Case Air Flow made simple ??

[Terrorbyte]

Ok, now as we all try to acheive a "positive" air flow inside the case to cool our valuable PC there are some calculations that may help, at least make your brain work instead of the computer.
The majority of cases have an 80mm inlet fan and the same inside the p/s, + there are usually more air inlet holes in the case both at the bottom (L) side and at the back of the box.
Any p/s fan does not work at its full rating due to all the restrictins inside the p/s, say it works at around 40%, which means there must be at least one more 80mm outlet fan placed somewhere to be able to remove the hot air that comes from the cpu fan, and also from residual heat created from the p/s case
Stay with me now, even if there is no extra exhaust fan in the top or top rear, depending on the brand and where the p/s is situated in the case, you would need to drill at least 64-10mm holes which is equal to the size of an 80mm fan area.
Considering that there are usually a lot of pre made air holes in the case there should be at least the same or more in the top to remove the hot air, thereby acheiving positive air flow through and over all the right parts that need cooling, and this takes into account the p/s fan that is not working at full capacity.
Hard drives do get fairly hot, as do CD burners, so it would not be overkill to have two extra exhaust fans, as long as there is enough air that can enter the case and keep a constant flow of fresh air passing over the whole thing.
It's not much use jamming the air in if it cant get out and just as important to not have several exhaust fans and not enough inlet areas to keep the flow happening.
In simple terms one 80mm fan needs at least 64 - 10mm holes in the other end to get it out.
So get out the drill and go for it, first feel the case, if it's hot you have a problem or at least a hot spot in the case where the air is trapped, the Aopen HX45a case is a classic here as the space between the p/s and the top two CD drives will trap all the hot air from the m/b, cpu, and residual heat from the p/s itself, and it simply has no where to go but sit there and heat up the whole thing.
Don't be shy, just try to drill the holes neatly in straight lines and it will look as if it was made that way.

And thats that.

[Toaster]

Hello there,
While what you say has merit, you made a few assumptions that I'd like to clarify. Case venilation varies from case to case. Some are well ventilated and some its nearly non-existant.
Your calculations on fan efficiency are reasonable but this varies by 50% or more demendant on the P/S design and the fan used.
I always use the following rule of thumb:
1. Intake vent area should be double that of the exhaust fan(s) size.
2. Fans are 60% more efficient "blowing" than "sucking".
3. An intake fan "boosted" by 2 or more exhaust fans is more efficient and less noisey then one intake and one exhaust fan set-up.
4. I have seen some poorly designed cases where intake air vents DID NOT EXIST or were not cleared from the maker of the case.
5. Some cases have these vents in a posistion where the system simply sitting on a rug would block such vents.

note for #2:
To test this theory, connect a fan to power. With the fan running, move your hand closer and closer to the induction side of the fan. (Intake side)
As your hand moves closer, noise increases markedly and the amount of air it moves reduces. Now try the exhaust side of the fan again with your hand.
Noise remains fairly constant and so does its output.
In most cases, a frontal mounted fan has to pull air through a number of small holes. Assume the fan can move 5 cu/feet of air per minute in open air.
A fan mounted and drawing air through the drilled holes would have a drastically reduced capacity of about 2 cu/feet per min.
Now, assume you have 2 exhaust fans, one in the P/S and an additional 80mm fan.
If the system pulls most of its air through the frontal portion of your case, a frontal fan would be "boosted" and more efficient and thus quieter.
Ideally, a large open frontal area and 2 exhaust fans is preferred to a frontal mounted fan "forcing" cooling. Some folks have devices which run exceptionally hot. Be this a graphics board or other do-dad and a frontal fan may be useful.
Otherwise, a frontal fan "fights" mother nature in some instances.
If one could "steer" the incoming cooling air rather then "force" it, efficiency rises and noise drops. Having a well ventilated Pee-Cee does NOT mean it must be noisey as well. Actually, high noise is the sound of "resistance" to air flow.


There are tens upon tens of variables to consider and thus the rule of thumb I use takes worst case as assumed. In the end, noise is low and efficiencies high.

[Terrorbyte]

Thanks for the info Toaster, I understand fully the theory + practice of how the shape and design of a fan works, allbeit the exact figures can as you say vary greatly.

Being also a boating fanatic I know that the propellers on my 175hp engines work MUCH better probably about +90% better in forward than in reverse, anyone with a boat will understand this only too well, simply due to the shape and design of the prop.

However this does not apply to a fan/prop design that has a straight blade without the additional shape and curves for the ablilty to bite into and grab more air/water as the case may be.

I have stated previously that I always as a matter of automatic modification remove the pre cut inlet holes which are made on the computer case, put simply they are totally removed leaving one big hole the same size as the internal diameter of the fan that has to go there, this should be done by everyone (that cares about their p-c cooling). Then any of the plastic of the front panel that may restrict air flow gets drilled or removed to assure the total area of the (hole) is roughly a little bigger than the fan put there.

I have had a few fans on the bench just for the testing of some of these theories, (and to pick the best for the box) and some are simply better even though they are the same size in many ways, there are those that have a better designed fan blade and just work better.

My main reason for the thread was to show the comparison of just how many holes are needed just to match the 80mm fan, given that it (in theory is not running), just for the exercise. As while the fan is running all those calculations are altered and more holes may be required.

I can only draw from my many years as a mechanic, with engines fitted with superchargers, if you jamm the air in, then you have to be able to get it out faster to acheive maximum horsepower, I find similar theories inside the computer case with the air flow.

I agree that there are so many design variables of the case that could change all these basic calculations many times over, and I am equally sure that most people are not aware of how the fans work, as we both have mentioned, by their design RE; sucking v blowing, once these variables are known more widely the subject of case cooling will be easy to deal with.

Of course one could just take off the cover and direct a desk fan at it, that always works well, I have tried it and watched the temps drop drasticly, and remember that all that is moving over the components is the ambient air temp of your room, my new XP will run at 36 deg under heavy load this way, with the covers on and all the fans fitted the best I get is 38-9 deg, and that is a room temp of 26-7 deg.

An interesting subject.

[Toaster]

Howdy,
My intent was not to shoot holes in your post but to give more information so folks could decide the best solution for them.
On the fan and design area, most Pee-Cee fans are designed alike with little variance. If fan makers would put a bit more of a "twist" in the blades as well as more "cup" type features, fan efficiences would rise. (pitch)
Also a redesign of the shape of the fan itself (the venturi) would increase efficiencies markedly.
I'm afraid your description of the fan mechanics may be in error to some degree.
Alas, I may have simply percieved it as so.
At any rate, a fan creates a low and high pressure area. The low pressure side (induction) attempts to equalize by drawing in ambient air. A portion of this air is drawn from the exhaust side but most from the induction side.
What air is drawn from the exhaust side is an inefficiency in design.
This low pressure effect is partially nulled when the fan is in close proximity to a baffle (your fan mounting point). Turbulence is created and efficiencies fall and noise increases. Blower type fans are far more efficient as thier main element design differs from a simple propeller. This as you would surmise is why a "prop" boat and a "jet" boat act so differntly.

[Terrorbyte]

Don't worry, no intent taken, any disscussion on the subject is a usefull thing, it may take many cooks to write the book, but only one to cook the "computer case" so to speak, lol.
perceived ! perhaps the wording may be different ? it's all under control.

You might find that a jet boat has indeed a propellor inside, sometimes more than one, only difference is the amount of water taken in via the intake that is far greater than the propellor (alias the jet) can handle, the water is then pushed through a venturi principle into a smaller diameter outlet creating far greater pressure, and there lies the jet principle. All too simple eh ?

Where I live there are many large fishing trawlers, they will sometimes use a simple venturi placed to surround the propellor, slightly larger at the front of the venturi, this does give up to 20% more power for pulling larger nets, seems the jet principle is an old one.

I have taken notice of the Global Win FOP-38 fan on my cpu and that the blades, while only small in length, they are many, and are indeed curved and twisted, and also running at 6,800 rpm, making it a very noisy bugger.

A common problem I have found is the fan inside the power supply is just not up to the job, replacing it with a better one is not a hard job, a soldering iron and some common sense is a good thing to have and all should go smoothly getting that hot air out of the box.

[Toaster]

Howdy,
Care and common sense is manditory when working with power supplies.
While true that P/S fans are often minimal and easily improved, the close proximity of the fan intake to the internals of the P/S come into play.
This goes directly with the "hand in the fan" experiment.
Its not odd to notice that the majority of P/S failures are heat/fan related.
Iv'e noticed on many occaision that if the fan is very close to an internal part of the P/S that it will die an early death due to inefficiencies in cooling.
While more blades create more airflow, there is a finite limit.
Variables such as fan speed and diameter come into play here.
I set up on my last Pee-Cee a simple CPU heatsink utilizing the cooling of the case alone. This resulted in a VERY quiet system where a 20X CDROM was louder then the system as a whole. The CPU is a PIII-800 running at 1250 and CPU temps stay below 130 degrees (F) and often lower.
Granted, its running Linux and as such it doesn't have to work as hard "just to make it pretty"
Noise and cooling neddn't go hand in hand like some folks want to believe.
I work in a noisey environment and after such work I needn't my system(s) to also add to a noisey day.
I'm currently in a design stage of a fanless system thats internally insulated.
The end result is to only know that the system is on by the status LEDs.

[Terrorbyte]

G'day,
I don't suppose you wish to expand on this fanless system, I am always interested in new stuff, or is a patent intended for the project at hand making it somewhat of a secret ?
_____________________________________

With all this talk of fans, I got motivated and have spent the afternoon doing just that, changing fans and modifying cases.

Now with 4 of the same size, 80mm, to pick from I thought it would be an idea to do some more technical tests on their performance, so out came the multimeter to check the label to see if they were correct, all were much per the specs, but one of the fans simply pushed more air, I would guess about 50% more in fact so as they were all the same size, same blades, same pitch and curve, same current usage 0.12A, but different brands.

The obvious was somewhat hidden until I weighed them with 3 being about 80gms, the better fan weighed only 70gms or so, I would think you have guessed it by now, the better fan had much thinner blades (although identical in shape etc) making it simply run with less resistance and obviously much faster

BTW, the good one came from a 230w Aopen p/s

Which leads to an obvious question, is there a "best brand" p/s that you would know of, or would the majority of good name brands be as good as each other, Or on the other hand it may be just good sense to replace the p/s fans as an automatic thing with a new system with a top class high performance fan, solving many case cooling problems in one go. I know that it is not value to buy a cheap generic p/s for say $10, you usually get what you pay for, not always, but mostly.

It could be like buying a new car, "and as a retired mechanic I do know a little here" put simply if you are in the market for a new car and are not too fussed with it's colour (within reason) you WILL find of say ten vehicles that you test drive on the same lot, one WILL go better than the others.

Could buying computer components become a similar task ? it may well be.

All good fun anyway I reckon.

And the better fan has dropped the temps one deg on the cpu and one on the board....so it does matter......

[Toaster]

H`dy,
I am a "fan" (pun intended) of Aztec power supplies. Built like a tank and grossly under-rated. There are quite a few good supplies and a few loosers. Some of the loosers are downright dangerous IMHO. Those are "Deer" and "austin", avoid these like the plauge.
PC-Power and cooling is another fine supply. Delta power supplies are also OK but their design is beginning to faulter.
Take a close look at "panaflow" (panasonic) fans. These are top quality, have a good basic design and good bearing supports. Their venturi are designed on a "performance" format. Their better fans have deeper venturi and their fans are often "cupped". They move a great deal of air and are reasonably quiet.
Sunon fans are quite good too.


The fanless design of a PC isn't new. It failed because of the lack of cooling medium. I have a rough build of a fanless system but it was a major pain.
I removed the slot-1 CPU socket and reverse mounted it so the CPU heatsink side comes directly in contact with the case. The case is actually a tray made of .16" aluminum. (Al-you-mini-um to you brits )
The power supply was another monster and is not yet complete. It will be a module affair that connects to the system board "tray" and share its thermal capacity. So far, temps stay low but the design is not yet complete.

[Terrorbyte]

Just one thing which seems to be a common mistake, we are "AUSTRALIAN", down here in our bit of paradise, NOT British.
The pomms are a totally different breed entirely ! apart from that I personally got used to and still use mph, inches, feet, gallons etc, all the metric crap that they forced on us (of my vintage) makes no sense whatsoever, My 4x4 landcruiser drives in mph (in my head anyway)

the only thing I dont understand is why you blokes drive on the wrong side of the road, that would be hard to get used to.
________________________________________________________

At least an AMD or an Intel are still the same no matter where we live, and we all have to pay Bill (Mr $$$$) Gates to have an o/s.

As our bushranger Edward (ned) Kelly said just before they hanged him,

"Such is Life"

[Toaster]

Well, from the folks I talk with from Australia, everything derived from "mother England". So, no, you are not a "Brit". As a matter of fact, I always regarded Australians more of a "down home" American of sort. Both you Australians and we Americans have things in common, we keep the "Brits" at arms length but stay friendly at the same time. It never ceases to Amaze me that Australia and England come to the USA'a aid when the chips are down. (most of the time)
So calling you a "Brit" albeit mistaken, was not intended as an insult.
Please excuse this kitchen appliance while he takes his foot outta his mouth.
(Egads, athletes tounge!)

[Hpro]

Here it's about cooling I see - Nice, have some words to say to this - NO not argueing with neither of you just out of experience..

Had a nice case even I never liked it too much - had an overclocked computer board in it - plus
1. Front fan regular size of a PSU fan 80mm in diameter
2. Rear fan same size
3. High speed CPU FAN 60 mm with aerodynamic Blades - rotation speed ~ 5500Rpm blowing towards the CPU (sounds kind like an airplaine only with less volume in sound strength..
4. A fan mounted on the right side of the Computer just besides the Motherboard where the CPU (socket 8 ) was mounted diameter 80mm exhaust - no way to lcose the Right side anymore
5.and of course the PSU fan

The computer had always temperaturs of over 55 degrees as of overclocking - had to remove both side covers because otherwise it went too hot - and also the comp was and still is always on - and I do not use software Measuring the Temp but a inserted Hard ware probe proven to be within +- .5degrees CEL of faulty reading - so Temperature read is accurate -

Because of some other reason I exchanged the case - this one is a Pro Station 2002R case - now in this case there is only ONE supporting fan sides are closed the same CPU fan and the PSU fan - and temperatur is down to 46 - 50 (when it get's hot here )degrees still measured with the same probe inserted at the same place..

The this case is somewhat smaller when it comes to width but somewhat higher - the only other difference of it is that the power supply is above the Motherboard and not blocking the AIRFLOW from and to CPU FAN -

So, conclusion on my side is you can drill as many holes you like if your CPU is blocked in front with the power supply then get another CASE - no hole and no other fan will do better..

Today it's hot again - almost 36 Degrees surrounding temp - now the Case or Computer temp is on 42.3 degrees and CPU temp is on 48.4 ..

I think good AIRFLOW in the case is the answer to the question here and not 4 or more fans - only will get dust into your computer ...
Thanks for reading..

[Terrorbyte]

So true, that's where all the variables come into play.
Some boxes are good to start with, some are Ok just needing a little help, and others are just dogs.
Of course positive air induction and hot air removal is not an option, it's a requirement.
However good air flow depends on where things are placed inside the box and also where fans and their air flow is directed to.
It is a subject that has to be made on a case by case plan for both correct air flow and the right amount of it.