How to lower voltage without wasting power?

[Prozzaks]

Here is the deal : For my new custum designed and built from scrach case, I bought 5 Panaflo FBA12G12L. Unfortunately, I found out that they sometime fails to spin up with only 5V, and at 12V they make way too much noise( mostly switching low frequency switching noise, some whooshing and almost no bearing noise ).

Using a regulated DC power supply, I found out ( Oh Surprise, surprise! ) that the ideal voltage is 7V.

Here are the fan specs ( According to the spec sheet ) :
Model : FBA12G12L
Operating voltage : 7 to 13.8V
Current : 0.170 A
Input Power : 2.04 W
Rated Speed : 1700 RPM
Max Air Flow : 68,9 CFM
Noise : 30.0 dB-A

However, the specs written on the fan themselves are not identical to the one given in the spec sheet. So here they are :
Model : FBA12G12L
Voltage : 12V
Current : 0.24 A

So basically the specs written on the fans suggest that they draw more power. So to make sure I can provide sufficient power I calculated it from the specs written on the fan using the following formulae

P = VI

P : Power in Watts
V : Voltage in Volts ( Duh! )
I : Current in Amps

P = 12V * 0.24A
P = 2.88 W

Therefore, each fan draws 2.88W. However, I intend to use them at 7V.

The way I understand things, the current is constant no matter the votage. Is that right? Therefore if I want to drop the voltage from 12V to 7V I have to "loose" 5V. If I use resistors, I can figure out there resistance from the required voltage. Given the formulae :

V = RI

V : Voltage in Volts
R : Resistance in Ohms
I : Current in Amps

V / I = R
5V / 0.24 = R
R = ~20.8 Ohms

Now, I must get resistors that are powerfull enough to handle the current. Using the first formulae I get :
P = VI
P = 5V * 0.24A
P = 1,2W

So, for each fan I must get a resistor of at leat 1,2W. According to my research resistors go from 1W straight to 5W.

Now for the actual question :

If my understanding is correct, then I'm basically wasting 1.2W ( 5 resistors * ( 5V * 0.24A ) ) in the resistors as pure heat. Therefore, is there a way to lower the volatge without wasting power? I know that 6W is not much, but according to what I know it's entirely dissipated as heat. Does anybody know how to lower the voltage without wasting power? I heard something about using diodes, resistors, and capacitors, to make circuits that would provide the full 12V at power-up and than gradually lower the voltage to the desired target. Do such ciruits waste power as heat to lower the voltage?

All help is greatly appreciated!

[Devonavar]

Are you unable or unwilling to use the 7 volt trick?

[jaganath]

If my understanding is correct, then I'm basically wasting 8.4W ( 5 fans * ( 7V * 0.24A ) ) in the resistors as pure heat.

1) You're probably wasting way more than that from your PSU.

Example: computer power draw 150W.
PSU electrical efficiency: 80%
Wasted heat energy: 30W

2) Why not use a Zalman Fanmate 2? I have mine at 5V and it's cold to the touch.

[Staats]

You could always build a PWM type fan regulator. If you do not want to waste power, it is the way to go. It is more complex than resistors or diodes, but it is mor efficient. I cannot verify this, but I have read that at lower voltages fans actually draw more current that at higher voltages. With a PWM, you are actually sending pulses of 12v.

[Le_Gritche]

However, the specs written on the fan themselves are not identical to the one given in the spec sheet. So here they are :
Model : FBA12G12L
Voltage : 12V
Current : 0.24 A

You are right to trust these specs, because they give you the current associated with a given voltage (12V) rather than with a wide array.

Unfortunately your following calculations are wrong because the current is NOT constant when the voltage changes.
A first (and rough) guess is that the fan motor act as a resistor with a constant value, no matter what the voltage may be. It's wrong obviously, but will give a decent enough approximation.

So U = Rfan * I gives Rfan= 12/0.24 = 50 ohm
You can then do the math with your fan acting as a 50 ohms resistor.

Then when you add a resistor, you get a total of R+Rfan under 12V.
The current accros the line becomes I = 12/(R+Rfan) and the voltage used by the fan is then U = Rfan*I = Rfan*12/ (R+Rfan)
you extract R (because you are looking for the value of the resistor to add)
U*(R+Rfan)=12*Rfan
R=Rfan*(12-U)/U

With Rfan=50 and U= 7v you get R=36 ohm

As I said, the fan doesn't act as a resistor, as its resistance drops when it gets less voltage (it spins slower so waste less relatively). So 36 ohms would probably give a little less than 7v.
Don't worry : resistor values are not equalliy available, you have 'common' values easier to find. For 36 ohm I think the closest value is 33 ohms and being a little bit smaller it will balance the aforementioned effect.

Now for the power rating, P=R*I², I=12/(50+33) that is P=0.69W
You can get a 1W resistor, but you can also find 2W or 3W rated resistors, the main difference being that they are bigger (to be able to dissipate more heat).
A 2W (or more) rated resistor will then be a little less warm to the touch, than a 1W rated resistor.

For your 5 fans that will amount to about 3.5W dissipated. You can use your PC case as a giant heatsink by holding the resistors on it with tape, if you want to avoid the heat being dissipated inside the case, but that's a bit overkill I would say.

Does anybody know how to lower the voltage without wasting power? I heard something about using diodes, resistors, and capacitors, to make circuits that would provide the full 12V at power-up and than gradually lower the voltage to the desired target. Do such ciruits waste power as heat to lower the voltage?

You can use the 7 volt trick as mentionned above, that will avoid wasted heat but it's a bit tricky for the PSU because you are then using the 5v line as a ground for the 12v line... Carefully read the link of Devonavar, because five 120mm fans seems a lot of power being fed back at the 5v line.
If you want advice on Zener diodes, capacitors and such, there's the link posted by MikeC in his article about the PC for Thailand, about the 7-volt trick and fixed voltages.
As you live in Canada, you might speak french, so I will add this link : le tomik des montages électriques pour ventilation

To conclude I will say that given the cheap price of resistor, you might as well try with different values (20, 25, 30...) and see what suits your needs.
Remember that a resistor is not flexible, contrary to a potentiometer, so beware of what will happen this summer with higher ambiant temps if you don't have AC at home.

[Le_Gritche]

I just got another idea ; since your fans are of the same model, you might as well use another trick which is simpler.
You wire 2 fans in line (I don't know the proper translation for 'en série').
The 12v get in the first fan, but when the wire get out of the fan, you wire it on the second fan rather than on the ground. You get that :
12v ----- FAN1 ----- FAN2 ---- GROUND
rather than
12v ----- FAN1 ---- GROUND
12v ----- FAN2 ---- GROUND

As your fans are identical they will equally share the 12v and get only 6v each.
I don't know if that will be enough to start them reliably, but if it works there will be no wasted heat, no required additional component and just some basic wiring action. So it's worth a try.

[Elixer]

Le_Gritche, your math is good, but you can't base the resistance of the fan based upon its current rating. What is generally listed on fans is the max current draw. You probably won't ever see the fan draw this much power unless you stall the fan at 13.2 volts. Also, fans have wide tolerances; I wouldn't be suprised to see a 10% resistance difference between apparently identical fans. The fan will need to be measured. The best option is to measure the current draw when the fan is given 7 volts and calculate the resistance based upon that current and voltage.

Or you could just avoid all this non-sense and use the 7 volt trick. Many SPCRers, including myself, have done this successfully with no problems.

[Prozzaks]

I had already considered the 7V trick but, I felt that 6W was a little too much. I rather do things "by the book" ( With a soldering iron and cool hacks )!

I searched on Digikey and I found variable rsistances @ 5W. I'm probably going to use thoses.

Given the fact that I only have 3 different airflow tunnels, I could use only 3 resistors insted of 5. There also have 20 Ohms fixed resistors that are a lot less expensive ( 0.56$ vs 4.48$ ).

[Le_Gritche]

you can't base the resistance of the fan based upon its current rating. What is generally listed on fans is the max current draw. You probably won't ever see the fan draw this much power unless you stall the fan at 13.2 volts.

I guess you are right on that.
About the 10% tolerance, I'm not convinced it would be a problem. A 10% variance seems to me huge manufacturingly-wise and yet it would only led to something like 6.3v for a fan and 5.7v for the other.
But in the end you are right : after rereading his post, I noticed he wants to run the fans at 7 Volts, not at the lower possible voltage starting the fans.
I have absolutely no experience with the 7 volt-trick so I can only rely on MikeC's advice in his article :

Since usual fans only use 1...3W, you are usually perfectly safe with using the 7V trick there. There a lots of devices in typical computers, which draw much more power out. Peltiers or things with power consumptions running up to dozens of watts, require careful calculation of what goes out of the 5V lead and what's going in there from the 12V lead.

He has five 120mm fans rated for 2.88W each (though at 12v) so I thought it might be in the area 'requiring careful calculation'.

@Prozzaks
5W rated components are probably expensive when you could use 2W or even 1W rated ones, with the airflow in a PC case that will be enough to cool them down enough. I think it's the kind of passive component bearing 90°C like a breeze.
Potentiometers allow you to fine tune the fanspeed according to ambiant temp and idle/load state of your computer, but they require a more complex setup. You can't just let them holding on the wire, you have to tie them somewhere, preferably so that you can access them without opening the case.
Also some potentiometers include a switch, so it might be of use if you plan to run fanless sometimes.
What do you mean with '3 different airflow tunnels' ? You don't plan to run a fan at 5v ina tunnel with another running at 7v, do you ?
Your links point to an online shop I think, maybe you could go to a local electronic shop, you would avoid the shippng fee and get useful advice.

[cloneman]

From what I have read on SPCR, it is my understanding that there is zero risk with the 7V trick as long as you draw the same amount of watts on the 5V line with another load. If that is the case, why not add lighting or something else @ 5V that doesn't add to heat?

Or am I wrong about having zero risk when properly doing 7V?

[justblair]

I've tried the 7v trick on a 5v line that had no other devices attached. Had no unpleasant side effects. I'm guessing that the 5v line feeding the mobo is the same line feeding the molexs?

I just know someone is going to post a link to the PSU essentials sticky... I plain and simple have a mental block when I read that post!!