Innovatek? the best? cool . . .

[aspire.comptech]

I ain't drinking no pond water!

[coolfris]

It's not worth trying since he will eventually attack the review for the flawed theory.

The more I read, the more I consider him as a Innovatek shill...

He's not Innovatek all the way. Here he actually advises a thermochill radiator and a D5 pump. After that he explains how to get low flow with the D5.

As for rads, I suggest, you go for Thermochill. It is excellent, and cheap. Regarding the pump, go for Danger Den D5 12V Fixed Speed Pump. This pump is tiny but strong; size is no guarantee for power.
Using this, you can split the flow from 1/2" into 2 hoses of 1/4" to get low flow, and supply it to 2 loops using the same pump, and get superb temps. You save cost of one pump, and get better temps with lower flow. That's how I'm doing for years, and it's been absolutely great.

[smee]

Wow.

[aspire.comptech]

Since when is thermochill cheap?

Last I checked they were near enough 3x the price of a traditional rad...

Also since when is the d5 a tiny pump? Last time I looked at mine it was about 3x the size of a ddc...

Maybe it shrinks when Im not looking?

[GeorgeStorm]

Hi,
Ive just read through the thread, and i have to say it is ridiculous, for while his formulae may work, he isnt taking into consiseration that while for every 'loop' of water the low flow is absorbing more heat (even though there is a limit due to water's own specific heat), the high flow will loop many more times in the sam length of time, thereby reducing his arguement alsmot completely.
I am no thermodynamics expert, but it does seem to me that he is fundamentally wrong. His idea of rolling a stone down a hill will gather no moss isnt really applicable, as the stone would gather more and more moss the slower it went, whereas water would not absorb more and more heat the slower it went, and so while it may sound true, it is not 100% useful/applicable.
Just me trying to make some sense of quite a lot of maths/physics
If anyone disagrees with me, or would like to clarify things to me, please do, as the entire subject of thermodynamics etc fascinates me

Thanks

[smee]

Maybe it shrinks when Im not looking?

It shrinks when i'm not looking too....

@George, thats a very good way of putting it. I agree 100%

[Shlomo]

I think this just needs to be posted real quick so DB can have a SOLID answer instead of being mocked.

Heat travels the fastest from one location to another with a higher heat gradient, with restrictions on thermal conductivity of a block.

http://en.wikipedia.org/wiki/Thermal_conductivity

The equation listed (number 2 on the page) shows that the higher the gradient, the better the thermal conductivity

If I have a copper block at 100°C and I run a LOT of water over it, it keeps the thermal gradient high (so 25°C water keeps flowing over it) so heat transfers more efficiently and faster. In the case of thermal equilibrium, that hurts. It'll take longer for the block and water to reach thermal equilibrium than to get the block LOWER than thermal equilibrium by running water over it instead of sitting the block IN the water.

[GeorgeStorm]

Schlomo
Yes, so the speed of the heat transfer decreases as the difference between the temps decreases, therefore high flow rate would be more efficient, as the difference between the water temp and the CPU/waterblock temp would be greater, i hadnt thought of that, the reasons keep on stacking up!
And smee, thankyou, as a 16 year old, i think we are standing up for teenagers here, and of cause their vast knowledge of computing/water cooling/ and thermodynamics

[NaeKuh]

we told him all this, grim even laid the smackdown on him with that awefully confusing stuff i havent seen since i was in college.

He just wont accept it. Theres nothing we can do except make fun of his responses.

See look at his reponse:

"You should've first read my post properly. I said it was a formula to calculate the water temperature and not something to calculate cooling efficiency. So you needn't have wasted so much of time wrongly assuming that I've understood wrong. I only wanted to bring out that cooling is better because there is greater heat dissipation per flow cycle.

I didn't put in the efficiency part that is a direct consequence of entropy becasue it'll make the post too technical for others to understand. I didn't want to unnecessarily lengthen because I didn't know that you've sufficient knowledge in physics to understand everything and thought it was best to explain things from a layman's aspect.

Anyway, with regards to efficiency, the same copper is used in both Innovatek and DD and all other companies. If we neglect crystal abberations like Interstitial, Frenkel, Schottky, etc.. from our comparison, we'll find that transfer of heat from CPU to water-block by the base-plate is same because in all cases it is identical copper. You've advertised that you're removing terms that are identical from L.H.S and R.H.S; but I didn't say so because I guessed you could make that out.

Again if you've read my post, you'd have understood that I'm very well acquainted with these equations as these form the basics of heat transfer and refrigeration. I also know for a fact that there is a big flaw in your beautiful diamond. You've forgotten that heat is freshly added for every cycle and you never added that factor called rate of heat production.

So your equations show the cooling efficiency for bodies with static heat content and not for bodies that are continuously getting hot. You might've also missed that saying of mine "high-flow is good only for cooling bodies that don't get continuously hot" because it was a lengthy post. So there is nothing surprising that the formula you invoked gave you the result saying that high-flow was better.

If you'd thought in my lines, you're equation should've never been dependent on flow-rate. I'm not saying that the equation Q = ms∆t gives you cooling efficiency because I never intended to show cooling efficiency in that post.

And don't think I'm an idiot to type [T1 - T2]; it's not because I don't know. For a fact, I've got a patent pending invention on refrigeration and you won't find people without knowledge in basic Physics pantenting refrigeration devices. It is simply because I wasn't able to get ∆ and the superscripts copied from Word to this post. But I find I'm wasting my time. You can only take a horse to the pond; you can't make it drink...."

[aspire.comptech]

I love how he constantly brings up his refrigeration thingymabob but doesn't say anything about what it does.

I have a patent pending refrigeration device, its called an ice cube.

[Shlomo]

I've thought of this while reminiscing however, and I think it probably should be brought up for the idea.. even if it may be flawed

In the radiator, the slower the water moves, the more heat is removed from it... it is basically a fact. 50°C water will lose more heat going half as fast as the same water would going a normal flow speed.

So I got to thinking a while ago, what if we split the radiator from the cooling loop? How would you go about doing that? The same way our heart splits our blood flow. 2 sections to lungs because of the many capillaries, it needs a separate "pump" to keep blood flow higher in the body. So how can we simulate this in a watercooling loop? Reservoirs. Reservoirs simply pool water and basically stop flow unless they are totally full. If totally full they do nothing and flow continues. Almost like our Atria

So heres the idea. Res > D5 highflow pump > waterblocks > Res to stop flow > lowflow pump > radiator(s) > starting res

Debunk this idea... I think it might improve temps ever so slightly, even by .05°C just because the water is cooled .2°C more. No math behind this, purely speculation

The only thing I can come up with is that the waterlevels in the res to the lowflow pump might overflow because water is being fed into it faster than the lowflow pump can feed into highflow pump reservoir. Probably the nail in the coffin, but i'd love to test it one day

[NaeKuh]

I have a patent pending refrigeration device, its called an ice cube.

oh wow! i developed the ice despenser!

i didnt know i was stealing your idea!

[aspire.comptech]

I've thought of this while reminiscing however, and I think it probably should be brought up for the idea.. even if it may be flawed

In the radiator, the slower the water moves, the more heat is removed from it... it is basically a fact. 50°C water will lose more heat going half as fast as the same water would going a normal flow speed.

So I got to thinking a while ago, what if we split the radiator from the cooling loop? How would you go about doing that? The same way our heart splits our blood flow. 2 sections to lungs because of the many capillaries, it needs a separate "pump" to keep blood flow higher in the body. So how can we simulate this in a watercooling loop? Reservoirs. Reservoirs simply pool water and basically stop flow unless they are totally full. If totally full they do nothing and flow continues. Almost like our Atria

So heres the idea. Res > D5 highflow pump > waterblocks > Res to stop flow > lowflow pump > radiator(s) > starting res

Debunk this idea... I think it might improve temps ever so slightly, even by .05°C just because the water is cooled .2°C more. No math behind this, purely speculation

Your forgetting that the whole loop is under pressure.

As long as the reservoir is not open it won't impact flow.

If your drawing from an open tank the flow will be impacted by the restriction that comes after it.

I'm probly thinking about this totally wrong but as far as I'm aware, flow throughout the loop is constant.

[Shlomo]

Your forgetting that the whole loop is under pressure.

As long as the reservoir is not open it won't impact flow.

If your drawing from an open tank the flow will be impacted by the restriction that comes after it.

I'm probly thinking about this totally wrong but as far as I'm aware, flow throughout the loop is constant.

Yeah I was thinking open tank, I knew about the closed loop pressure deal

[NaeKuh]

I'm probly thinking about this totally wrong but as far as I'm aware, flow throughout the loop is constant.

Flow is constant thoughout the whole loop in a closed loop system.

Pressure isnt however as it changes each time you hit a restriction, ie blocks and angles.

[IanY]

You can't make the horse drink, but you can beat it to death, and then continue beating it till you are all out of sticks.

[NaeKuh]

You can't make the horse drink, but you can beat it to death, and then continue beating it till you are all out of sticks.

[LinusTech]

In the radiator, the slower the water moves, the more heat is removed from it... it is basically a fact. 50°C water will lose more heat going half as fast as the same water would going a normal flow speed.

While it is true that slow moving water through a radiator will have more heat removed from it per pass, we all need to remember that this is a closed loop, NOT some magical water cooling system with unlimited warm water that needs to go through a radiator to cool the CPU block... or something.

Let's say you double the flow rates of your slow moving loop. Now you are removing less heat from the water per pass, but you are also doing twice as many passes through the radiator.

You are still removing more heat from the system by having a faster flow rate and you have the added benefit of not hurting the performance of the other components of the loop.

think about this, slow the loop down to almost nothing. You've got cold water in your radiator and hot water in your CPU block. What bloody good did that do you?? Faster flow is better period. The water spends the same amount of time in the radiator and you get better thermal performance out of the radiator from more turbulence. The only consideration is heat dump from the pump.

[largon]

The thread at SLI forums yet again shows how even educated people can be totally clueless. Too bad.

[DarthBeavis]

SLI forums turn into . . .

[NaeKuh]

Sorry i lost it, and blew that post up with pictures.

I got tired of his bull on this base and that base is soldered. So i poped up high res closeups of some bases.

[Shlomo]

While it is true that slow moving water through a radiator will have more heat removed from it per pass, we all need to remember that this is a closed loop, NOT some magical water cooling system with unlimited warm water that needs to go through a radiator to cool the CPU block... or something.

Let's say you double the flow rates of your slow moving loop. Now you are removing less heat from the water per pass, but you are also doing twice as many passes through the radiator.

You are still removing more heat from the system by having a faster flow rate and you have the added benefit of not hurting the performance of the other components of the loop.

think about this, slow the loop down to almost nothing. You've got cold water in your radiator and hot water in your CPU block. What bloody good did that do you?? Faster flow is better period. The water spends the same amount of time in the radiator and you get better thermal performance out of the radiator from more turbulence. The only consideration is heat dump from the pump.

I went about it through an open loop... reread :-D Open reservoirs = open loop.

[tiro_uspsss]

Sorry i lost it, and blew that post up with pictures.

I got tired of his bull on this base and that base is soldered. So i poped up high res closeups of some bases.

good stuff, about time someone did

[AllAgainstPaul]

this idiot reminds me of that guy over at [H] who put a NB waterblock on top of a motherboard's heat sink fins and trying to defend it as being the next coming of Jesus.

No we can't make the horse drink the water, but we can sure as hell try. If he drowns, it's his own damn fault.

[LinusTech]

I went about it through an open loop... reread :-D Open reservoirs = open loop.

I'm using closed in more of an "electrical" sense. It's closed in that it recirculates. An open loop in this context would be one that flows from the kitchen tap and goes through your computer and down the kitchen sink. You're not reusing the same water.

Open reservoir is a whole other issue. It would just be a way for contaminants to enter your loop and for water to evaporate. You'd be using more additives and topping off your loop more often. It's not worth experimenting with and has no impact on any of this.