Direct Heat Spreader Cooler

[jaydee]

Time to put the theory into practice and see if this is a viable option. There has been some discussion on the subject over the years and some may have done it but if they have i cannot find it. Therefor I have constructed this to see if all my theories about this approach are reasonable.


Theoretical cons:
1) IHS has no center pressure to make the IHS contact the die well.
2) Laminar flow and low surface area making heat removal less efficient.

Theoretical positives:
1) Removes TIM layer.

Known positives:
Very cheap to make.

I already ran a few other blocks (MP-05-SP, Stock Retail Sempron Cooler and a DD Maze4) through the test bench on this Sempron CPU at stock speeds and overclocked. So if the thing doesn't fry right off the bat then I will have some comparative results.

[jaydee]

Leak test has passed just fine. over 10hrs no problems. Will go live with it tomorrow.

[jaydee]

I ran this tonight. Results were pretty much as expected.

Sempron 2500+ @ 1862mhz approx 82watts stock Vcore 1.45V

Direct IHS Cooler:

@3.6GPM water to cpu differential was 12C. (Water 25C, CPU 37C). I decided to stop there simply because there is no point in going further.



Stock AMD Sempron Air Cooler 50C CPU temp (32C air to CPU differential)



AquaXtreme MP-05 SP 432pin base:

@1.5GPM water to CPU temp differential 10C

@1GPM 11C

@.5GPM 12C



Danger Den Maze 4

@2GPM 14C

@1.5GPM 15C

@1GPM 15C

@.5GPM 16C



So obviously if the direct IHS cooler is performing worst at over twice the flow rate than the MP-05 SP then it shows this is not a viable option. I don't think modifying the IHS will make it enough better to even beat out the Maze 4.



So in my mind project successful and enough info to know this is not the path to follow. A water block will perform better and less chance of failure.

[ben333]

Bastard! Said the mill was all packed up! I'm glad that somebody finally did this so I didn't have to do it myself with goop and zip ties I think you should give direct die cooling another shot. Any chance you'd sell that over IHS block you made? If it was cheap then it sounds like a good idea for budget cooling.

[bigben2k]

What if you used a thermal adhesive, and mount a small copper fin between the inlet and outlet? It could also act to put some clamping pressure on the core.

Just an idea.

[jaydee]

Originally Posted by ben333

Bastard! Said the mill was all packed up! I'm glad that somebody finally did this so I didn't have to do it myself with goop and zip ties I think you should give direct die cooling another shot. Any chance you'd sell that over IHS block you made? If it was cheap then it sounds like a good idea for budget cooling.

And it is all packed up again. I can get away with milling plastic in the apartment but metal is a quick visit from the manager.

There is no point in direct die. Not enough surface area for the amount of heat put out to be effective. A 1/8" flat copper base plate water block would probably work just as good. That and I got no bear dies to work with. These IHS's appear to be epoxied on so they are not coming off with out the core attached.

Also out of motivation, moving back to RC's for a while being the weather is finally getting good around here again. Long winter...

I did order new PC parts though.
2gig dual channel DDR2 PC6400 G-Skill. $35shipped
MB BIOSTAR TFORCE TA780G M2+ AM2+ R - Retail $89.99 shipped
CPU AMD A64 X2 4200+ 2.2G 65N 1M R - Retail $59.99 shipped

Not a bad upgrade for $184.98.

[jaydee]

Originally Posted by bigben2k

What if you used a thermal adhesive, and mount a small copper fin between the inlet and outlet? It could also act to put some clamping pressure on the core.

Just an idea.

My orginal plan was jets with a center inlet but not enough room for an outlet. I didn't have time to make a 2 peice top to accomadate it.

Peice in the middle might help but highly doubt it would help enough to make much difference. With the onboard thermal probe I doubt I could measure it.

This system is pulled apart and already running Linux with the stock air cooler. I learned all I needed to know.

And no ben333 not selling it. This block will only work on a socket 754 platform anyway. Not satisfied with the quality of the work on it either. Wouldn't take much to break the o-ring free and soak a system.

[ibmkg]

What kind of a tool you used for making this? What is that in the pic and can it cut Cu?

[killerbong]

hehe i remember trying this with an old pentiums, remember lowsy temps.

[ben333]

Originally Posted by ibmkg

What kind of a tool you used for making this? What is that in the pic and can it cut Cu?

I believe thats his CNC mill and I'm pretty sure it can do copper

[jaydee]

Originally Posted by ibmkg

What kind of a tool you used for making this? What is that in the pic and can it cut Cu?

Got it from here: http://www.acumotion.com/index.asp

Milled many a copper blocks with it but it is very slow going. Would suggest a better mill for it.

[Brians256]

Good job, Jaydee!

Jaydee, you probably remember, but others may not remember some attempts made to directly cool the silicon die with water. Removing the IHS is not a good idea, because the silicon die is water permeable. So, a normal waterblock is about as good as it gets unless you remove the IHS and cool the die with the waterblock.

[derelict]

Cool, puts my mind at rest about trying to put water straight onto a core.

Q: How are people making the O rings? Are you using a premade set size ring, or is there goop that you fill a groove with and it "hardens" into a rubber O ring? Newb question I'm sure but I don't know enough about this whole o-ring gasket process.

[derelict]

Oh I also got some new parts, bought it off craigslist

DFI P965-S Dark + C2D E6550 + 4gb DDR800 4-4-4-15 (OCZ) = $165

1 stick of the ram was bad, so have to RMA

I could at least POST at 500x7 without trouble, once I get an OS I'll see how stable that really is and maybe go beyond if I can.

[jaydee]

Originally Posted by derelict

Cool, puts my mind at rest about trying to put water straight onto a core.

Q: How are people making the O rings? Are you using a premade set size ring, or is there goop that you fill a groove with and it "hardens" into a rubber O ring? Newb question I'm sure but I don't know enough about this whole o-ring gasket process.

I buy oring material by the foot. Cut it to size and super glue the ends.

[derelict]

Cool ok thats easy enough

[bobo5195]

I see thats pretty much as expected.

This is clearly a case for jet impingement if there ever there was one. heff for that is 120 W some area (centimeters squared??).


My plan for ultra cheap mass market wter block was to use my local uni, maker of the worlds smallest holes to make a simple holed plate for jet impingement, sub milimeter. The key thing with jets is the smaller they are the better as cooling scales with size (relative area gets smaller per jet but relative velocities increase much faster). The jets need to be desinged to provide a good cross flow (flow across other jets to help them cool). If done right the numbers in papers i have read indicated it could just about beat current blocks.

Because of the method of manufacture though it should be cheap.

[jaydee]

Originally Posted by jaydee

I did order new PC parts though.
2gig dual channel DDR2 PC6400 G-Skill. $35shipped
MB BIOSTAR TFORCE TA780G M2+ AM2+ R - Retail $89.99 shipped
CPU AMD A64 X2 4200+ 2.2G 65N 1M R - Retail $59.99 shipped

Not a bad upgrade for $184.98.

I got this system setup and using it to type this right now. Been playing a few games with it and so far am I very impressed with the lack of noise. I am using the STOCK AMD heatsink and fan. The fan is of the 4 pin type for automatic speed adjust. The CPU is dual core 65 watts. This thing runs near silent. My power supply fan on the Antec True Power Trio 650 power supply is the loudest and it is hardly noticable. CPU temps are under 40C even when gaming. I now remember what hard drives sound like.

If this is the way CPUs are going there is little point in water cooling on a PC level.

[Brians256]

Originally Posted by jaydee

I got this system setup and using it to type this right now. Been playing a few games with it and so far am I very impressed with the lack of noise. I am using the STOCK AMD heatsink and fan. The fan is of the 4 pin type for automatic speed adjust. The CPU is dual core 65 watts. This thing runs near silent. My power supply fan on the Antec True Power Trio 650 power supply is the loudest and it is hardly noticable. CPU temps are under 40C even when gaming. I now remember what hard drives sound like.

If this is the way CPUs are going there is little point in water cooling on a PC level.

Yup.

[compro]

Have you considerd jet impingement-based direct die cooling, the heat spreader shoild be thick enof for small cups and with 24 jets it shoild work vary whell?






sorry jaydee I mised it,

"My orginal plan was jets with a center inlet but not enough room for an outlet."

an off center inlet woild be just as good.

Man got to read the hole thred before posting, I have akses to a mill and some drill bits down to .35mm to I may just try this but not on a cpu,I'll get a pice of aluminum and put an a cople power resistor as a heat sorce, 150w shoild do.

[ben333]

Seems at that point you might as well just use a waterblock? A flat waterblock should be much better seeing that the contact between the IHS and die isn't great.

[compro]

if I did this it woild just be for fun, but it shoild workalmost as whell as a storm g4 (g4 has 62 jets I think but one more thermal joint)

[compro]

2 100w cartrige heaters and a light dimmer

[compro]

I have a 500 mhz celeron I'm going to try this on. When I have some pics I'll make a new thread.

[derelict]

/offtopic

Originally Posted by derelict

Oh I also got some new parts, bought it off craigslist

DFI P965-S Dark + C2D E6550 + 4gb DDR800 4-4-4-15 (OCZ) = $165

1 stick of the ram was bad, so have to RMA

I could at least POST at 500x7 without trouble, once I get an OS I'll see how stable that really is and maybe go beyond if I can.

Weeks later this hardware is nothing but trouble >_>
I got my ram back from RMA and I think they sent me the same set back, because I still have 1 bad stick giving me memory errors. Thats 1 demerit on OCZ since they said they sent me a brand new set
/offtopic
This C2D is cooled by a e6600 stock HSF (copper core). It idles low, 25-29c depending on room temp, but the second I start prime my temps rocket. 1s after starting my temps have already risen 10c. After 3s I'm in the 50s, and after about 1min I'm near 60c. This is crazy considering its a C2D and its not overclocked atm (2.33ghz), anybody else have this experience? Most Intel mobos come with a plastic standoff that you put on the backside of the mobo behind the processor, this I thought was mainly to prevent shorts. I'm wondering now though if they're really meant to keep the motherboard from bending as much as they do with Intel HSF attached (I have some serious bendage going on with my mobo). I'm also wondering if that bending of the motherboard is causing poor contact for my hsf and processor

When I took the HSF off a week ago to test a different processor, I found that only a small sliver of the processor's arctic silver was actually disturbed/used. The vast majority of the grease still had the same spread that I originally gave it. Does anybody know if putting that standoff on the backside of the motherboard helps correct the bending of the motherboard, and in turn creates better contact with stock HSFs?