Ultra 120 Extreme fan setup

[malik22]

Hi guys



Im soon getting my new pc the cpu is a Q6600 whats the best fan setup for the Ultra 120 for best temps.Two 70 cfm fans in a push pull or one higher cfm fan pushing?

[Kosior]

It depends what would be the "higher cfm" fan, if it'd be Delta 220cfm fan, it would be better. Two 70cfm fans in push pull would give you like 105cfm~

[malik22]

When I say higher I mean like a 130cfm.

[Kosior]

so it would probablly be better than two 70cfm fans

[g1raffe]

Are you concerned with noise, or just performance?

2 quieter fans in push-pull can achieve as good or better performance than a faster single fan whilst being quieter.

[spookmineer]

Two 70cfm fans in push pull would give you like 105cfm~

No.

Two 70 CFM fans in push-pull configuration will give you 70 CFM - with higher back pressure.
A second fan can never increase the air velocity if it's spinning at the same speed as the first one. It can however increase the back pressure.

[Kosior]

No.

Two 70 CFM fans in push-pull configuration will give you 70 CFM - with higher back pressure.
A second fan can never increase the air velocity if it's spinning at the same speed as the first one. It can however increase the back pressure.

really? isn't it like with the pumps? if you run 2 pumps in series you'll have like 1.5 waterflow of one pump.... Thought it's the same with fans

[loonym]

I love fans! :D I'm using 2 scythe 110 cfm on my TRue. The noise? Maybe some would complain but they are quiet compared to the delta on my ttbt or the tornado on the xp-90.

[Kosior]

I love fans! :D I'm using 2 scythe 110 cfm on my TRue. The noise? Maybe some would complain but they are quiet compared to the delta on my ttbt or the tornado on the xp-90.

hope that those "Scythe 110cfm fans" aren't Slipstreams...

[loonym]

hope that those "Scythe 110cfm fans" aren't Slipstreams...They are indeed, as seen here http://www.scythe-usa.com/product/acc/026/sy1225sl_detail.html model is SY1225SL12SH.
Why would you hope they are not? My temps are outstanding.

[malik22]

I was thinking of the ULTRA KAZE 120mm Case Fan 3000rpm is this fan noisy?

[Clue69Less]

No.

Two 70 CFM fans in push-pull configuration will give you 70 CFM - with higher back pressure.
A second fan can never increase the air velocity if it's spinning at the same speed as the first one. It can however increase the back pressure.

You should get a little more than 70cfm through the HS. Some of the push fan's air will go laterally out through the side of the fins. The lowering of pressure by the pull fan will typically raise the efficiency of the push fan also.

[Vapor]

You do get noticeably more airflow than 70CFM.

[Kosior]

They are indeed, as seen here http://www.scythe-usa.com/product/acc/026/sy1225sl_detail.html model is SY1225SL12SH.
Why would you hope they are not? My temps are outstanding.

According to Vapor's data posted a few days ago in some theard , Slipstreams are designed to be case fans, which means they loose massive amount of their airflow when mounted on a radiator. That's why Scythe realesed Scythe Ultra Kaze. These are proper radiator fans :P

[spookmineer]

You should get a little more than 70cfm through the HS. Some of the push fan's air will go laterally out through the side of the fins. The lowering of pressure by the pull fan will typically raise the efficiency of the push fan also.

You do get noticeably more airflow than 70CFM.

The only reason why there is a slight increase in airflow from 2 fans in push-pull on a heatsink, is because the back pressure of the heatsink only weighs in half as much on each fan. As a result, more energy is available to move the air for each of the two fans.

In free air (the specifications of fans are always measured in free air, without obstructions) 2 fans in series will not increase the airflow but will increase the backpressure.
Two parallel fans will not increase the back pressure, but will obviously double the airflow.

Analogy to pumps:
Two parallel pumps: twice the flow, but the head (how high the water can be lifted) will be equal to a single pump.
Two pumps in series: same flow as a single pump, but the head will now be doubled in comparison to a single pump.
[In flowcharts you can see that the head is at its maximum when flow is zero, and flow will be at its maximum when the head ("back pressure") is zero. This means when there is back pressure (the pump has to do work) the flow will always be less than 100%]

So, I guess in reality a slight increase in airflow from 2 fans in push-pull (but not 105 CFM if a single fan is rated at 70 CFM of airflow).
The CFM of a fan is measured in free air, on a heatsink this amount of airflow will never be reached.
A single fan has low back pressure, decreasing the airflow by X %.
Adding another fan will increase the back pressure, so the loss of airflow will be less than X %. Still lower than 70 CFM on all accounts though.

[Vapor]

Okay, just ran some tests with extra YL D12SL-12s I have here...

Radiator was an MCR120. Both fans are are PTS D12SL-12s and test within 1CFM of each other at 12V in open air.

One fan:
Open air: 47.59CFM
Radiator: 35.13CFM

Two fans:
Open air: 62.86CFM
Radiator: 47.31CFM**

**I accidentally broke a blade on one of them before the test (they were running and sitting on an angled surface...one slid and fell off and got a wire caught in it and ripped a blade off :(). And due to limitations of the equipment I have with me right now, this was pull-pull (i.e., arranged: RAD-FAN1-FAN2->, rather than push-pull which is FAN1-RAD-FAN2->).

So with a non-ideal pull-pull configuration of identical non-counter rotating fans (and with one fan having a missing fan blade for the last test), you get within 1% of the airflow you get with just 1 fan in open air.

I have run just-for-kicks tests previously with counter-rotating (i.e., one fan spins CW, the other CCW) and the numbers were staggering. CFM went up a lot in open air and barely dipped on a radiator. Those tests were also pull-pull. Also had my dB meter out at the time and noise went up accordingly...3dB for like-rotating (perfectly expected--double the number of identical 'noise makers' and you get a 3dB increase) and ~3.5-3.8dB for counter-rotating (more turbulence probably made more noise).

[spookmineer]

First of all, I appreciate you taking the time to do these measurements!
The increase of airflow with 2 fans is far more then I expected but I still have some questions...

What kind of anemometer were you using, and what was the test setup?

Converting wind speed to CFM can be tricky, with an anemometer so close by, did you account for the hub of the fan?
Is the anemometer as big or bigger than the 120 mm fan (a smaller anemometer will pick up a higher air speed near the tips of the fan as opposed to closer to the fan's hub)?

The questions are just out of curiosity though... The D12SL is rated 47 CFM @ 1350 RPM, your measurements seem to be spot on.


Measuring the airflow "inside" a radiator/HSF is almost impossible to do, but this is what ultimately counts. Without discrediting the measurements, I still have my doubts to what is actually being measured here... because I can't explain the numbers from the theory. Maybe my implementation of the theory is wrong somewhere but I can't figure out how.


Aside from the measurements this does not affect the conclusions from the measurements a lot... The relative increase is clear, no matter what the "real" CFM may be. Again, the difference is a lot more than I would have guessed. Maybe I can look up my old physics teacher and flick him on the head, I can't explain the results from the theory.

[Vapor]

The anemometer is a windmill type that is smaller than 120mm, but it's not placed up against the fans anyway.

I use a very-low restriction duct-like system....all the air that a fan moves must go through the anemometer. For the radiator tests, I mount the fan pulling through the radiator and mount the radiator onto the ducting.

I think all Delta and NMB-MAT/Minebea fans I've tested get within 1% of rated CFM (including the Delta Tri-Blades that spin in the opposite direction, which proves that my testing isn't influence by air spin--more on this later). I've tested A LOT of 120mm and and posted the results--it's in the stickies. SPCR recently (well, maybe 9 months ago) adopted a testing methodology similar to mine as well.

Because fans also have a tendency to 'spin' the air, I only use pull configs (where the spun air is exhaust--not passing through the anemometer) with the setup I have here at school (don't have any of my sound testing testbed, or even the full-on airflow testbed I used in the reviews--just too cumbersome to bring).

I also tested San Ace 1011s after I posted that and got 113.7CFM for 1 fan and 108.9CFM for 2 fans pulling through the radiator. Fortunately neither of them broke a blade (I've broken 2-3 San Aces out of 8-9 I've bought--so fragile and expensive!).

Again, for the radiator tests...I place a radiator in the system. All air must pass through the anemometer, then through the 'ducting', then through the radiator, and finally through the fan.

My understanding of the theory is that each fan adds some air-moving energy to the air--no matter how much 'energy' it already has. But energy vs. airspeed is a logarithmic scale (each fan design having its own scale of course). So adding an 800RPM fan onto a 2400RPM fan might increase airflow only a tiny bit (but should still increase airflow), but two 2400RPM fans and two 800RPM fans of the same design have the same additive scaling. You can keep adding fans and they keep adding energy, but the airspeed will barely increase after awhile with each addition.

This is similar to dBA and noisemaker counts. You can have 1 bumblebee making 40dBA and then double bumblebee count to 2 and have 43dB, and continue adding bumblebees until you get to 1024 bumblebees making 73dBA of noise. The next bumblebee you add will be entirely insignificant to the energy of the noise being made though, even though it's identical to the bumblebee you added 1023 bees ago that bumped 'energy' up 3dBA. (okay, kinda a weird analogy, and I'm stretching the language a bit, but it's just an analogy and I hope it works for some :))

[g1raffe]

I was thinking of the ULTRA KAZE 120mm Case Fan 3000rpm is this fan noisy?
Yes, 3000rpm is very loud.

A lot of people would consider above 1200 or so load, some people 800, some 2000, etc, different for everyone, but 3000rpm on a 120mm fan is definitely loud.

[cabbspapp]

Sleeve bearings tend to get loud at high RPM's from what I've heard.

[Kosior]

Sleeve bearings tend to get loud at high RPM's from what I've heard.

Well they all get loud at high rpms but it's not only about the bearings. The noise depends on the blade design also.