Quote:
|
Originally Posted by Cathar
Centrif. blowers have a different optimisation point. They also tend to be a lot noisier than axial fans for the same level of air-flow, and they are not exactly things that can be mounted conveniently inside a case.
Keeping in mind that we want sustained air-flow through the radiator that is at least around 43CFM. 43CFM of air-flow has an inherent C/W of 0.04, meaning that a 0.04 C/W is as good as you can get given perfect heat transfer from the radiator into the air flowing through it.
If you take a look around the web, you won't find many blowers that can deliver 43CFM against pressure for low noise, where by "low noise", we're talking about ~20dBA noise levels. If you don find one, chances are it's also a fairly large unit, quite costly, and being run at very low speeds.
Remember, we're also talking about cost here. Hard to compete against a ultra-low-noise, small, compact, and light $5 fan with a large, heavy, noisier $50 blower, even if you can argue that the blower is better for pure performance.
The whole point with making the radiator larger than the 12cm axial fan was to decrease the air-flow resistance to the point that the low-speed low-noise axial fan is operating very high up on its flow curve. i.e. you don't need the pressures provided by a blower to get good performance out of the radiator.
Sure, as Hoot say's, if the radiator is now 16x16cm then let's stick a 15 or 16cm fan on it. Yes - you could do that and get better performance, but what I'm talking about here is extracting the most performance possible for the least effort (on the fan's part) and cost and noise.
To me, it just feel right that I'm matching a super-quiet, $5, 1W powered fan to a radiator designed to work with its specific traits, that provides cooling performance that matches larger/noisier setups. Yes - of course we can stick a larger/noisier fan on it, but then that's moving away from what I perceive to be the primary design goal.
Match the fan/radiator setup with a super-efficient super-quiet 5-10W pump (my other dream - but 15-25W will have to do for now until that day) and we have a setup that kicks anything else in the teeth quite hard in every respect (less power, less noise, low-space, smaller ID tubin, better cooling, moderate cost).
Maybe I should move to Europe.  |
We both want silence
I know very little about fan, radiator, and watercooling design.. Im just suggesting that instead of low restriction rads, to achive the same with as-quiet fans, but much higher restriction, same noise, but also same airflow.
Where you say effecient, what do you mean? efficent in fan power? noise? airflow?
I think noise is most important, then air flow.. I couldnt really care how much power it takes within a resonably amount.. as long as the noise and performance are both good.
Would having less restriction, make the airflow less effective? much like a single large channel in a waterblock i less restrictive but compared to many smaller ones or jet impingement it performs a lot worse aswell.
Im currently using 2x200mm heatercores, with airflow in series, with a 172mm fan which is not THAT noisy. Its rated for 12w - 24v 0.55A (which is noisy!) but I have been running it at 7-10v (At a guess ~3-6w = 6-12v?) which is much much reduced niose - but the bearing used in it are crap.. with good bearings it would be almost silent - I think.
What causes noise in a fan?
Like a water pump, I think its not so much the actual motor but instead the bearings and the flow rate.
.'. keep flow the same, use the worlds best bearings, but chose a motor that is LOW RPM (dont want high pitch whine - I think below 1000RPM, maybe down to 500?!) also over-rated enough so it will require a LOT more restriction to slow it down than your going to get from a few rads in series. Might want a bit more power but are you going to hear any difference..
I think the ideal rad would be 180*180*100, much like my two rads together... *but* also with a higher FPI - how much will that affect performance assuming constant air flow?
I want the most airflow, for the least possible RPM, with the most possible pressure.
I think low rpm is important in reducing noise caused by bearings?
edit:
Done some quick tests and I have observed a few things which should be obvious...
1) air restriction does not slow down fan that much, instead the air bounces back through the fan
So all I can think of really, is a really wide fan, with blades curved so that they give the maximuim flow and pressure - allowing the RPM to be dropped a lot. ..
Anyway.. time for me to let this thread get back to where it was..