Confessions of an insane big-fan fanatic

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Confessions of an insane big-fan fanatic

Post by Felger Carbon » Sat Feb 24, 2007 8:47 am

I posted a wrong formula. [Long pause for the reading public to recover from the sudden shock.]

We know (or should have learned by now) that doubling the RPM of a fan doubles its CFM and increases its noise by 18dBA. Or we can halve the RPM and "dial down" the noise by 18dBA. There's ample empirical evidence supporting this; I'm comfortable describing it as a fact.
-------------------------

So: If we "dial down" two fans and place them side-by-side, we get our original CFM (from one fan) back, and only increase our noise by 3dB on account of two fans. Net gain in noise level, 15dBA.

We repeat this exercise, replacing the two fans with four running at a quarter of the original one-fan cooler's RPM. Net gain in noise level, 30dBA.

Say the original fan was a 120mm. Now we have a much quieter array of 4ea 120mm fans that provide the same CFM. No change in cooling air, large reduction in noise level. Now we replace the four 120mm fans with one 240mm fan. Instead of 4 fan motors we have one, and we have only half the linear "turbulence area" at the tips of the fan blades, so the real-world performance should be better than 30dBA.

This is entirely logical and the math is correct. The problem is, in the real world it's wrong (sigh).
---------------------------

The problem is the enclosed computer case, and its passive air vent. If the fan (or fan array) is at the intake, the air vent is the exhaust. Or reverse.

In our computer case, when we went from a 120mm exhaust fan on the rear to a (fictional) 240mm intake fan on the left side, did we also quadruple the area of the passive air vent (in this case, the exhaust vent)? No, we didn't. And that's why we don't get a 30dBA improvement in noise level when going from a 120mm fan to a (fictional) 240mm fan.

My real-world 220mm fan doesn't have quadruple the area of a 120mm fan so obviously I'm not writing about it! :)
----------------------------------

Consider a large box with a 240mm fan on one side and a 240mm diameter vent on the opposite side. Can it move the same CFM thru the box as a 120mm fan running 4 times the RPM would move thru a 120mm vent in the same size box? Yes. Would the 240mm fan be 30dBA quieter? Yes. But only with the vent being 4 times the size of the vent used with the 120mm fan.

Therefore, the comparison is not oranges vs oranges. The "30dBA" improvement is only valid in open air, with no computer case and therefore no vent.

Does my real-world 220mm fan in my real-world computer case do a good job cooling quietly? It sure does! Does it cool almost 30dBA more quietly, compared to a 120mm fan? Hell no!

So what I did was fudge one formula, the one for the noise of different sized fans at constant CFM. And nobody caught me at it, because the results of the formula I used conform reasonably well to the real world. (I came up with 10.53dBA as the advantage of a 220mm fan vs a 120mm.) An actual, completely accurate, formula would be very complex and would have to include the vent area as one of the more important variables. In other words, an accurate formula would not be very useful to us SPCR individuals.

So I do confess but I don't repent. :twisted:

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sat Feb 24, 2007 9:19 am

Confessions of an insane big-fan fanatic
The first step on the road to recovery is admitting you have a problem. :mrgreen:
Say the original fan was a 120mm. Now we have a much quieter array of 4ea 120mm fans that provide the same CFM. No change in cooling air, large reduction in noise level.
I would agree with that, with the proviso that the fans are not models with noisy bearings; bearing noise, it seems to me, does not decrease anywhere near as drastically as aerodynamic noise when rpm is decreased.

It seems to me there is a somewhat law of diminishing returns going with ever-bigger fans; because most retail heatsinks are not sized for fans bigger than 120mm, sure you get more airflow from a big fan but it's not precision airflow so a lot of it will flow uselessly past the heatsink fins in the free field.

Still, I'll be going with a single big fan in my next build, somewhat hypocritical I guess. :D

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sat Feb 24, 2007 10:39 am

jaganath wrote:It seems to me there is a somewhat law of diminishing returns going with ever-bigger fans; because most retail heatsinks are not sized for fans bigger than 120mm, sure you get more airflow from a big fan but it's not precision airflow so a lot of it will flow uselessly past the heatsink fins in the free field.
Rats! You've correctly anticipated my next posting on this "thread". Which will await gathering some emperical data centered on using the Freezer 64 Pro in the "big fan" case. :D

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sat Feb 24, 2007 10:47 am

jaganath wrote:I'll be going with a single big fan in my next build.
I'm willing to do some light kibitzing via PM or email, or do you prefer to have "learning experiences" on your own? Email would allow me to easily pass some photos along via my personal 10MB web space.

For example, I've modified 3 "Xclio 3060" (includes PSI Solo 604) cases, and am bringing up the 4th now. The first was less than totally successful, but 2 and 3 are working fine and are on this table now.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sat Feb 24, 2007 11:00 am

Felger Carbon wrote:
jaganath wrote:I'll be going with a single big fan in my next build.
I'm willing to do some light kibitzing via PM or email, or do you prefer to have "learning experiences" on your own? Email would allow me to easily pass some photos along via my personal 10MB web space.
Well when I say 'big' it will probably be 140mm max; these are midgets compared to the behemoths you have had such success with. :D

I usually just rely on having really cool parts so even sub-20CFM volumetric airflow is sufficient; I am about as far away from pushing the limits of air cooling as it is humanly possible to be. :lol:

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Sat Feb 24, 2007 11:55 am

A side mount fan would tend to do best with a cooler that faces it,such as Aerocool Dominator,which is an oversized Thermalright type,140mm fan,though the popular 120 Thermalrights,Big Typhoon,also should do well

A HS like a ninja should like a bottom to top airflow or front to rear.

Now,there are quite a few chips,from Semprons to mid-level X2's and C2D's that do fine with a passive HS,a slow 120 and the PSU helping some.
On such a rig,a mega-fan is overkill unless there's issues with multiple HDDs,a hot running vi card...

Could a Dominator HS be run fanless with a 220 mm door fan reversed as exhaust? (no fan for intake air) ? I'd think so. Some creative ducting,airflow management,RPM fine tuning might make that real effective
I'd figure,however,that the location of the computer would be especially relevant. A side fan would be percieved quieter if the case is under a desk and to the left.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sat Feb 24, 2007 12:35 pm

ronrem wrote:Could a Dominator HS be run fanless with a 220 mm door fan reversed as exhaust? (no fan for intake air) ? I'd think so. Some creative ducting,airflow management,RPM fine tuning might make that real effective.
Ron, you've obviously given this matter some serious thought. Why not actually try out your theories? The Xclio 3060 case is only $35 and includes the fan. The 220mm fan will start from 3.0 to 3.1V, and I run mine at 3.6V, which is 5V less two power diode drops (1N400x).

Iateronmly
Posts: 17
Joined: Wed Feb 28, 2007 7:51 am
Location: Ninilchik Alaska
Contact:

Post by Iateronmly » Wed Feb 28, 2007 8:21 am

Yah - big fans are addicting. I currently have a 220mm from a Thermaltake side panel and an Antec Big Boy 200mm slated to be shoe horned into the custom case I'm making. I'm just going to set them as intakes, I've never seen much gain with series configurations. The only down side is the lack of velocity with big fans. But I've strived to eliminate as much impedence as I can to the airflow. Like placing the 1U PSU and HDD behind the the motherboard and putting about 38 square inches of 78% open mesh on the back panel. I hope it'll be enough to cool a .64 cubic foot area. :lol:

I'm just replying because while I've lurked I haven't seen anyone with the 200mm Big Boy just mention it's available. At it's narrowest part of the frame it's 199mm, about 220mm at the widest part of the frame, 30mm deep, the mounting holes are about 8.5" acrost diagonally. Accuracy of their dBA claims - I cannot verify, but subjectively to my ear, it's quite tolerable even on medium.

Felger - don't worry about the big fan confessions til you take it to far and you have people run simulations though.....

Image

disphenoidal
Posts: 333
Joined: Wed Apr 05, 2006 9:54 pm
Location: USA

Post by disphenoidal » Wed Feb 28, 2007 10:18 am

How did you get that simulation? That kind of thing would be very in demand around here, I think.
[size=75][i]Athlon64 3000+ @ 2.6 GHz, Radeon 9550, Seasonic S12-380, Zalman CNPS7000-Cu, Yate Loon 120mm, Centurion 5 case
Core2Duo E6320 @ 2.4GHz, Seasonic S12-330, Zalman CNPS7000-AlCu, Yate Loon 120mm, Ultra MicroFly[/i][/size]

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Wed Feb 28, 2007 10:53 am

that is an awesome simulation. what cfd software is that?

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Wed Feb 28, 2007 11:46 am

That 'simulation' looks wrong to me.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Wed Feb 28, 2007 12:03 pm

Tzupy wrote:That 'simulation' looks wrong to me.
Looks plausible to me, flow is slow at the intake and fast at the exhaust due to reduced exhaust vs intake area.

spookmineer
Patron of SPCR
Posts: 749
Joined: Sat Nov 11, 2006 6:02 pm

Post by spookmineer » Wed Feb 28, 2007 4:13 pm

It would be cool to see how messy it gets with all the hardware in place.

J. Sparrow
Posts: 414
Joined: Wed Jan 17, 2007 7:55 am
Location: EU

Post by J. Sparrow » Wed Feb 28, 2007 4:25 pm

it seems to be Ansys.

Iateronmly
Posts: 17
Joined: Wed Feb 28, 2007 7:51 am
Location: Ninilchik Alaska
Contact:

Post by Iateronmly » Wed Feb 28, 2007 8:33 pm

Tzupy wrote:That 'simulation' looks wrong to me.
It is wrong - it was a quick one I had my friend run (I don't have access to Ansys atm). I didn't have all my exhaust vents placed into the model, and obviously there's no account for the PSU/VGA's direct exiting or VGA card/HSF and motherboard interferring with the airflow. But I was needed a quick demonstration of what a side intake does to airflow so I had him run a simple model. Just wanted a visual to show how it helps to eliminate dead spots and how a person that asked can't just add cfm ratings.

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Thu Mar 01, 2007 1:12 am

Simply 'wrong' wasn't specific enough, I agree, but I was tired and didn't want to type much. :lol:
It looks as there's nothing IN the case. The HDDs, gfx cards(s), CPU HSF and cables affect those flows IMHO.
And the *large* fans were simulated as providing uniform flow, which IMHO can't be right: the larger the blades, the larger the turbulence.
While general case airflow can't hurt, the actual cooling happens in the contact areas, so local flow / turbulence is also important.
For a quiet system the general airflow displayed in the 'simulation', and the one in Felger Carbons's case, is overkill IMHO.

Iateronmly
Posts: 17
Joined: Wed Feb 28, 2007 7:51 am
Location: Ninilchik Alaska
Contact:

Post by Iateronmly » Thu Mar 01, 2007 3:21 am

Tzupy wrote: For a quiet system the general airflow displayed in the 'simulation', and the one in Felger Carbons's case, is overkill IMHO.
One person's overkill is not enough for another. I already replied to all your concerns on how it's wrong - it was a simple model for a graphic example on a point. If you want I can post what'd happen if you add a blowhole to the design - now that'll cause unneed turbulence/noise/ reduced cooling.

Big fans are great - if you acknowledge their limitations. Their static pressure is miserable, along with the exiting velocity. Spot cooling, yah, not great at all. But if you're looking to increase your air exchange rate, they're fan-tastic. I have no remorse for my love of big fans, and I stall not recant. They have their place, they might not be the right option every time, they're an option just the same though.

I've read many of the discussions here about the 220mm fans and 200mm to go along with my own experiences so far. I don't recall Felger Carbon ever saying they're a must have, I definately don't. But in my experience so far, they're very viable and worth looking into. If you can fit them, they're definately worth looking at.

zds
Posts: 261
Joined: Tue Aug 03, 2004 1:26 pm
Location: Helsinki, Finland
Contact:

Post by zds » Thu Mar 01, 2007 8:07 am

jaganath wrote:It seems to me there is a somewhat law of diminishing returns going with ever-bigger fans; because most retail heatsinks are not sized for fans bigger than 120mm, sure you get more airflow from a big fan but it's not precision airflow so a lot of it will flow uselessly past the heatsink fins in the free field.
I think the "diminishing" applies only as long as you are using the traditional ATX case and air cooling. I believe that if you use some heat pumping to spread the heat to wider area and possibly not make the big fan blow through the case at all, the limits for reasonable size of the fan are lot higher.

For example, you could put a heavily undervolted medium-sized fan (like 172-175mm one used in server racks) sideways to the front of the case and have it blow fresh air through heat exchanger and straight out from the other side. Or something.

"Heat exchanger" here refers to heat sink, radiator or some other means of getting the pumped heat transfer to air.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Thu Mar 01, 2007 9:01 am

I think the "diminishing" applies only as long as you are using the traditional ATX case and air cooling
Which is what 99% of people use. Don't get me wrong, I respect the aerodynamic efficiency of big fans but I don't think we should go down a path where fans get bigger and bigger; it's just not necessary with today's cool components (except for GPU's, where the form factor precludes using axial fans).

zds
Posts: 261
Joined: Tue Aug 03, 2004 1:26 pm
Location: Helsinki, Finland
Contact:

Post by zds » Thu Mar 01, 2007 12:58 pm

jaganath wrote:Don't get me wrong, I respect the aerodynamic efficiency of big fans but I don't think we should go down a path where fans get bigger and bigger
Yeah, I didn't mean it as a general solution for everyone.. It was just a comment on the physics of the fans. The reason I'm into them is more to do with my desire to try out things and build a bit unconventional systems. I think regular 120mm ones are plenty enough for the regular desktop computer systems.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Thu Mar 01, 2007 5:45 pm

Iateronmly wrote:I've read many of the discussions here about the 220mm fans and 200mm to go along with my own experiences so far. I don't recall Felger Carbon ever saying they're a must have, I definately don't. But in my experience so far, they're very viable and worth looking into. If you can fit them, they're definately worth looking at.
Most SPCR types would kill :twisted: for a fan that pushes the same CFM as a Yate Loon 120mm but is more than 10dBA quieter. The 13-blade 220mm fan is that quieter fan. It's cheaply available - $35 including Xclio 3060 case at Newegg - and only needs five modifications to produce a case that's far quieter than, for instance, a Solo:
1. cut out the exhaust fan area (easy)
2. cut the 3.5" internal HDD cage to make room for HDD suspension
3. suspend the 220mm fan
4. remove the fan from the PSU and go passive
5. Seal the front panel (behind the bezel) with 1/8" rubber (McMaster), or two layers of Dynamat Extreme (or equivalent).

Step 2 is the most difficult; there are two ways to go about it (dremel and non-dremel).

Does anyone doubt that a 120mm fan can cool more quietly than a 70mm fan? A 120mm is closer to the 70mm than a 220mm is to a 120mm! :D

edit: added step 5

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Fri Mar 02, 2007 3:10 am

I agree 100% with the cfm / noise ratio of large fans, but the issue is how effective is the usage of all that airflow.
IMO only when we'll have flexible heatpipes from the CPU and GPU to large heatsinks mounted on the exhaust areas, the large fans are going to make a big difference.
Such flexible heatpipes have been built for some 15 years, but are not in the silent PC domain yet.
And for such a setup I would rather have two 140 mm fans on the bottom of the case, with a sound muffler around the base of the case.
Not to forget a good dust filter, I remember your post about how much dust gets in the computer when using a 220 mm fan. :wink:

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Fri Mar 02, 2007 4:09 am

Tzupy wrote:...I would rather have two 140 mm fans on the bottom of the case, with a sound muffler around the base of the case.
A 200mm fan would produce the same CFM at significantly lower noise. Currently the only commercial case with a 200mm fan, the Antec 900, uses that fan as an exhaust on the top of the case! :wink:

Dust? What dust? Don't you vacuum? 8)

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Fri Mar 02, 2007 4:16 am

The 220 mm fan wouldn't leave room for the muffler, which would be best distributed along the case base. Two 140 mm would fit better.

Of course I use a vacuum cleaner, but I have a long-haired tomcat. 8)
Last edited by Tzupy on Fri Mar 02, 2007 7:34 am, edited 1 time in total.

Iateronmly
Posts: 17
Joined: Wed Feb 28, 2007 7:51 am
Location: Ninilchik Alaska
Contact:

Post by Iateronmly » Fri Mar 02, 2007 6:55 am

Tzupy wrote:I agree 100% with the cfm / noise ratio of large fans, but the issue is how effective is the usage of all that airflow.
True - that is a concern. Attempting to increase cooling and reduce noise by increasing the exchange rate through the case solely isn't really an efficient nor an elegant way to achieve the end. Efficent would be spot cooling everything of concern and moderate exchange rate. But with my experiences so far, I've achieved superior end results with big fans. With each larger fan I go to it becomes harder to make good use of it, but I haven't found the end yet. Reason I'm building my next case the way I am.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Fri Mar 02, 2007 7:59 am

Tzupy wrote:The 220 mm fan wouldn't leave room for the muffler, which would be best distributed along the case base. Two 140 mm would fit better.
I said "200mm", by which I meant the Antec fan. Just under 8", about right for a fairly wide computer case, as Antec proved with the 900. :wink:

BTW: the Antec 200mm fan is almost exactly the same fan area as two 140mm fans, the other reason I mentioned it.

christopher3393
Posts: 289
Joined: Fri Sep 23, 2005 9:46 pm
Location: Bloomington, MN

Post by christopher3393 » Thu Mar 15, 2007 5:28 pm

Felger Carbon: I would really love any suggestions you would have for using this Antec fan for this build:viewtopic.php?p=322015#322015

The MM case is an 18" cube and can be custom built. It looks to me that I might be able to squeeze 2 BigBoys on top and 2 in front. Yes, ridiculous. Clearly front fans would be intake, top most likely exhaust. Single biggest heat/noise problem in this build is the 8800GTX. ANY suggestions much appreciated!!!

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Thu Mar 15, 2007 7:52 pm

christopher3393 wrote:Felger Carbon: I would really love any suggestions you would have for using this Antec fan for this build - Single biggest heat/noise problem in this build is the 8800GTX. ANY suggestions much appreciated!!!
I'm afraid I can't help. First, I use integrated mobo graphics so the 8800GTX is "terra incognita" for me. And I have all the work I can handle with the 220mm fan, which comes in a case, so I don't mess with the Antec "big boy", the details of which (mounting etc) continue to be held by Antec as a closely guarded secret.

I can wish you the best of luck tho. And I do! :D

christopher3393
Posts: 289
Joined: Fri Sep 23, 2005 9:46 pm
Location: Bloomington, MN

Post by christopher3393 » Fri Mar 16, 2007 6:21 am

Thanks anyway. I suppose there isn't much to go on as far as mounting the Big Boy other than a few pictures like these: http://www.xyzcomputing.com/index.php?o ... mitstart=1

http://img.hexus.net/v2/internationalev ... P190.4.jpg

Doors
Posts: 7
Joined: Sat Mar 17, 2007 3:06 am

240MM fan

Post by Doors » Sat Mar 17, 2007 3:29 am

Hi,
I've read this thread with great interest, though am quite new to silent PC modding.
I have the Thermaltake Mozart TX case which has room for 4 fans mounted in a square on the back side of the case (the Mozart TX case is an unusual shape - more square then oblong).
I was looking for a 240mm fan that would fit here instead of 4 x 120mm fans, so was keen to read this thread.
The idea would be to run it very slowly to be quiet but get enough air movement to equal what 2 or so 120MM fans.
Does anyone know where you can get a 240mm fan that might mount on the standard fittings of 4 x 120 mm fans?
Many thanks for any advice.
Doors.

Post Reply