Restriction?

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
Praxis1452
Posts: 13
Joined: Sat Dec 23, 2006 11:07 pm

Restriction?

Post by Praxis1452 » Sat Dec 23, 2006 11:17 pm

Would SPCR be able to add a restriction to a fan to see how much something such as a fan grill take away from the rated CFM or a water cooling radiator? Most cases have fan grills of some sort. I know most silent enthusiasts usually remove these but I'm just wondering what sort of effect they may have. As I see SPCR doing more fan reviews it seems that this would be kind of logical as it would help people decide what fan to use under certain situations.

You could organize it by CFM range. You take ones that are +-10cfm categories and compare them. Also many fans may be silent and move air but it on a cpu cooler the slightly louder fan maybe worth it just so that it would actually keep the cpu cool enough while the slightly quieter one won't.

Not sure if this has been asked before sry if it has.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun Dec 24, 2006 2:24 am

Well I've done plenty of "restriction" experiments involving filters and grills on the intake side of fans. There's also a restriction / back-pressure effect when blowing air through a grill, a heatsink, or a complicated exhaust path. The problem here is that different fans respond differently to restriction, be it on the intake or exhaust side. Any conclusion or number you could apply to a particular restriction, would be subject to many different factors.....and probably not universally accurate.

For instance.....a wire fan grill supposedly reduces airflow about 10%, but I've never been able to measure anything close to that. Check this by running a fan in free air at a fixed voltage while you measure the rpm. Then add a fan grill to either side of the fan. See if the rpm changes. You can do this experiment with filters and definately see the rpm drop.

It's much easier for a fan to suck through a restriction, than it is to blow through the same restriction. But as to an exact percentage.....varies.
[size=75]"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke[/size]

Praxis1452
Posts: 13
Joined: Sat Dec 23, 2006 11:07 pm

Post by Praxis1452 » Sun Dec 24, 2006 11:01 am

Bluefront wrote:Well I've done plenty of "restriction" experiments involving filters and grills on the intake side of fans. There's also a restriction / back-pressure effect when blowing air through a grill, a heatsink, or a complicated exhaust path. The problem here is that different fans respond differently to restriction, be it on the intake or exhaust side. Any conclusion or number you could apply to a particular restriction, would be subject to many different factors.....and probably not universally accurate.

For instance.....a wire fan grill supposedly reduces airflow about 10%, but I've never been able to measure anything close to that. Check this by running a fan in free air at a fixed voltage while you measure the rpm. Then add a fan grill to either side of the fan. See if the rpm changes. You can do this experiment with filters and definately see the rpm drop.

It's much easier for a fan to suck through a restriction, than it is to blow through the same restriction. But as to an exact percentage.....varies.
I see that it would be hard but I think SPCR could come up with a standard. I was thinking about it because The Noctua fan seems to push 62cfm but that's only in free air I assume so maybe the other fans would push more air in an restrictive situation because at $17 per fan I was thinking about them but it's expensive as hell compared to just a regular yate loon.

I'm sure SPCR could do both sucking and pushing... Pushing would be used more often actually so that what I would expect. heh. I got this idea from Vapor on XS testing fans and using a BIX or BIP as restriction and most fans lost anywhere from 25%-50% of their cfm. The panaflos did the best.

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

Post by jaganath » Sun Dec 24, 2006 12:00 pm

Would SPCR be able to add a restriction to a fan to see how much something such as a fan grill takes away from the rated CFM or a water cooling radiator? Most cases have fan grills of some sort. I know most silent enthusiasts usually remove these but I'm just wondering what sort of effect they may have.
If you search you will find there have been previous discussions about this topic. You can generally tell by looking at a grill how restrictive it will be; the brass "ring-type" grills present little restriction, stamped grills more so. You'd be surprised how much they decrease airflow. As you say, anyone who's serious about noise reduction will remove these.
Noctua fan seems to push 62cfm
Where is that number from? I can only find 48CFM :

http://www.candccentral.co.uk/acatalog/Noctua.html
at $17 per fan I was thinking about them but it's expensive as hell compared to just a regular yate loon.
Save your money. The Yate Loon will be almost as quiet and do much better against backpressure.

Most decent fans will have a static pressure rating (in mmH2O), this tells you their pressure capability, higher is better.

Praxis1452
Posts: 13
Joined: Sat Dec 23, 2006 11:07 pm

Post by Praxis1452 » Sun Dec 24, 2006 12:41 pm

jaganath wrote:
Would SPCR be able to add a restriction to a fan to see how much something such as a fan grill takes away from the rated CFM or a water cooling radiator? Most cases have fan grills of some sort. I know most silent enthusiasts usually remove these but I'm just wondering what sort of effect they may have.
If you search you will find there have been previous discussions about this topic. You can generally tell by looking at a grill how restrictive it will be; the brass "ring-type" grills present little restriction, stamped grills more so. You'd be surprised how much they decrease airflow. As you say, anyone who's serious about noise reduction will remove these.
Noctua fan seems to push 62cfm
Where is that number from? I can only find 48CFM :

http://www.candccentral.co.uk/acatalog/Noctua.html
at $17 per fan I was thinking about them but it's expensive as hell compared to just a regular yate loon.
Save your money. The Yate Loon will be almost as quiet and do much better against backpressure.

Most decent fans will have a static pressure rating (in mmH2O), this tells you their pressure capability, higher is better.
higher is better doesn't really mean much. It's nowhere near accurate enough for me.

62 if you read the fan review right on the front page of SPCR...

And by looking at the grills it's not a very accurate depiction. So if 50% of the area is covered flow is reduced by 1/2? I doubt it.

Simply with SPCR doing more fan reviews just like how they keep improving their PS reviews they should do some restriction tests. And how exactly do you know that the yate has more mmH2O or has more static pressure? Atleast how about static pressure in relation to some restriction.

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

Post by jaganath » Sun Dec 24, 2006 2:58 pm

higher is better doesn't really mean much. It's nowhere near accurate enough for me.
Oh, sorry, we better re-arrange all of SPCR's testing priorities to suit you.... :roll:
62 if you read the fan review right on the front page of SPCR...
62CFM sounds way too high to me. The Yate Loon D12SL-12, spinning about 100rpm faster, pushes a little less than 48CFM; no way a slightly different blade shape and 100rpm less is going to give you almost 30% more flow. SPCR's CFM methodology has been called into question more than once.
So if 50% of the area is covered flow is reduced by 1/2? I doubt it.
I never said that.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Sun Dec 24, 2006 3:17 pm

Maybe you should consult wikipedia about this.
[size=75]
[color=#3366ff]e8400, Corsair XMS2-5400, Asus 8400GS, Gigabyte EP45-UD3R, Chenbro SR209/Nexus 120, T-Balancer, Enhance ENP-5150GH 500W, Seagate 7200.7 160MB [/color][/size][size=75][color=#cc0033](Retired: XP-120, FSP530-60GNA, Antec SP2.0 500W, Antec SLK3000B)[/color][/size]

Praxis1452
Posts: 13
Joined: Sat Dec 23, 2006 11:07 pm

Post by Praxis1452 » Sun Dec 24, 2006 3:48 pm

jaganath wrote:
higher is better doesn't really mean much. It's nowhere near accurate enough for me.
Oh, sorry, we better re-arrange all of SPCR's testing priorities to suit you.... :roll:
62 if you read the fan review right on the front page of SPCR...
62CFM sounds way too high to me. The Yate Loon D12SL-12, spinning about 100rpm faster, pushes a little less than 48CFM; no way a slightly different blade shape and 100rpm less is going to give you almost 30% more flow. SPCR's CFM methodology has been called into question more than once.
So if 50% of the area is covered flow is reduced by 1/2? I doubt it.
I never said that.
:roll: it's so bad of me to ask more of reviews... yeah... perhaps we should just all be happy with most of the sites out there then. Too bad your satisfied. I'm not.


I'm not sure how exactly SPCR does their fan review as it's not listed in this forum. I'll relook the review.

No you didn't say the last part but just relying on a general rule that open=more cfma and restriction=less cfm still means pretty much nothing while true. For example a BIP vx BIX rad. The BIX is supposed to be more restrictive and it probably is. However to what extent may determine what fans you choose and how you'll go about them.

NeilBlanchard
Moderator
Posts: 7681
Joined: Mon Dec 09, 2002 7:11 pm
Location: Maynard, MA, Eaarth
Contact:

Post by NeilBlanchard » Sun Dec 24, 2006 6:41 pm

Greetings,

The fan test methodology was written up in an earlier article:

http://www.silentpcreview.com/article687-page1.html
Sincerely, Neil
http://neilblanchard.blogspot.com/

Rasterman
Posts: 37
Joined: Sun Nov 12, 2006 3:47 am

Post by Rasterman » Tue Dec 26, 2006 5:26 pm

Bluefront wrote:It's much easier for a fan to suck through a restriction, than it is to blow through the same restriction. But as to an exact percentage.....varies.
This is very interesting. Do you have any links about this? Can you please explain why?

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Tue Dec 26, 2006 8:52 pm

Bluefront wrote:For instance.....a wire fan grill supposedly reduces airflow about 10%, but I've never been able to measure anything close to that. Check this by running a fan in free air at a fixed voltage while you measure the rpm. Then add a fan grill to either side of the fan. See if the rpm changes. You can do this experiment with filters and definately see the rpm drop.
The rpm does not have to slow down (at least not much) for the airflow through the grill to be impeded. Impedance can cause turbulence and reduce airflow w/o slowing the fan speed. The difference can be felt on your skin, and it is measurable with an anenometer. Typically, a wire grill impedes about 20% of airflow.

To test this, take a typical mid/slow speed 80~120x25mm fan w/a wire grill on it (whichever side you want, afaik), hold it it 6" from your face and fire it up at 12V. Make sure the hand that's holding the fan does not impede airflow. Remove the grill, try it again. You should be able to feel the difference in the amount of airflow on your face.
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

pcy
-- Vendor --
Posts: 137
Joined: Sat Aug 13, 2005 2:46 am
Location: Central London, England

Post by pcy » Wed Dec 27, 2006 1:58 am

Hi,

Could we possibly extend this to measuring the resistance or actual airflow through a CPU heatsink.


I'm not certain that I beleive the manufacturers quoted figures fro CFM: but even if I did they are CFM in free air; and the Heatsink represents a massive resistance to airflow.


The actual aiflow through a Heatsink is typically (IMO - I've calucalted this from the difference between input and output air temeratures) between 10CFM and 20 CFM. The rest of the fan's energy is going into overcoming the back pressure.


It's entirely possible that the optimum fan design for a CPU cooler is quite different from the optimum design for a case fan where the back pressure is much lower.



Just a thought....




Peter

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

Post by jaganath » Wed Dec 27, 2006 4:14 am

It's entirely possible that the optimum fan design for a CPU cooler is quite different from the optimum design for a case fan where the back pressure is much lower.
See my reply in "blowers vs fans" thread.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Thu Dec 28, 2006 2:17 am

I'm not convinced what MikeC said about wire fan grills is 100% accurate. The experiments I've done prove otherwise. Here's one test I've used......with a 120mm Yate Loon and a wire grill on the intake side. Blow the fan at the cpu through a duct. Measure the temp of the cpu under various usage.....then add the wire grill. If the CFM is impeded(reduced) by the 20% MikeC states, even if the rpm remains constant, the temp of the CPU should rise due to the reduced airflow. Well it has remained virtually constant every time I tried this experiment....proving to me the CFM of the fan is unaffected by any significant amount, certainly not 20%.

As to blowing a fan at your face through a wire grill.....it will feel somewhat different due to the change in the airflow, not a reduction in cfm. The airflow through a grill is not as consentrated as without the grill, and it feels different/less strong. IMHO....this does not prove a reduction in total airflow, again not 20%.

Blowing through a wire grill reduces airflow somewhat due to back-pressure, but not by any significant amount. The effect of sucking through the same wire grill is even less than blowing through the grill....I'd guess much less than 10%.
[size=75]"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke[/size]

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Dec 28, 2006 8:10 am

Bluefront wrote:I'm not convinced what MikeC said about wire fan grills is 100% accurate.
I'm hardly the only one to say this: viewtopic.php?t=8493

Also, the absence of CPU temp change doesn't prove there's no airflow reduction. My very earliest heatsink reviews showed how this can happen easily -- an increase in airflow beyond a certain point for a CPU/HS combo doesn't help. If you have more than this amount, then dropping the airflow a bit won't change anything.

Just try the "feel" test. Besides you don't have to agree with me.

I don't choose to live with dust-generating parrots either. :lol:
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

Praxis1452
Posts: 13
Joined: Sat Dec 23, 2006 11:07 pm

Post by Praxis1452 » Thu Dec 28, 2006 2:03 pm

we are talking about these http://www.ncix.com/products/index.php? ... gies%20Inc right? 10% seems a bit much but I guess it depends on how low you run your fans.

Anyone have the specs for the CM fans that come with the stacker 810? they seem completely silent to me... heh

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Thu Dec 28, 2006 4:09 pm

Ok here's how I conducted the fan grill/restriction test. I adjusted the airflow on the CPU to a point that when the fan speed was reduced further(a reduction in airflow), the CPU temp went up. I used this point for the test.....any reduction in airflow would cause the CPU temp to rise. So if I put a fan grill in front of the fan intake, and the airflow was reduced by 20%, the CPU temp would surely rise. Didn't happen.....the temp rise was not measurable by speedfan (it was less than 1C).

I consider this to be a valid test......and I tried this test with several fans, and numerous filters. All produced a reduction in airflow, resulting in a CPU temp rise......all except the wire grill, which had no effect at all.
[size=75]"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke[/size]

pcy
-- Vendor --
Posts: 137
Joined: Sat Aug 13, 2005 2:46 am
Location: Central London, England

Post by pcy » Fri Dec 29, 2006 2:41 am

Hi,

I'd have to say that that's a valid, but very insensitive test.


The probelm is that the CPU temp is only reported to 1C.


Let's look at the numbers:

The CPU generates a constant amount of heat, so the total rise in the air temp flowing through the cooler multiplied by the air flow through the cooler must equal (in the right units) that heat output.

The input air temperatue is constant (room temperature) so we have output air temperture increase (vs room) x aiflow through cooler = constant

However the mean air temp within the cooler must be (first order approximation) the average of the air input and the air output temperatures. Heat output is constant, and heat transfer from the heatsink to the air is a linear function of the temperature difference, so any increase in mean air temp in the heatsink must be accompanied by an equal increase in CPU temp.

So a 10% change in airflow will produce a CPU temp rise of 5% of the air input to air output difference.


Typical values for that air temp differnce are 20C to 30C at idle, so you'd need a 10% change in airflow before you could necessarily detect any change in CPU temperature, given a sensitivit of 1C.


But there is more.


The experiment was conducted under low airflow conditions with the fan pushing air through a heatsink. I'm quite certain that grills and filtes offer very little resistance to airflow at low speed, but much more (as a proprotion) at high speed. I'd expect their resistance to be proportional to the square of the air velocity. By contrast a heatsink definatly offers a linear resistance unsder the sort of conditions we are considering and a substantial resistance as well. Typical airflow through a heatsink may be only 1/4 of the free air flow through a fan.

All this means that if a grill caused a 10% reduction to airflow under free air conditions it might cause less than 1% under low speed + heatsink conditions.


And the resulting CPU temperature change would be a fraction of 1C...




Peter

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Fri Dec 29, 2006 4:26 am

Yeah....that test is only valid to 1C using speedfan to detect the temp. But I also use different temp testing hardware......an external LCD temp gage with a remote sensor, with a readout to .1C, and a non-contact laser beam temp tester.

And of course there is a slight change when anything is put close to the fan intake.....like your hand 6" from the fan. But for all practical purposes, any temp change under 1C is insignificant.....IMHO.
[size=75]"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke[/size]

pcy
-- Vendor --
Posts: 137
Joined: Sat Aug 13, 2005 2:46 am
Location: Central London, England

Post by pcy » Fri Dec 29, 2006 6:13 am

Hi,

I agree that if there is no detectable change in CPU temp the effect of the grill must be neglible in that particualt case.


But I'm also saying that your results are not necessarily insonsistnt with Mike Cs, if his 20% reduction in airflow was taken in free air at 12V.




Peter

iamweasel
Posts: 14
Joined: Mon Aug 07, 2006 1:28 am

Possible pressure test rig

Post by iamweasel » Fri Dec 29, 2006 7:16 am

I was wondering about this as well, and I think there might be a way to test how much pressure a fan generates at a given rpm.

What you need is a airtight box with hole to mount the fan into and a U shaped transparent tube going through a wall as well. Then put coloured liquid in the tube leaving both ends open. The tube ends up being a simple pressure gauge.

Now, turning on the fan and after the liquid settles at a certain height, you can measure the difference in the levels of the liquid in the U tube to measure the pressure difference inside and outside the box. Then turn the fan the other way around and repeat.

I think this simple experiment would help suggest how well a given fan would cope with airflow restrictions; the higher the pressure difference the fan can generate, the better it is likely to be when the airflow is restricted.

Just my 2 cents.

I would actually be very interested in this, since I've been trying to figure out the best fan to use on the lower part of P180, in which the airflow is quite restricted. (Using a fanless PSU, no less)

Post Reply