Guide to fan noise and case thermodynamics

[DerwenArtos12]

Yeah, that was me. There are two options, you can just fold the cable to get it compacted as best as possible or, go the razor method like I do. Lay teh cable out as flat as you possibly can then slide a razer, using as little force as humanly possible between the wire pairs. remeber there are 80 wires making 40 pairs so there is no real reason to seperate each and every wire.

The easiest way is to buy rounded cables and cut off the bulky plastic sleeving, most of the mid prices ones use individual twisted wire pairs while the cheapies are usually just machine folded, either way, chop off the big plastic sleeving and they'll tuck much nicer and be more flexible for routing.

I have the bad habbit of modifying everything though, every wire in my case has been cut to the exact proper length and routed as cleanly as possible and then some.

 

[cloudkat]

Originally posted by: DerwenArtos12

The easiest way is to buy rounded cables and cut off the bulky plastic sleeving

LOL.... I thought I was the only one who did that.
Back a while, a local PC shop was going out of business and I bought up all his cables. So I had a monster stock of rounded IDEs. So I figured this one out quick. I ended up with like 50 brand new ones for only about $0.75 each.

As for SATA, I was never really happy with them because the cables are too short to give you any really good options for great wiring. But when I bought out that PC store's stock, He had like 4 or 5 SATA cables that were 3' long. Now those bad boys you can do almost ANYTHING with. So now all I buy is 3' SATA cables. They are much better than the short ones that come with a drive.

I should probably start another thread about this, but on this subject, do any of you know where I can buy a roll of flat IDE cable?? I have found a place to buy the ends and the crimp tool, but can not find a place to buy the cable at. I would like to be able to custom make my cables to exact lengths. Especially with some of the case mods that I have lined up.

 

[DerwenArtos12]

I usually just buy long ones and cut them down, don't know of anywhere to buy 80wire. you could always just pick up some ~28guage and line it up yourself.

 

[bunnyfubbles]

Originally posted by: cloudkat

Originally posted by: bunnyfubbles
I'll keep my optical drive in the top most slot. Have been doing it since forever and have yet to have a drive fail before I've replaced it solely as a performance and/or technology upgrade.

I can sympathize with some of what you're trying to share, but come on...'if it ain't broke, don't fix it' - otherwise you're essentially fear mongering.

That is the cool thing about it being YOUR PC. You can do whatever you wish with it.
I never said that by having your optical at the top it would cause it to fail. All I was pointing out was that it gets hotter in that position and it blocks up the "hot air zone". So by moving it you can not only open up a larger, higher-in-your-PC-case "hot air zone" to be exhausted, but you can also help lower the operating temp of your optical. Producing less heat is a good thing, right?

I disagree with your statement, "if it ain't broke, don't fix it' - otherwise you're essentially fear mongering." And I am sure that most of the other modders and OCers out there will as well.
If your statement is true, then why do people OC? Lap their HSs? Spend all that money on aftermarket HSF? And a million other things like those?
It is simple, to get every last friggin' OUNCE of performance we can squeeze out of our systems. Each and every little thing we can do to gain better temps, or better performance we are willing to try. So, it is with that in mind that I do not put my opticals up top. I use that as the "hot zone" and have exhaust fans there.
You, as I said, may do as you wish.

You can go all out, but its not going to directly scale in added performance, if at all. If anything the only value of doing so is because its fun to do. The way you initially worded it, you make it sound as if it's a no-brainer solution, yet you only provide theory and do not back it up with any cooling performance or system performance numbers that have increased because you unnecessarily went out of your way for what is ultimately (IMO) just something fun you could do and/or an increased piece of mind. MAYBE this is a good idea if you have your system operating in insanely high ambient temps, but to suggest everyone should do this no matter the situation is a little absurd (again, IMO).

Heck, let's not stop there!, let's keep all our drives in a separate compartment (with their own direct cooler per each drive) that we attach to our cases! That makes the most sense! That way we can have direct lines of intake to exhaust across the entirety of our system guts! [Howard Dead]YEAH![/Howard Dean]

I'm sorry if I came across as being out of line with my first post, but you created something that really seems to be oriented more towards cooling newbies, and they tend to blindly take on information that appears to be good.

 

[WoodButcher]

Ha ha ha, I was waiting for the flames.
Cloudkat, I disagree w/ many of your points and won't argue them as many more educated/ qualified members will discuss these things. (I favor a positive pressure case, a link from one of your references). Rather than prove my stupidity by debating things that I have no qualification for I'll watch, listen and learn.
Thank you for a very informative thread and welcome to the forums!

Keep in mind if one or more of these turkeys around here happen to change your opinion or comes up with some additional info, that first post of yours can be edited. You may want to put your reference links up there, they're good. Keep up the good work, thanks again. :thumbsup:

 

[LOUISSSSS]

i don't care much about the noise of my case. i'm used to having two delta 120x38mm 3200rpm 151cfm Fans on my case (+ all the other fans that aren't case fans.) [fans r connected to a sunbeam rheobus] my fans x 2

anyone have any idea what kind of fan that is? ball bearing, sleeve..

when i turn my case fans down too much and walk out of my room, i get scared because i think my pc has turned itself off. seriously. haha

 

[aigomorla]

Originally posted by: WoodButcher
Rather than prove my stupidity by debating things that I have no qualification for I'll watch, listen and learn.
Thank you for a very informative thread and welcome to the forums!

Keep in mind if one or more of these turkeys around here happen to change your opinion or comes up with some additional info, that first post of yours can be edited. You may want to put your reference links up there, they're good. Keep up the good work, thanks again. :thumbsup:

LOL woodbutcher you said my exact words for me! But i wont let this turn into a flame feast.

Please keep this thread clean, id hate to do anything to it. Its pretty informative with the debates back and forth. Less anger, more proof!

But i dont know enough to speak up.

So im also watching! >:]

And welcome to the forums Kat.

Originally posted by: bunnyfubbles

I'm sorry if I came across as being out of line with my first post, but you created something that really seems to be oriented more towards cooling newbies, and they tend to blindly take on information that appears to be good.

Which is why people need to speak up if they see something is wrong. Not saying whose wrong in this situation tho, because i dont know. But with good proof's its easy to tell. :]

Originally posted by: Navid

CFM/dBa is a meaningless ratio.

This i do know enough, and i have to side with Kat. The amount of air it pushes at that sound is important in comparison across same voltage on different fans. So yes its very important.

Navid are you thinking about static pressure? Because on heat sinks static pressure plays a bigger role on fan selection.

 

[DerwenArtos12]

Originally posted by: aigomorla

Originally posted by: bunnyfubbles

I'm sorry if I came across as being out of line with my first post, but you created something that really seems to be oriented more towards cooling newbies, and they tend to blindly take on information that appears to be good.

Which is why people need to speak up if they see something is wrong. Not saying whose wrong in this situation tho, because i dont know. But with good proof's its easy to tell. :]

Originally posted by: Navid

CFM/dBa is a meaningless ratio.

This i do know enough, and i have to side with Kat. The amount of air it pushes at that sound is important in comparison across same voltage on different fans. So yes its very important.

Navid are you thinking about static pressure? Because on heat sinks static pressure plays a bigger role on fan selection.

I think in this particular case it's not as black and white and navid and kat are making it out to be. CFM/dbA ratio is a good way to measure the effiency of a fan but, it's DEFINATELY not the only thing you should judge a fan on. This particular measurement is useful but, it's in a big gray area for the exact reasons that both kat and navid already posted.

I'd like to just say this for this whole thread and to everyone that reads it. Read the whole thread! Just about every post in this thread has useful information for judging the merits of fans and how they're used colelctively but, there is no bible for choosing fans and this thread is certainly not a candidate for that position. I have been attempting to write a good fan selection thread for over a month now and I keep running into a GIANT brick wall called subjectivity. There is so much concrete evidence that there is no one fan for every job or even one perfect fan for one job. I've got nearly a hundred fans and I've played with and tweaked each one of them and I've chosen a few that are my favorites, some of which have actually picked up pretty good adoptions/endorsements from other users here, some most people won't touch with their friends 10 foot pole.

Anyone who is trying to choose a fan has to look at the whole picture, there isn't one magical fan, or fan type, or fan size that could possible be totally merited by less than the combination of there's DBA, CFM and subjective noise. IMHO YLs tick when upright and whirr when on their side but, I haven't bought any from petra's. Cooler masters stock case fans hum, really bad. Antec tricools tend to whine at higher RPM. SFDBs whirr but at a lower resonance than the cooler masters which makes them a lot more acceptable to me. Slipstreams tend to hiss at 1900rpms and the slower ones loose all flow with restriction. Ultra kaze's are among my favorites for low RPM noise but, the 2000 and especially the 3000 sound worse than a similar delta from whirr.

I could pick apart any fan I bought but, they each have their own purpose so I keep buying all of them for different purposes.

 

[Navid]

Originally posted by: aigomorla

Originally posted by: Navid

CFM/dBa is a meaningless ratio.

This i do know enough, and i have to side with Kat. The amount of air it pushes at that sound is important in comparison across same voltage on different fans. So yes its very important.

Navid are you thinking about static pressure? Because on heat sinks static pressure plays a bigger role on fan selection.

You asked!

No, I am talking about a ratio that has no value!

If you are interested to buy a car and the most important parameter for you is how much you have to pay for gas, you obviously want a car that can go the longest distance with the least amount of gas.
You ask about the Miles/Gallon ratio then.

Say car A goes 100 Miles with 2 Gallons of gas and car B goes 200 Miles with 5 Gallons.
Then, A has a ratio of 50 Miles/Gallon. While B has a ratio of 40 Miles/Gallon. So, for the same amount of gas, the first car can travel 25% farther.
Now, it is easy to compare the two cars for the gas consumption.

Let's say one car can go 50 Miles with one Gallon. Let's say another car can go 100 Miles with 2 Gallons. But, if a car can go 100 Miles with 2 Gallons, it can also go 50 Miles with one Gallon. So, I calculate the ratio, which normalizes to 1 Gallon. That's what a ratio is all about (to normalize). So, the two cars will be the same as far as the gas consumption goes. And that is shown by their ratios being equal.

So, the ratio is very useful.


You want a fan that moves the most amount of air with the least amount of noise.
So, you can find the ratio of air moved to the noise and use that ratio for the purpose of comparison.
But, you can do that only if you measure both air movement and noise with linear units.
Remember we said that if a car goes 50 Miles with 1 Gallon, then it can go 100 Miles with 2 Gallons? This is because Mile is a linear unit and Gallon is a linear unit too.
You cannot say that about CFM/dBA. dBA is not linear.

Why is dBA not linear?
Because if you double the dBA figure, you are not doubling the amount of noise. If the noise is represented with 30dBA, doubling that figure will give you 60dBA, which will be much much MUCH louder than just twice as loud.
If you double 5 Gallons, you will have 10 Gallons. There is twice as much gas in 10 Gallons as there is in 5 Gallons. So, Gallon is a linear unit.
Because of this, we cannot do everything I said we could do with Miles/Gallon with CFM/dBA. So, that is why I say the latter is a useless ratio.

It could actually be very misleading.
Let's say you have a choice to buy fans that move 60CFM and sound 30dBA. Ratio = 2.
Alternatively, you can buy fans that move 80CFM and sound 36dBA. Ratio = 2.22. These are just meant to be examples.
If you go by the ratio, you will choose the second fan because it has a higher ratio of 2.22. So, you think that it moves more air for the same amount of noise.
But, that is not true. If you buy two of the fans from the first type and put them in parallel, the two fans in parallel will move 120CFM. But, their noise will only be 33dBA.
Wouldn't you prefer to have 120CFM for only 33dBA as opposed to only 80CFM for 36dBA?

Grasping the Logarithm concept is difficult for many. The CFM/dBA ratio showing up in posts and publications as a mean to select which fan to buy is an evidence.

 

[aigomorla]

Your correct, but, isnt that why when people do reviews they take cfm/db ratio at 3 different voltage settings?

So they can get a basis on how it scales? Last time i checked they usually tested it at 7V 9V 12V

Some reviews take that even further and go to 5V to see if it makes clicking noises, and overclock them even at 13V to see how loud they scream.

I agree with you one solid CFM/DB ratio would be pointless, but if you had enough you could plot and scale it.

 

[DerwenArtos12]

Navid, I think the miscomunication here is that Aigomorla and I are very familiar with Dba scaling, therefore the CFM/DB ratio can be useful because we already know that DB is exponential and not linear. If I'm understading you right your main problem is that noobs probably won't know that and therefore get very confused trying to use cfm/db to buy a fan.

I totally agree that for a complete noob cfm/db should not be what they base their decision on but, for those that understand DB scaling it can still be useful.

 

[Navid]

Originally posted by: aigomorla
I agree with you one solid CFM/DB ratio would be pointless, but if you had enough you could plot and scale it.

But, that is the whole point of a ratio! So that you do not have to have it for different points and you don't have to plot it.

So, you are saying that the CFM/dBA is not a constant ratio and varies depending on what voltage (CFM) you are at. That is exactly what I am saying. Then, forget the ratio.

Imagine going to a car dealership and asking the gas mileage of a car:
You: Hello, what is the gas mileage of this car?
Dealer: It depends how much gas you have!!! For one gallon, it will be 50. But, for 2 gallons, it will be 47, for 3 gallons, ......!!!!

Would that answer make any sense to you?
The gas mileage of a car is how far it can drive with one gallon of gas; period

Is the CFM/dBA the amount of air a fan can move with only 1dBA of noise?
NO!

 

[Navid]

Originally posted by: DerwenArtos12
I totally agree that for a complete noob cfm/db should not be what they base their decision on but, for those that understand DB scaling it can still be useful.

Can you give a general example to show how it could still be useful?

 

[DerwenArtos12]

Originally posted by: Navid

Originally posted by: DerwenArtos12
I totally agree that for a complete noob cfm/db should not be what they base their decision on but, for those that understand DB scaling it can still be useful.

Can you give a general example to show how it could still be useful?

for comparing fans that either have the same CFM or the Same DB. It is an accurate measurement of the efficiency of the blades in one aspect.

 

[Navid]

Originally posted by: DerwenArtos12

Originally posted by: Navid

Originally posted by: DerwenArtos12
I totally agree that for a complete noob cfm/db should not be what they base their decision on but, for those that understand DB scaling it can still be useful.

Can you give a general example to show how it could still be useful?

for comparing fans that either have the same CFM or the Same DB. It is an accurate measurement of the efficiency of the blades in one aspect.

Don't you see we keep going back to this?

If the CFM is the same, you can just compare the dBA. So, you do not need the ratio.
If the dBA is the same, you can just compare the CFM. So, you do not need the ratio.
You have not shown an example of how the ratio can be useful.

Give me an example, where you only have the ratios of two different fans and you can tell me which one is going to give you more air movement for the same amount of noise.
The same as if you give me the gas mileage for two cars with nothing else, I can estimate to a very reasonable accuracy which car will drive more for the same amount of gas.

 

[aigomorla]

Navid im glad your getting all over this.

Because im sure there are many people who think like you do. So lets see if we can clear things up.

If your looking at 6 different fans. And you want to see what would fit your budget, you would need to decide on a votlage to base a comparison.

lets say your a spcr freak who hears pin drops on hay pale, sorry kat i had to throw it in there.

Your going to look for the quiest "reviewed/tested" fans you can find with the highest cfm. Then i would be comparing all the fans on either the 5V or the 7V range. Which ever had the best cfm to noise raito, i would go with on exhaust fans.

on sinks i would look at static pressure while holding cfm/noise. Hence why i feel you should spend a little more on these fans because of the static pressure element.


But basically its like asking you this first.

you have 5 cars that look the same but have different engines, and you want to go 70miles while only using 15mpg, and there all scattered out. You would pick the one closes to your target no?

 

[Navid]

Running two fans at the same voltage does not guarantee that they will make the same amount of noise or move the same amount of air or even are at the same RPM.
I would never do what you have suggested!

I read forums and review sites and have experience with different fans.
I know some of the basics like:
a fan blade with a huge dead zone in the middle is not going to be very efficient;
a filter in front of a fan reduces its air flow;
the noise you hear is a combination of the noise components coming from all the fans;
sometimes, one fan could dominate depending on how close it is to your ear or what is in front of it;
a setup best for cooling the CPU may not be best for cooling the graphics card or chipset and vice versa;
etc.

Based on those, I buy a few fans that I suspect can satisfy my requirements.

I will install the first fan and I will reduce the voltage until I am happy with the noise. Then, I will monitor how well it cools. I will log my temps under different conditions like idle, full graphics load, full CPU load, ....

Then, I will remove the fan and install the second fan and repeat.

So, for each fan, I will drop the RPM until I am happy with noise. Then, I log temps.
I make all my measurements at the same perceived noise level judged by my ears. After all, this whole thing is to keep my ears happy!
At the end, I will pick the setup that gives me the lowest temperatures at the same perceived noise.

I already gave an example where picking a fan based on the CFM/dBA ratio could result in an incorrect selection. Doing that at the same voltage makes no difference; still incorrect!

 

[Tweakin]

Wow...tough crowd on this thread. Great post for the new users out there!

 

[zorrt]

Navid I get where you are going by saying the ratio is useless in picking a fan, however the ratio is only there purely as a GUIDE. Just like the way I compare prices of HDD, see which is cheaper per mb. People like to put figures to aid in decision making.

Reason why this helps, lets say you are lazy and can't be bothered looking for reviews for fans. All you want is a fan that can push the most air with the least amount of sound. The ratio can help there. If the specs are correct (which I think in most cases they aren't) then ideally you will be getting what you want.

So for example, I am lazy and can't be bothered reading reviews, I am choosing between the Nexus 120 (36.87cfm 22.8dba) to the yate loon d12sm (70cfm 33dba). I find it difficult to compare the 2 cfm and 2 dba values as they are. No way I'd work out which is better just by staring at the numbers, so I'd try to make something out of them. Heres where ratio comes into play. I can now say the nexus push 1.6cfm/dba, yateloon pushes 2.12cfm/dba. Now I can clearly see the Yateloon pushes more air per dba and would make it a better buy (to keep it simple we'll pretend both fans are the same price).

So isn't that easier than staring at the original numbers. Remember, ratio is a GUIDE. By adding perceived sounds you're just complicating the matter which most people couldn't be bothered with. And if you go that far, we might as well say everyone has the equipment to measure sound sitting in their room.

Don't get me wrong, I do get where you're coming however that doesn't help average joe who has no experience at all.

 

[aigomorla]

Originally posted by: Navid

I will install the first fan and I will reduce the voltage until I am happy with the noise. Then, I will monitor how well it cools. I will log my temps under different conditions like idle, full graphics load, full CPU load, ....

Then, I will remove the fan and install the second fan and repeat.

So, for each fan, I will drop the RPM until I am happy with noise. Then, I log temps.
I make all my measurements at the same perceived noise level judged by my ears. After all, this whole thing is to keep my ears happy!
At the end, I will pick the setup that gives me the lowest temperatures at the same perceived noise.

I already gave an example where picking a fan based on the CFM/dBA ratio could result in an incorrect selection. Doing that at the same voltage makes no difference; still incorrect!

wow, and people called me crazy for putting about 30 min up front on a prebuild prep for my board b4 i start mounting it.

To each his own, Navid your way is of course better. Its more accurate, however, for a review purpose, i havent had an issue with following fan reviews from either SPCR or XS. And they all do it via CFM/DB ratio. :T


You method would be too time exhausting and also hard to reproduce for a comparison on a poor reviewer. Remember most of them dont get paid for this and they dontate all that time for the public.

:T

 

[godisknugen]

hey great guide, got some good pointers there.

anyway i´m just a little curious. is there an ideal arrangement of the three vertical componemts: psu cpu and gpu. some cases have psu in the bottom while traditionally it has always been at the top. and some cases, like the t40, places the mobo upside-down having the video card placed at the top and cpu in middle (with the addition of a strange airduct).
as a little break from the whole noise debate ;P

 

[DerwenArtos12]

Originally posted by: godisknugen
hey great guide, got some good pointers there.

anyway i´m just a little curious. is there an ideal arrangement of the three vertical componemts: psu cpu and gpu. some cases have psu in the bottom while traditionally it has always been at the top. and some cases, like the t40, places the mobo upside-down having the video card placed at the top and cpu in middle (with the addition of a strange airduct).
as a little break from the whole noise debate ;P

none of them are the best unfortunately. it depends entirely on what other components are going into the system.

 

[Sheninat0r]

Originally posted by: godisknugen
hey great guide, got some good pointers there.

anyway i´m just a little curious. is there an ideal arrangement of the three vertical componemts: psu cpu and gpu. some cases have psu in the bottom while traditionally it has always been at the top. and some cases, like the t40, places the mobo upside-down having the video card placed at the top and cpu in middle (with the addition of a strange airduct).
as a little break from the whole noise debate ;P

There's no "best" way to do it - the original ATX specification put the PSU at the top so that its fan could act as an exhaust for the rest of the system, but the original ATX specification also didn't consider the fact that computers would be getting zillions of times hotter, what with the Pentium 4. The airflow in the ATX spec is inefficient, since cooling really wasn't a concern when they made it.

Reverse ATX, like you see in Lian-Li's V series, reverses airflow as well. Air enters from the CPU area and flows back to the hard drive area, so that the air going to the CPU is cool and not "pre-heated" from other components. With the video card on top, however, the heat from the CPU rises and heats it instead of, like in regular ATX, heat from the video card rising into the CPU. BTX also has similar airflow, as it was designed to cool Intel's infamous Prescotts.

 

[zorrt]

Originally posted by: godisknugen
hey great guide, got some good pointers there.

anyway i´m just a little curious. is there an ideal arrangement of the three vertical componemts: psu cpu and gpu. some cases have psu in the bottom while traditionally it has always been at the top. and some cases, like the t40, places the mobo upside-down having the video card placed at the top and cpu in middle (with the addition of a strange airduct).
as a little break from the whole noise debate ;P

As preference i prefer PSU on the bottom. That way it wouldn't be taking in the hot air from every other component in the system thus causing the PSU to heat up and ramp up its fan speed.

Motherboard mounting helps with situations where you PSU is bottom mounting and the power cables for motherboard are too short to reach. Not sure if it has much effect to cooling though. But theoretically air towards the lower section of the case is cooler than the top. So with this motherboard mounting direction the CPU would benefit from the cooler air rather than pulling in recycled air from the videocard. But then the very hot videocards of today would now be pulling in recycled air from the CPU.

 

[Bluefront]

Concerning cloudkats posts...... keep in mind there are very few absolutes when judging the whole picture of a computer, with respect to fans, airflow, temperatures, noise, etc. One might say the ideal setup runs quieter, cooler, with the least amount of airflow. But what if your location is dusty from sources that cannot be changed very much? (you live on a dirt road without AC and have screens on your windows).

In this scenario the ideal setup would be clogged with dust quickly, the temperatures would raise, and you'd have to rise the fan-speed/noise to compensate. Which means that an ideal setup for the guy on a dirt road may involve a noisier but well filtered computer running total positive pressure......different from the OP.

This indicates that there are numerous variables to consider when making a judgment about your own intended setup.......what works best for one person, might be a flop for another person. The presence of differing internal components changes the whole picture. A computer with a really hot video card, requires a different airflow setup from a computer using on-board graphics.......if you desire the quietest, coolest setup.

Then there's the fan thing......you can throw out mfg specs. There's no way to figure out anything using made-up specs. Some manufacturers simply lie, some use different testing techniques...the list is endless. When you hold a fan up running in free air, it may well sound quiet to you. Put that same fan behind a filter, or blowing into a tight fin-pack, and it may well become a noisy P.O.S. Mfg specs won't tell you that. And specs won't tell you how you particular fan reacts to a PWM controller. Some click like mad, some are quiet as a pin. The variables that affect a fan are endless.

The most you can hope for......find several good review sites that have forums. Read with others have to say about a certain part/a certain airflow setup. Expect to experiment yourself to find your own ideal setup. Don't rely on specific guide-lines. There is no absolute "best" way.