Calcuationg Fan Speeds

N

NeverBackDown

AMD Enthusiast
Hey everyone i just had a weird math moment with fan speeds... I'm only interested as i will be running watercooling in the upcoming months at low volts with the fans.

Anyone care to correct me if i'm wrong here: but this is the math i thought would be correct on how to find rpms on voltage reducers.

Corsair Sp120 HP: 2350rpm/12volts = 195.83rpm

195.83rpm * 5volts= 979.15rpm.

Corsair Sp120 QE: 1450rpm/12volts = 120.83rpm

120.83rpm * 7volts= 845.81rpm

I picked popular fans but would this math be correct? I used different voltage to show the difference between what fans speeds would be run at different volts.
I know there's more electric laws than this simple quick one i thought of. Obviously ohms and resistance(electric terms if you didn't know) and can affect the outcome but i'm assuming those are equal to less complicate the math.

Any advice on how to calculate the fan speeds? I did try researching quickly too see if there is a set rule but I could not find one. Thanks^_^
 
Last edited:
NeverBackDown said:
Could you explain that? I'm not sure i'm understanding you correctly.
Click to expand...

means that so long as RPM scales linearly with volts E.G 12v=2000, 6V=1000 then your calculations would be correct.


made a graph as a rough guide, the fans have an operating range of 7 - 12 volts so anything under 7 V is useless as the fan wont even start however, as I don't know the RPM for 7 V I had to do the whole scale from 0 - 12 volts.

graphkn.png
 
Last edited:
To put what the others have said in mathematical terms, if you can write the fanspeed as:

Fanspeed = K * Voltage across the Fan (where K is some number that doesn't change)

Then what you've done is correct.

Technically it's the current through the fan that determines the fan speed not the voltage and the relationship between the fan speed and the current isn't strictly linear. But for simplicity's sake that's not something to worry about. :p
 
yes k refers to the voltage you find when dividing 2000rpm/12v which gives you 167v. in this case, k is 167v.

I think it should work with all fans but it varies depending on the fans, example if you have fan1 running at 5v it might run perfectly but fan2 has much higher resistance on the blades and does not manage to spin at 5v. I think its all about trial and error until youre happy with the results
 
So for example, you could just make a graph with the volts/rpms and start with 12/2000 and draw the start linear line to 0/0 and for that fan you can find any rpm at any volt? Or would that still not work?
 
Volts, HP,QE
7,1370.83,845.83
8,1566.67,966.67
9,1762.50,1087.50
10,1958.33,1208.33
11,2154.17,1329.17
12,2350.00,1450.00


like this?
 
Maths goes right over my head but i don't think this can be right.

Fans don't tend to run at low voltages so you canni divide by 12?
As the fans spin quicker the air resistance ect grows and the motor will need more "umpf" to power it. Not to mention the varying tolerances between even the same make and model of fan.

Im sure if i spent more time i could pick even more holes.
Or am i missing something completely here?
 
Last edited:
maniac said:
In theory it would work :)
Click to expand...

I'm not so sure...
Been a long time since I did Physics...

But,
Considering it's less and less power being provided in order to shift the same mass (fan blades) it won't be linear. Also, there's the fact that once the blades have their own momentum, they are able to spin by themselves - ie - they have their own momentum too, which is increased more and more as the voltage increases.
It won't be completely linear, but I have no idea about the maths behind it in order to establish the exact shape of the curve


Also, fans have starting voltages too. Noctua NF-F12s won't even start at 5v, so it's only really measurable down to that.
 
If you're into your cars, think of it in terms of the Bugatti Veyron supersport. They had to add 200bhp to increase the top speed by just 8mph (Or something like that, I don't remember exactly).

What this means is that the faster the fan is spinning, the more volts it needs to go that little bit faster. This results in a graph that is exponential, meaning it looks like one side of a quadratic graph. This is all likely to air resistance increasing as you increase the RPM - the fan has to do more work against the air resistance and therefore needs more volts/current for a lesser increase in RPM.

I'm not great at explaining things, but I hope I helped...
 
Well yes but like maniac said in theory it's correct. The math still isn't correct so I don't know how to fix it:p

Edit: Didn't see the second page lol
So in other words there are too many variables to be able to calculate the rpms at X volts? Then don't forget to add different places the fans are located in a case with different objects around it... I'm assuming we just have to by noise?
 
Last edited:
NeverBackDown said:
Well yes but like maniac said in theory it's correct. The math still isn't correct so I don't know how to fix it:p

Edit: Didn't see the second page lol
So in other words there are too many variables to be able to calculate the rpms at X volts? Then don't forget to add different places the fans are located in a case with different objects around it... I'm assuming we just have to by noise?
Click to expand...

You can still definitely use your way as a rough estimate, but I'd honestly just start a new thread about ultra silent fans, I'm sure there's some silence freaks around here who would show you exactly how/which fans to setup.

Chris230291 said:
HAHA no way! Thats EXACTLY what i was thinking!
Click to expand...

LOL, Top gear FTW :D
 
iBeInspire said:
You can still definitely use your way as a rough estimate, but I'd honestly just start a new thread about ultra silent fans, I'm sure there's some silence freaks around here who would show you exactly how/which fans to setup.
Click to expand...

Why would i do that when this is a thread about the same thing?:p
My way is a rough estimate but not a very accurate rough estimate...

Does anyone know a semi accurate way of finding out anyone?
 
Manufactures data sheet?
I would have thought that they each have there own graphs that show the rough noise generated at any given voltage? I know i've seen something like this before.

EDIT: After a quick google image search i found loads of graphs. Mostly noise vs RPM or noise vs CFM though.
 
Last edited:
NeverBackDown said:
Why would i do that when this is a thread about the same thing?:p
My way is a rough estimate but not a very accurate rough estimate...

Does anyone know a semi accurate way of finding out anyone?
Click to expand...

What are you actually wanting to know?
Which fans to buy? and what speed to run them at?
 
You must log in or register to reply here.

Similar threads

W
Need your advice - 280 AIO in the top or 360 in the front?
Replies
17
Views
7K
Warchild
W
Excalabur50
Best Case for 360 top mount push pull?
Replies
11
Views
6K
Dark NighT
Dark NighT
deigima
Mobo/cpu update and cooler advice
Replies
1
Views
5K
Excalabur50
Excalabur50
T
Need help with Fan Headers please !!!
Replies
2
Views
3K
Terry1978
T
trawetSluaP
Case fan % seems off
Replies
17
Views
7K
Warchild
W
Back
Top