PWMs v DC fans ?

Associate
Joined
17 Feb 2009
Posts
869
Got myself 2x140mm and 3x120mm corsair RGBS and they're awesome.
They are all DC.
I have made a fan curve in the bios and they're pretty much silent all the time.

What is the difference between PWM fans and DC ? All I know is PWM are 4 pin and DC are 3 pin
 
Soldato
Joined
31 May 2005
Posts
15,426
Location
Nottingham
Essentially, Pulse width modulation (PWM) fans are controllable by the system meaning fan speeds are changed "on the fly" if you wish with the fan speeds changing as and when needed.

For example, a CPU heatsink fan is always (I think) PWM.

You set a target temperature and the fan speeds will do what is needed.

If your system is idle, fan speeds are low, if system is using more resources, fan speeds increase.

If you have many larger fans and/or large CPU fans, you do not want them churning away at full speed as they can be loud. The best fans move more air and you do not want them running at max speed ALL the time but do need them to do so when gaming for example.

PWM fans will need PWM support from the header they are connected to. Cheaper boards may have just one PWM fan header, some have none. The better boards, like yours has 4 PWM headers. In this regard, connecting a 3 pin fan to a 4 pin header will not give PWM control, likewise, connecting a 4 pin fan to a 3 pin header with an adapter will not give PWM control.

Some people dislike their fans running at "full speed" so will set a curve anyway even if using PWM but it is personal preference. If your whipping your system within an inch of its life, overclocking etc, there are cases for sticking with the PWM control and setting a fan curve, all depends.
 
Last edited:
Soldato
Joined
6 Jun 2008
Posts
11,353
Location
Finland
Essentially, Pulse width modulation (PWM) fans are controllable by the system meaning fan speeds are changed "on the fly" if you wish with the fan speeds changing as and when needed.
PWM just allows a fan to spin at lower speeds.
Not really much difference outside of that.
Same applies to standard fans.
Fans starting with some 3V are capable to running really slow compared to full speed.
During its active years SilentPCreview did tests of fans including that.
 
Soldato
Joined
1 Jul 2011
Posts
8,548
While PWM theoretically has the ability to idle at lower speed than variable voltage, I've ran into several instances that motherboard PWM circuitry/software wouldn't idle fans at as low a speed as variable voltage did. This was on motherboards with fan headers user could choose to use either PWM or variable voltage control.
 
Soldato
Joined
11 Sep 2013
Posts
11,634
Well, there you have it - Factoring in all the outlying examples and assuming them to be representative, there is absolutely no difference whatsoever and it's all down to what your motherboard will permit.

Personally I just run them off a dedicated controller anyway, as I trust that more than motherboard software. My PWMs are alright being controlled DC only, but get a bit clicky at lower levels and will not achieve minimum revs on DC only.
 
Associate
Joined
19 Jul 2015
Posts
282
...I've ran into several instances that motherboard PWM circuitry/software wouldn't idle fans at as low a speed as variable voltage did...
Would that happen to be Asus motherboards? By default they won't let the duty cycle fall below 20%, which is well above the minimum speed of some fans. You can persuade it to go lower by running "QFan tuning", though it doesn't always work quite right.
 
Associate
Joined
22 Nov 2018
Posts
2,408
A PWM fan is still powered by DC current to pins one and two. So think of it as PWM vs non-PWM fans. Also make sure you buy the correct type of fan;

SP (static pressure) fans are for blowing air through a restricted area such as a radiator or heatsink.
AF (AirFlow) fans are for blowing air into an open area such as with case fans.

I`ve seen people use regular case fans on a rad without knowing there`s a difference. It still works but not optimal. Some people like to use SP fans in a case with air filters as they restrict the airflow slightly.
 
Associate
OP
Joined
17 Feb 2009
Posts
869
I put an SP120 on the right of my heatsink, and put the stock Noctua one on the left ( push pull ) not only does it look good, it has also bought my temps down nicely too.
 
Soldato
Joined
1 Jul 2011
Posts
8,548
A PWM fan is still powered by DC current to pins one and two. So think of it as PWM vs non-PWM fans. Also make sure you buy the correct type of fan;

SP (static pressure) fans are for blowing air through a restricted area such as a radiator or heatsink.
AF (AirFlow) fans are for blowing air into an open area such as with case fans.

I`ve seen people use regular case fans on a rad without knowing there`s a difference. It still works but not optimal. Some people like to use SP fans in a case with air filters as they restrict the airflow slightly.
Indeed, all of our computer fans are DC fans.
There are 2 man difference between variable voltage and PWM.
Variable voltage lowers 12v to lower voltage to power fan at lower speed .. generating heat as a bi-product.
PWM sends different length/speeds 12v pulses to motor .. the more time the 12v power is on the faster fan runs .. no heat is generated by pulsing power.​
PWM motor has mor torque at lower speeds, but in our fans that is not an issue.

I've found AF fans on case and even coolers to be a waste of money.
Reasoning is case grills and fins create as much or more resistance than air coolers and similar resistance as radiators do. Add a filter (many if not most case intakes are filtered) and case fans have to overcome way more resistance than cooler fans.
A fan with static pressure rating of 1.365mmH2O has 1.365mmH2O more pressure on exhaust side than on intake side. This is the same pressure difference as is on your feet versus on your neck / chest standing at sea level. The air pressure at sea level is 1.365mmH2O more than at 5 feet above sea level. So, as you can see our fans make very very little pressurel. ;)
Also, for airflow is created by pressure differential. Higher pressure air flows into lower pressure air.
 
Soldato
Joined
6 Jun 2008
Posts
11,353
Location
Finland
SP (static pressure) fans are for blowing air through a restricted area such as a radiator or heatsink.
AF (AirFlow) fans are for blowing air into an open area such as with case fans.
Static pressure means zero airflow and fan blowing into closed box.
Not very usefull value for real usage...
And that airflow isn't any more usefull value:
Because it's "car's top speed in free fall" number with fan running along with airflow and doing zero work to actually push that air.

What matters is P-Q curve and how fan behaves between these two extremes.
And that isn't linear straight curve and rarely published.
With also some changes in shape depending on speed.

Of course noise generated would be another variable.



Indeed, all of our computer fans are DC fans.
There are 2 man difference between variable voltage and PWM.
Variable voltage lowers 12v to lower voltage to power fan at lower speed .. generating heat as a bi-product.
PWM sends different length/speeds 12v pulses to motor .. the more time the 12v power is on the faster fan runs .. no heat is generated by pulsing power.​
Slower running fan does always less work and hence consumes less power.
In fact voltage control is really based on limiting how much current fan can draw by effectively adding series resistance.
It's that voltage controlling component, which suffers from power loss with both voltage loss in it and current through it.
Though with lower current total consumed power still decreases compared to fan running at full speed.

And with no transistor having zero resistance when conducting and unlimited speed during on/off transition PWM actually also generates some heat in controlling component.
While fan itself always wastes some power into its less than 100% efficiency when running.
Which is up to quality/friction of bearing and design of motor itself.
There's nothing preventing PWM fan from being power and current hungry hog and nothing preventing non-PWM fan being very frugal.
 
Associate
Joined
5 Mar 2017
Posts
1,825
Location
Cambridge
PWM would allow very low speed, but if the fans in question would be running at higher speed all the time, the 3 pin should be fine.
An efficient way to control 3 pin fans is the use of a molex adapter to 3 pin (something like https://www.overclockers.co.uk/bitf...v-adapter-20cm-sleeved-red-red-cm-194-bx.html )
Simply changing the position of the pins from the molex side would allow 12V, 7V or 5V. No need to change anything with the motherboard.
https://silentpcreview.com/get-12v-7v-or-5v-for-your-fans/
 
Soldato
Joined
11 Sep 2013
Posts
11,634
Static pressure means zero airflow and fan blowing into closed box.
Not very usefull value for real usage...
Unless you know to include it as one of several contributory factors, like Martin's Liquid Lab says to... :p

And that airflow isn't any more usefull value:
Unless you know to include it as another of several contributory factors, like Martin's Liquid Lab says to... :p

Because it's "car's top speed in free fall" number with fan running along with airflow and doing zero work to actually push that air.
You keep saying this, but it doesn't get any less incorrect the more you say it.

What matters is P-Q curve and how fan behaves between these two extremes.
And that isn't linear straight curve and rarely published.
Actually it often is pretty straight, as demonstrated by Martin's Liquid Lab...

Do you ever actually read this stuff, or do you just guess at the TL;DR version?
 
Associate
Joined
23 Dec 2018
Posts
1,020
Got myself 2x140mm and 3x120mm corsair RGBS and they're awesome.
They are all DC.
I have made a fan curve in the bios and they're pretty much silent all the time.

What is the difference between PWM fans and DC ? All I know is PWM are 4 pin and DC are 3 pin


Only real downside of DC fans is that bios usually won't let them run as slow at minimum than it will with PWM ones.

In my bios DC minimum speed is about 50%, PWM around 25%. I have a silent system at idle with mostly PWM fans and almost had to get rid of 2 x 140mm DCs for being too noisy at bios minimum.

Luckily the excellent Argus System Monitor lets you set DC fans safely at PWM speed ranges in Windows on launch, so you can override DC limitations in software.

Not sure what % your fans are now, likely 50% minimum or so in bios, but if you ever do want even quieter, Argus is also a good option.
 
Associate
OP
Joined
17 Feb 2009
Posts
869
Only real downside of DC fans is that bios usually won't let them run as slow at minimum than it will with PWM ones.

In my bios DC minimum speed is about 50%, PWM around 25%. I have a silent system at idle with mostly PWM fans and almost had to get rid of 2 x 140mm DCs for being too noisy at bios minimum.

Luckily the excellent Argus System Monitor lets you set DC fans safely at PWM speed ranges in Windows on launch, so you can override DC limitations in software.

Not sure what % your fans are now, likely 50% minimum or so in bios, but if you ever do want even quieter, Argus is also a good option.

I use smart fan mode with a custom curve. Pretty much silent, temps are still excellent, about 31c idle, and 65c under full load ( 3600 )
 
Soldato
Joined
6 Jun 2008
Posts
11,353
Location
Finland
Actually it often is pretty straight, as demonstrated by Martin's Liquid Lab...

Do you ever actually read this stuff, or do you just guess at the TL;DR version?
It's you who doesn't seem to be able to read:
Doesn't need big bend in P-Q curve and fan with both better "free air" airflow and static pressure can be actually worser in real world situation.
 
Soldato
Joined
29 May 2005
Posts
4,290
PWM maintains the voltage and sends pulse waves to modulate motor speeds

DC fan alters the actual voltage to lower fan speed

some fans don’t play well with low voltages.

PWM is the preferred because of that. Personally PWM always.
 
Soldato
Joined
11 Sep 2013
Posts
11,634
It's you who doesn't seem to be able to read:
Doesn't need big bend in P-Q curve and fan with both better "free air" airflow and static pressure can be actually worser in real world situation.
"Can be"... not 'is utter trash and should be derided and ignored at all costs', which is the route you always take.
Both elements are highly variable and should be factored in with all the others accordingly - Even your 'Word Of God' P-Q curve does not account for the variables with far greater impact on the cooling system.
 
Soldato
Joined
6 Jun 2008
Posts
11,353
Location
Finland
"Can be"... not 'is utter trash and should be derided and ignored at all costs', which is the route you always take.
Both elements are highly variable and should be factored in with all the others accordingly - Even your 'Word Of God' P-Q curve does not account for the variables with far greater impact on the cooling system.
Actually definitions of those marketed numbers are strict and don't allow any variability:
Airflow is measured with fan in series with variable air source, which is adjusted so that there's there's zero pressure difference between fan's intake and exhaust.
Meaning fan running along with airflow instead of doing actual work to move it.
And static pressure is how much pressure (above atmosphere) there's in closed box fan is blowing into when airflow has dropped to zero.
Once again number not that great for telling real world situation performance.

Airflow vs. pressure curve is precisely what tells fan's behaviour in real world's variable situation in between those extremes.
That's what Noctua is using to market their NF-A12.
https://noctua.at/en/nf-a12x25-performance-comparison-to-nf-f12-and-nf-s12a
Notice that high pressure fan not actually being such good in comparison...
 
Top Bottom