PWM Control for Asus M5A97 LE R2.0 and Ausus M5A97 R2.0 MBs

[andrew124C41]

I have a number of these motherboards.

I am new to building PCs.  I made a mistake purchasing a bunch of 120 mm coolermaster led 2000 rpm 3 pin fans. 

I want to use 4 of them for intake and 3 for exhaust.  two of the exhausts are horizontal which I have heard is not a good idea for these fans..but that is ok.

The problem is that the PC sounds like it is going to take off!

The MBs had 3 case fan PWM controled 4 pin headers.  They cannot be controlled by voltage.

Each header cannot handle more than 1 Amp.  The fans use .35 amps each so that putting them in parallel would use more than that amount.

I have looked at some circuits here and elsewhere.

It would be nice to be able to control the two fan sets using either the EUIF (BIOS) of the MBs or FanExpert but it would mean taking power from the PSU and using the tach lead (not forgetting that 3 and 4 fans will be in parallel) and putting that input in the fan headers...but I don't know if that will work.  I am refering to circuit images 3a and 3b.

If that won't work, perhaps circuit 1 or 2 will.

Any ideas would be greatly appreciated.  I want to keep it simple.

I have seen inexpenisve, less than 10 dollar PWM controllers on ebay and amazon as well.

[kam]

Ok... first, chances are that your mobo cannot control a 3-pin fan. you can only get rpm feedback from the fan. you need a 4pin fan to control if from the mobo.

If you use a pwm controller for a 3-pin fan, then you will not have rpm feedback
If you use a linear fan controller, you have rpm feedback but the controller will dissipate a lot of heat.
A basic pwm controller is here: http://www.pcbheaven.com/circuitpages/High_Frequency_PWM_Fan_Controller/
A basic linear controller is here: http://www.pcbheaven.com/circuitpages/Simple_Linear_Fan_Controller/

[andrew124C41]

Thanks for getting back to me.
I understand that my MBs fan headers only work for PWM controlled fans.
The schematic and picture of the connections...that is my first two attacments were taken from a post that preported to work for my situation...that is since the FET is missing from my 3pin fans, that by using the circuit and connecting the tach to the MB header, and the PWM from the header to both the juncture of R1 and 1 (also connectedd to the MOSFET although I am not sure how) this would allow my MB and software to control my fans.

Here is the link to where I got the circit from:http://www.overclock.net/t/1505404/guide-convert-3-pin-fan-to-pwm-56k-warning-courtesy-of-lazzer408

So, I am wondering if it will work.

If it won't, how about one of the other two circuits for instance the one with only two componants...an NPN power tansister and a 100 K pot

My fans are in sets of 4 and 3
Each fan draws .35 amps  4.2 watts at 12 volts

The only other thing I can think to do is use a resister.

However, the sets of 4 and 3 in parallel will draw more current with less resistance if my thinking is correct.

I don't know where to find a resister that can handle the power and for sets of 4 and 3 what resistance would I use?

[kam]

yes this circuit will work, it is missing the pwm generator though. the setup is the same as the pwm circuit that i sent to you.

as for the resistors, you can use multiple resistors instead. You can have one resistor for each of your fans, and then connect them in parallel. but this method does not provide any control. only fixed lower rpm

[andrew124C41]

Before I build anything, this is what I am concerned about.
These fans spin at 2,000 rpm

1) with PWM do you think I could get under 1,000 rpm

2) How about using the power transistor with the 100 K pot....will that work?

3) If you were to guess, based upon the values for the fans, what resistor woul I use for a parallel circuit for 4 and 3 fans respectively....that is...one resistor per set on the positive leg?

[kam]

1. definitely
2. well, maybe not. Linear control has problems at low speeds because the torque is decreased significantly
3. I'd start with say 20 Ohms. You should try anything from 5 to 80 Ohms though

[andrew124C41]

The tans are about 4 watts each.
For each set 4 and 3...how about the wattage of the resisters.   Finding 5 watt resisters is not easy unless I get them online.

I might try the power transistor...just to see if it works

[kam]

The power transistor will work definitely.

As for the power, say that you want to drop like 6 volts on the resistor (so to provide half the voltage), the power on the resistor will be
6x6/R

So, if the resistor is 20 ohms, the power will be

6x6/20 = 1.8 Watts

[andrew124C41]

is that for 6 fans at a time...is that how you got 6 x 6?

2 watt resisters are not that common here unless you purchase them online.

[kam]

the formula is

P = V² / R

where V is the voltage drop across the resistor and R the resistance. So, for 6 volts drop on an 20 ohm resistor:

P = 6² x 20 = 6 x 6 / 20

[andrew124C41]

OK, I have been experimenting with my fans.  I find that I can use a 10K pot rated at .5 watt to bring the fans to a very, very low speed at 3 volts.  So, the voltage drop across the resister is around 9 volts.  Since it is such a high resistance, the wattage is very low, below the .5 watt rating.  However, the transistor I am using, a TIP 3055 rated at `10A with a 75 W dissipation gets very hot.  I measured it with an infrared gun and it is 126 C.  The max is 150 C.  Do I need to heat sink it?  Another issue is that the case is connected to the collector at 12 V...so, I don't know how to prevent the sink from having current on it.  It has been running for a couple of hours this way without a heat sink.

One thing though, if I turn the pot all the way down, the fans will go off.  I am going to try a 4K pot and see what happens.  I may have to use that with a fixed resistor, probably around 1, 2, or 3 K.  I just will have to experiment.  What is your guess?

[kam]

Ok, you have confused something here with the power dissipation. You use the pot to make a voltage level to transfer it on the transistor, amplify the current there and feed it to the fan. The transistor alone is the one that is called to dissipate (as heat) all the power left, and this is why it gets so hot at low speed. You definitely want to use a heatsink. Otherwise it will fail in the long run.
Also, yes you need a fixed resistor to maintain a minimum voltage level when the pot is down to 0.
Also, linear controllers tend to stall the fan at low speed. My guess is that at 3v, if you block the blades with your hand and then let thmem go, they will not start. Make sure that at minimum speed the blades start again even if you block them with your hand.

If you want to totally eliminate the extra heat from the transistor, go ahead and test the pwm controller.