheMonky


I like what I see. Bounced the same idea around myself wondering how well it would work. Great job.

Now what's the possibility to mass produce? Would like to see something like that in action on my board.

shoot me a PM.  :)