Cooling myths
Some of you might already have seen this article but here it comes.
http://www.procooling.com/articles/html/5_myths_about_cooling__-_phaes.php
Since most of you is probably more experienced in cooling than i am it would be nice to hear what you think.
http://www.procooling.com/articles/html/5_myths_about_cooling__-_phaes.php
Since most of you is probably more experienced in cooling than i am it would be nice to hear what you think.
0
Comments
I agree that it is total BS that aluminum PC cases act as giant heatsinks. Given identical PC cases where one was made of aluminum and one of steel both would, for all intensive purposes, act the same for cooling efficiency. Remember that it is the PC case configuration + thickness of material + cooling fans that make the difference.
To my best knowledge copper is a better conductor of energy (thermal and/or electrical) therefore given identical construction then copper will perform better. The statement of one material absorbing better and the other "giving up" better is erroneous.
Northbridge's need (water) cooling. Many manufacturer's have told me that the northbridge active cooling is bunk. Like a fancy paint job on a car...the car "looks faster" but isn't necessarily so. There will be many that argue this point stating their overclocks are superior due to increase active cooling on the northbridge.
I cannot comment to the program Radiate.
Positive case pressure better? HA! Air will find it's way into a PC case. Get that hot air out. Plain and simple. Best route...repeat after me. In through the bottom front....out through the top rear.
just my 2c.
There's a common thread to all these myths: people take their results, and then try to come up with a reason for them, when their test was invalid in the first place.
Most power supply reviews are totally worthless. The data is totally invalid.
Any heatsink review using an overclocked CPU as a test load is worthless (as I confirmed recently while testing that blower assembly
The logic behind an aluminum case acting as a heatsink is so obviously flawed that I don't know how it got started. You're talking about transferring energy from the cpu, to the heatsink, to the air, to the case. It just ain't going to happen.
That whole copper vs. aluminum thing is rediculous. People need to go read the CRC Handbook Of Chemistry & Physics before they say stuff like that. Copper has a higher thermal conductivity than aluminum does. It absorbs AND radiates heat better. PERIOD. A possible reason why copper heatsinks may not perform as well in low flow applications (which I haven't seen any evidence of, so I'm speculating here) is that copper is a much denser material. A copper heatsink of a given size will weigh a great deal more than an aluminum heatsink of the same size. The more mass an object has, the slower it is to react to changes, and copper heatsinks may not react as fast as aluminum ones do, even with their superior thermal conductivity. But I don't know that for sure. It seems logical, but I'm sure that the 5 myths in this article seemed logical to the people writing them, too, and I haven't tested my idea, so at this point, it's just a guess.
1. I don't know much about cooling but it doesn't sound ok to me.
2. No opinion
3. I'm kinda hoping they don't since i'm having trouble finding a place that sells swiftechs nb-blocks in sweden. Other than that i'm awaiting your opinions.
4. Never heard of it before i read the article and have no use for it nor an opinion.
5. I'm not into cooling so no opinion there either.
I can agree that having a lightweight case would be a good idea but bying aluminium thinking it will improve temps if it obvisly doesn't is a mistake. Please keep'em coming.
It's beginning to sound like I wrote the article, you know that? (I didn't). The guy is right. I've done my own testing, too. Negative pressure works very, very, VERY well if the case is properly designed. Neutral pressure is the best idea for most users because most cases are not designed for negative pressure use. Positive pressure just plain doesn't work.
New market: Negative Pressure kits.
2. Don't know. Hardly use either.
3. NB needs active cooling of some sort if overclocked, yes. At least on the last 3 Amd sollutions. (2XVia1XNvidia) Anything else is just BS. I have tested passive NB and overclocked. Result? Froze.
4. No idea
5. No
Another problem that needs to be adressed is that 9 out of 10 relies on the MBM application. If that was true readings, my pc would never boot. The vcore and 5V flucts way to much in MBM while it's dead stable while measuring voltages with a multimeter.
This does not mean that Aluminum cases or steel cases are bad. This is not a question about design. It is about thermal conductivity.
Overclocking changes the parameters about northbridge cooling. Given that a certain flow rate exists with watercooling through the waterblock to a CPU heatsink THEN if a second waterblock were added in the same loop then the overall effectiveness could be lessened.
The article argued that the flow per minute would be lessened by adding a second waterblock in the loop. (the northbridge). I'm not positive that a second waterblock would greatly affect the overall gallons/litres per minute flow rate. I say "greatly" affect. BUT it would be important to put the northbridge waterblock as the second waterblock instead of the first. The coolest water would be passed through the CPU waterblock first where it is most needed. The thermal energy picked up by the water in that waterblock would warm the water but that warmed water would most likely still do a better job at "cooling" the northbridge than an active fan.
Therefore...a better overclock.
But in non-oc'd conditions it has very little benefit as evidenced by the amount of motherboards with passive northbridge heatsinks.
In heatsink design the goal is to provide a large amount of radiating surface area. The material chosen (copper vs. aluminum) depends on a retail cost that a consumer would pay vs. manufacturing cost. If cost didn't matter we'd all have heatsinks made of Space Shuttle ceramic heatshield material.
Copper heatsinks are more costly to manufacturer and you know us enthusiasts...we always want the best for cheap.