How much air do I need to move?

sgtspiff · Post by **sgtspiff** » Fri Sep 29, 2006 2:20 am

I've finally come to a conclusion on wich system I will buy. But one final question still haunts me.

How much air must I move to keep it quiet? This will be the system;

CPU: Intel Core 2 Duo E6400
Chassi: Silverstone Temjin TJ08B
PSU: Seasonic S12 430W
HDD: Western Digital Raptor WD740ADFD
Mobo: Asus P5LD2-VM SE
RAM: Kingston 2x1GB PC4200
Graphics: XFX Geforce 7950 GT 512MB Passive
CD/DVD: NEC AD-5170 DVD+RW DL 16x Svart
CPU cooler: Scythe Ninja Plus (run passive)
HDD cooler: Scythe Quiet Drive
Damping: Acoustifeet och Acoustipack soft fan mount
Screen: Hyundai N91W

The TJ08B has two 120mm fans 1 in 1 out. And a vent by the VGA card. I will also open up a couple of PCI-slots.

My first thought was these;
Scythe S-Flex 8,7dBA.
But they only move 33,5CFM of air and that might be to little?

How much air do I need to move? Will 33,5CFM be enough?

jaganath · Post by **jaganath** » Fri Sep 29, 2006 8:12 am

How much air do I need to move? Will 33,5CFM be enough?

The only way to know is to build it and see the temps. IMO your main cooling problem will be the GPU. Also, can the Scythe QuietDrive handle a Raptor?

Felger Carbon · Post by **Felger Carbon** » Fri Sep 29, 2006 8:16 am

sgtspiff wrote:How much air must I move to keep it quiet?

Scythe S-Flex 8,7dBA.
But they only move 33,5CFM of air and that might be to little?

How much air do I need to move? Will 33,5CFM be enough?

First, the correct question is, "How much air do I need to move to keep my X watt PC cooled to Y degrees C?"

Second, note that the fan CFM specs are for zero back-pressure, or air-flow resistance. In a real PC, those CFM figures are never achieved (consider them a "guaranteed not to exceed" figure).

If you are at or near sea level, and you know X and Y (the power dissipated by your PC and the exhaust temperature rise you can live with) then the actual CFM you need is:

Actual CFM = 1.738W/C where W = power dissipation and C = degrees C rise in exhaust temperature over inlet temperature.

Now all your CFM problems have been solved by the above equation, you can happily go on your way.

You'd never guess I used to be an engineer, would you?

Post by **NeilBlanchard** » Fri Sep 29, 2006 8:45 am

Hello,

The short answer is: it has to move enough air to keep it cool. And as has been said, the only way to really know this, is to do it.

On the face of it, it looks very possible. Your CPU is relatively cool running, and with two 120mm fans moving air (plus the PS, so really it's three 120mm fans), you'll be fine; IF the case does not have too restrictive air flow, and if you have a good enough heatsink on the various heat producing items.

What is your HS on the CPU? Are the video cards active or passively cooled? Ditto for the mobo?

oakdad · Post by **oakdad** » Fri Sep 29, 2006 9:38 am

So you are saying a computer requiring lets say 200 watts (~1.8 amps at ~110volts) from your power outlet coming out of the wall in your house.
Allowing only a modist 10 degree C temp increase. We would have the following.

CFM = 200 * 1.738 / 10
So CFM = ~35 CFM

And a 33.5CFM fan in the front and a 33.5CFM fan in the back does would give you a total of 67CFM. So you could be nearly 50% restricted and still be ok?

Does this seem out of wack or have I always over done my CFM requirements in the past?

jaganath · Post by **jaganath** » Fri Sep 29, 2006 9:42 am

Hi FC, how is that CFM equation derived? Temp and CFM are inversely correlated, but I wonder if the relationship is absolutely linear (ie for a decrease X in CFM you get an increase Y in temp).

qviri · Post by **qviri** » Fri Sep 29, 2006 9:47 am

Also take into account the fact that air is a fluid...

If there are restrictions along the way, such as heatsinks with densely-spaced fins, air will try to mostly flow beside them. If the airflow path is front bottom to back top, the video card will be mostly out of the flow, and hot air may stagnate underneath it.

Also, airflow addition doesn't work like this. If you had two exhaust fans, you could add the CFM values; but one intake and one exhaust fan will reduce the effect of restrictions in the airflow path, but the airflow will stay at 33.5 CFM. It's the opposite of batteries in series or parallel.

peteamer · Post by **peteamer** » Fri Sep 29, 2006 9:50 am

oakdad wrote:And a 33.5CFM fan in the front and a 33.5CFM fan in the back does would give you a total of 67CFM.

Unfortunately that's very much not the case...

Bugger qviri got there before me...

Felger Carbon · Post by **Felger Carbon** » Fri Sep 29, 2006 10:21 am

jaganath wrote:Hi FC, how is that CFM equation derived? Temp and CFM are inversely correlated, but I wonder if the relationship is absolutely linear (ie for a decrease X in CFM you get an increase Y in temp).

To the best of my knowledge, the equation is absolutely linear.

Derivation? You have to determine the weight of the air per cubic foot, and how much the temperature of that one cubic foot goes up for a given power input. I derived this twice, a month apart, using a thick book called "The Physics Handbook", as I recall. Regrettably, these calculations reference BTUs - British Thermal Units - and so you get to research the conversion constant for joules to BTUs. The reason I did it twice, a month apart, was to guard against dumb mistakes.

(It was Brit scientists who did the original research, so all the measurements are in furlongs per fortnight.)

Alas, the equation is based on sea level air density, and I have since moved to a location that's at 4100 feet elevation. Sigh.

peteamer · Post by **peteamer** » Fri Sep 29, 2006 11:09 am

Felger Carbon wrote:(It was Brit scientists who did the original research, so all the measurements are in furlongs per fortnight.)

I was always taught it was 'Chains per Equinox' which originally came from/during the Roman occupation of Britain...

.....

Felger Carbon · Post by **Felger Carbon** » Fri Sep 29, 2006 11:33 am

peteamer wrote:I was always taught it was 'Chains per Equinox' which originally came from/during the Roman occupation of Britain.

At the time the original experiments and measurements of air properties were performed, I believe the Romans had cut and run.

I decided to try using the internet to derive the constant. I very quickly arrived at 1.740 for dry air, so my 1.738 must have allowed for some moisture. I stayed away from Wikipedia - I don't trust it (it now sez there was one Wikipedia founder!).

First I got the air density of one cubic foot of air. Each second, 1 CFM = an airflow of 1/60 of a cubic foot, so divide by 60 to get air in grams per second. Then I got the "heat capacity of air":

cp=1.006kJ/kg/degC or 1.006J/g/C and recognize that a Joule per second is one Watt. So divide 1.006 by the weight per second in grams for 1 CFM, and voila, the constant - at a lot less work than I originally did.

Tzupy · Post by **Tzupy** » Sat Sep 30, 2006 6:53 am

IMHO you should get a fan controller or use the motherboard temperature based fan control if available.
At low load your system should be well cooled by 120 mm fans at 700 rpm, but at load - like in modern 3D games -
the GPU may get hot and you would need to increase the rpm, maybe upto 1,000 rpm.

peteamer · Post by **peteamer** » Sat Sep 30, 2006 12:45 pm

sgtspiff, sorry to pollute your thread yet again

but I really feel I must set the record straight :.....

Felger Carbon wrote:At the time the original experiments and measurements of air properties were performed, I believe the Romans had cut and run.

Actually this is not the case at all, not even close...

Those experiments had been done and the conclusions found a long long time before the Romans were ever in Britain.

It had actually been worked out way before even 0001B.C. and was common (but never ever spoken about ) knowledge.
Unfortunately... much much later, some Klutz with too much Meed in him and a loose tongue (because of said Meed ) let the information slip....
At this point a cover-up was instigated, ( much like when it was found that Bush had a brain cell that wasn't dead... ) and all was made to seem like it was a new finding

.....

---------------------------------------------------------------------------------------------------

As for the Romans "Cutting and Running"... well that's just another 'Crock of Left EarLobes'...

They were actually here as our guests..... But only for our own very British style of humour/amusement.....
Once we had gotten bored with them.. We drove them out so we could carry on with our own inimitable life style.....

There has even been a very factually based film made about it...
But as some of the information is not for 'World Ears', the British were 'transformed' into mice for the film so as to negate having to rewrite the history books yet again...

You may have seen the film... It's called "The Hitch Hikers Guide to the galaxy"

Regards and sorry to have bored people...
Pete

Please don't believe all you read in the WIKI.......

sgtspiff · Post by **sgtspiff** » Mon Oct 02, 2006 11:38 pm

Okidoki.
So my question was maybe a bit vague and was specific for the system I'm buying.

But if you see it like this;

Could you go after size of the chassi (cubic feet) and the total airflow of the fans to calculate CFM needed?
Most of computers have the same stuff inside, Mobo, VGA, CPU, HDD...
And most of the stuff a pretty similar in heat (yeah I know that some are warmer but in the end it's not that much of a difference when you use Ninjas in all systems).

Can you make a simple equation on CFM needed (even if yoy use the manufactorers "fantasy" numbers)???

Tephras · Post by **Tephras** » Tue Oct 03, 2006 12:38 am

You could always look around in the general gallery for systems that is not too far off from yours and see what fans is used, if it's mentioned. On the other hand, I think the best is to do what Tzupy wrote above because that is a flexible solution and will give you some headroom.

How much air do I need to move?

How much air do I need to move?

Re: How much air do I need to move?