A new idea on fan stacking
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
A new idea on fan stacking
We all know that the intakes on the P18X's are very restrictive. Yet some like me can't do with out them as we have dusty rooms. Which means having to live with reduced airflow or faster fans.
Now we also know that by stacking fans on top of each other you get batter restriction handling on the fans. The reason people don't do that is because:
A) More noise as you have more fans.
B) The exhaust from the front fan is spinning in the direction of the fans which means that the impeller of the next fan is effectively "chasing" the the air, reducing efficiency.
So today I'm proposing a solution to B):
SaxonPC Flow Straighteners
The idea is to put one between each "stage" of the fan stacks so that the air reach the 2nd fan at a right angle, allowing them to work to their fullest potential.
As for noise as the additional fans are toward the inside of the case their noise may be able to be reduced by wrapping the entire assembly in foam, as well as foam at the front cover. Plus if the pressure does greatly improve we may be able to dial down the fans. i.e. we get extra air at same or even lower RPMs.
Now the question is thus:
-Are my theory sound?
-What would be a good fan for this. The noctua P12's don't seem to live up their hype as high pressure fans, Ultra Kaze's are too loud and Slip Streams have low pressure (small motor=low torque). so that leaves the S-Flex's, correct?
I'm thinking about having a triple fan set up for the upper chamber.
Now we also know that by stacking fans on top of each other you get batter restriction handling on the fans. The reason people don't do that is because:
A) More noise as you have more fans.
B) The exhaust from the front fan is spinning in the direction of the fans which means that the impeller of the next fan is effectively "chasing" the the air, reducing efficiency.
So today I'm proposing a solution to B):
SaxonPC Flow Straighteners
The idea is to put one between each "stage" of the fan stacks so that the air reach the 2nd fan at a right angle, allowing them to work to their fullest potential.
As for noise as the additional fans are toward the inside of the case their noise may be able to be reduced by wrapping the entire assembly in foam, as well as foam at the front cover. Plus if the pressure does greatly improve we may be able to dial down the fans. i.e. we get extra air at same or even lower RPMs.
Now the question is thus:
-Are my theory sound?
-What would be a good fan for this. The noctua P12's don't seem to live up their hype as high pressure fans, Ultra Kaze's are too loud and Slip Streams have low pressure (small motor=low torque). so that leaves the S-Flex's, correct?
I'm thinking about having a triple fan set up for the upper chamber.
Last edited by Hypernova on Wed Jun 11, 2008 8:51 pm, edited 1 time in total.
The idea seems doable, but wouldn't it be easier to just buy a thicker fan?
Well from what I read the Ultra's seem to have a very clear motor whine to them. Plus the restriction of the filter + door appears to be too great for any single silence oriented fan to overcome. The biggest issue I think is that Scythe doesn't post data on pressure.Modo wrote:The idea seems doable, but wouldn't it be easier to just buy a thicker fan?
Last edited by Hypernova on Sat May 31, 2008 4:19 am, edited 1 time in total.
As I intend to get high end GFX at the end of this year I need all the CLEAN air I can get.jaganath wrote:yes. compressors in jet engines use exactly the same principle (rotor/stator arrangement). could be expensive tho!Now the question is thus:
-Are my theory sound?
It would also have the effect of a positive pressure set up. There have been stories of GFX card fans stalling due to low case pressure and actually having air sucked in through the exhaust ports. Plus with dust control positive pressure is critical.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
Re: A new idea on fan stacking (I think)
"small motor=low torque" is only true for constant power input. It's easy to put more power into a fan hub, so "small hub = low pressure" just ain't so. The SFLex fans are in fact high-pitch (just not as high as the 9-blade Slipstreams). I believe you will find GW NCB fans are better than SFlex at handling pressure - the NCBs are very low-pitch.Hypernova wrote:Slip Streams have low pressure (small motor=low torque). so that leaves the S-Flex's, correct?
Since there are no (none, zip, zero) standard methods of measuring how fans respond to backpressure, all reported results are subjective. I'm suspicious of reports on the new 9-blade Noctua fans. They certainly look as if they'd handle backpressure just fine. I just wish I could measure it and confirm it.
Re: A new idea on fan stacking (I think)
Is this the fan your are referring to?Felger Carbon wrote:"small motor=low torque" is only true for constant power input. It's easy to put more power into a fan hub, so "small hub = low pressure" just ain't so. The SFLex fans are in fact high-pitch (just not as high as the 9-blade Slipstreams). I believe you will find GW NCB fans are better than SFlex at handling pressure - the NCBs are very low-pitch.Hypernova wrote:Slip Streams have low pressure (small motor=low torque). so that leaves the S-Flex's, correct?
Since there are no (none, zip, zero) standard methods of measuring how fans respond to backpressure, all reported results are subjective. I'm suspicious of reports on the new 9-blade Noctua fans. They certainly look as if they'd handle backpressure just fine. I just wish I could measure it and confirm it.
The SPCR review was certainly favourable, although I haven't seen anyone review them in regards to pressure. The SPCR review also mentioned possible issues with bad batches which got me a bit worried.
Re: A new idea on fan stacking (I think)
technically AMCA 210 is the standard for measuring CFM and pressure of fans (this is how mfg.s generate their P-Q performance graphs e.g here), for examples of the test apparatus see here, here and here; but these are far too expensive and bulky for an individual SPCRer to construct.Felger Carbon wrote:Since there are no (none, zip, zero) standard methods of measuring how fans respond to backpressure
From some posts here it seems that the Aerocool Turbine might have some decent pressure too (0.889 mmH2O), Anyone care to comment on that?
Current choices are now:
-S-FLEX 1200
-GW NCB
-Aerocool Turbine
Since I'm living in New Zealand any order for more special parts has to come from overseas so I'm hoping to get it right the first time. I'm mainly ordering from FrozenCPU and since I'm ordering the straighteners from them anyway I gravitate toward getting the turbine or S-FLEX to even out the shipping.
Current choices are now:
-S-FLEX 1200
-GW NCB
-Aerocool Turbine
Since I'm living in New Zealand any order for more special parts has to come from overseas so I'm hoping to get it right the first time. I'm mainly ordering from FrozenCPU and since I'm ordering the straighteners from them anyway I gravitate toward getting the turbine or S-FLEX to even out the shipping.
-
- *Lifetime Patron*
- Posts: 5316
- Joined: Sat Jan 18, 2003 2:19 pm
- Location: St Louis (county) Missouri USA
Airflow straightener.....rather than order something at great shipping expense, why not a DIY solution? In my local hardware stores, I can find a panel in the lighting section that is meant to be placed under a floressant (sp.) lightening fixture. They are about 3/8" thick, plastic, with a squarish open hole pattern. The holes are about 1/2" diameter.
I've got a panel in the basement, and have used it in computer projects before. Here's an example....
I've got a panel in the basement, and have used it in computer projects before. Here's an example....
Last edited by Bluefront on Sun Jun 01, 2008 2:24 am, edited 1 time in total.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
Re: A new idea on fan stacking (I think)
I apologize, Jaganath. I should have specified "respond to backpressure in terms of noise".jaganath wrote:technically AMCA 210 is the standard for measuring CFM and pressure of fans (this is how mfg.s generate their P-Q performance graphs e.g here), for examples of the test apparatus see here, here and here; but these are far too expensive and bulky for an individual SPCRer to construct.Felger Carbon wrote:Since there are no (none, zip, zero) standard methods of measuring how fans respond to backpressure
That's more like using it as a grill isn't it? I don't have any material on hand that can fit well anyway.Bluefront wrote:Airflow straightener.....rather than order something at great shipping expense, why not a DIY solution? In my local hardware stores, I can find a panel in the lighting section that is meant to be placed under a floressant (sp.) lightening fixture. They are about 3/8" thick, plastic, with a squarish open hole pattern. The holes are about 1/2" diameter.
I've got a panel in the basement, and have used it in computer projects before. Here's an example....
Any opinion on the fans? If no one has anything else to suggest I'll use the GW and order the stuff at the end of tomorrow. Last time I ordered stuff from FC it took two weeks to get here. So sit tight until then.
Funny thing about shipping is that some stuff provided you buy in largish quantity the price actually works out cheaper then buying local in NZ. One example is the S-FLEX 800 (only model available in here) which after shipping and exchange works out to about 30USD.
-
- *Lifetime Patron*
- Posts: 5316
- Joined: Sat Jan 18, 2003 2:19 pm
- Location: St Louis (county) Missouri USA
Hypernova.... If you cut out a section of the stuff and place it between two fans, the effect will be the same as that "airflow straightener". It's about 3/8" thick, and would form little airflow channels.
Now whether that effect would do anything for your idea is the question. I've never been convinced that "technique" did anything positive for airflow. Good luck.
Now whether that effect would do anything for your idea is the question. I've never been convinced that "technique" did anything positive for airflow. Good luck.
Al right, finally settled on the NMB-MAT (PanaFLO) FBA12G12L. I know some people say they have dropped in quality but the posts were all 2+ years ago. I'll be the guinea pig for everyone and see if they changed. Also got a kaze master to dial down the fans as I doubt I'll run then anywhere near 12V.
Parts should be here in 2 weeks. We'll see what happens then.
Parts should be here in 2 weeks. We'll see what happens then.
Alright, The fan are in and the results are not exactly great.
-First on the NMB-MAT's I bought 3 and the 3rd one beyond 1000RPM noticeably clicks more then the other two. A good balance for airflow and noise for me was at about 800~900RPM. If any of the 3 fan's rpm deviate by more then 100 there is a noticeable turbulence noise.
-Room temp 22C
-Case temps dropped by 2C at 900 RPM compared to before. One noticeable effect is that speeding up the fans will now actually decrease temps where as the S-FLEX even when cranked to max at 1600 wouldn't do any thing.
-Biggest change was VGA (8400GS) temps as it dropped to 53C compared to 62 before.
-No real changes in CPU temps. 4C drop at best, but then again I'm using an TR ultra-90 with the fans at bare minimum levels.
About the air straighteners:
-Pressure measured by taping a piece of paper to the exhaust and see how high it gets pushed.
-With two NMB's stacked directly the back pressure actually dropped. Putting a straightener inbetween raised it to original levels.
-When the 3 NMBs are stacked directly putting my hand in front of the intake there is a large drop in pressure, with straighteners in between there is still a drop but the paper stays at about 50% of the level compared to no restrictions.
-The straighteners do work as advertised at directing the air flow into a straight beam. This may be useful for people looking for fan duck equivalents.
Overall I find it useful, considering that I will be getting hot VGA cards later this year. With the kind restrictions the fans are dealing with they are doing a decent job.
One more thing, the Kaze Master used to control the fans are rated at 1A. However even at the 0.25 A the NMB's are rated at the MOSFETs become too hot to touch. Also considering the fact that there are no heat sinks on them I doubt the controller can actually handle anything more then 0.5A safely.
-First on the NMB-MAT's I bought 3 and the 3rd one beyond 1000RPM noticeably clicks more then the other two. A good balance for airflow and noise for me was at about 800~900RPM. If any of the 3 fan's rpm deviate by more then 100 there is a noticeable turbulence noise.
-Room temp 22C
-Case temps dropped by 2C at 900 RPM compared to before. One noticeable effect is that speeding up the fans will now actually decrease temps where as the S-FLEX even when cranked to max at 1600 wouldn't do any thing.
-Biggest change was VGA (8400GS) temps as it dropped to 53C compared to 62 before.
-No real changes in CPU temps. 4C drop at best, but then again I'm using an TR ultra-90 with the fans at bare minimum levels.
About the air straighteners:
-Pressure measured by taping a piece of paper to the exhaust and see how high it gets pushed.
-With two NMB's stacked directly the back pressure actually dropped. Putting a straightener inbetween raised it to original levels.
-When the 3 NMBs are stacked directly putting my hand in front of the intake there is a large drop in pressure, with straighteners in between there is still a drop but the paper stays at about 50% of the level compared to no restrictions.
-The straighteners do work as advertised at directing the air flow into a straight beam. This may be useful for people looking for fan duck equivalents.
Overall I find it useful, considering that I will be getting hot VGA cards later this year. With the kind restrictions the fans are dealing with they are doing a decent job.
One more thing, the Kaze Master used to control the fans are rated at 1A. However even at the 0.25 A the NMB's are rated at the MOSFETs become too hot to touch. Also considering the fact that there are no heat sinks on them I doubt the controller can actually handle anything more then 0.5A safely.