PWM to control brushless 12VDC fan speed

[mcgyvr]

Looking for some help to control the speed of a brushless 12VDC (120mA) fan from a 0-5V 500Hz PWM signal.
Here is the fan.

http://www.newark.com/sunon/kde1204pkvx-ms-af-gn/axial-fan/dp/89K1822

I tried simply sending the PWM through a 220ohm resistor to the base of a transistor and 12vdc through the red wire, black wire to the collector and emitter to ground. It worked fine when at 100% duty cycle but anything lower and the fan just stalls and buzzes.

 

[VEBill]

Does the fan itself contain electronics to 'replace' the brushes?

 

[mcgyvr]

Yes I believe so which is why it is not simply working with the transistor switching. Probably a hall sensor,etc.. to switch coils.

 

[VEBill]

I had a quick look at the .pdf spec sheet. It doesn't seem to be the sort of fan where the speed can be varied.

 

[mcgyvr]

So I guess I need to go with a 4 wire fan? red,black,pwm control,tach or something?
I was hoping one of you smart guys might have a solution.

 

[mcgyvr]

I found this schematic but wasn't really happy that the mosfet would generate a decent amount of heat.

 

[IRstuff]

Doesn't the datasheet say that it's a 12 volt DC fan? Pages 5-7 seems to make it pretty clear that you apply a DC voltage and it runs. Page 4 seems to imply that there is no expectation of adjusting the speed, otherwise, there'd be a family of curves.

Nonetheless, you might see some speed difference between 4.5 VDC and 12 VDC.

TTFN

FAQ731-376

 

[mcgyvr]

Yes there is a huge difference between running it at 4.5 and 12. I thought that I could simply pulse it on/off to vary speed. (very common with older brush fans,etc.. to vary speed) Then come to find out this doesn't work with newer brushless fans because of the internal hall sensor,etc... I just thought someone might have a circuit to perform the function I needed. Guess I'll just source another fan with speed control.

 

[VEBill]

You could PWM a suitably large value capacitor, and create 4.5 to 12 volts (relatively) smooth DC. Probably not worth the effort though.

 

[IRstuff]

Why do you need to do PWM? Why not just a programmable DC voltage?

You cannot PWM any DC brushless motor, newer or older. It's already doing pulsing internally; external pulsing just screws up the control circuitry.

TTFN

FAQ731-376

 

[VEBill]

Even at 12 volts, it's only using 1.4 watts. That'd be even lower if it slows down at lower voltages. On the assumption it's being used to cool electronics that requires a fan (medium to high power), then probably not worth worrying over the fan's ~1% contribution.

There are plenty of fans and fan controllers sold for PCs. But most would use more than 1.4 watts.

But obviously I don't know what you're up to...

 

[IRstuff]

Back to your posting of 29 Jun 10 16:09:

No, you need a naked brushless motor if you want to do your own pulsing, which means a minimum of 3 windings, or phases. Anything that's described as a "DC" brushless motor has an internal controller.

TTFN

FAQ731-376

 

[Henneshoe]

To expand on VE1BLL's idea.

I assume during the dead-time of your PWM the fan voltage drops below 4.5V. This causes the integrated brushless DC motor controller to turn off. During the on-time, the motor controller powers up again, but not long enough to get the fan spinning. You could add a capacitor in parallel with the fan to keep the fan voltage above 4.5V during the dead-time. The capacitor would charge almost instantly through the transistor, and would slowly discharge through the fan. For an input voltage of 12V and a minimum output of 4.5 volts, the RC time constant of the fan/capacitor should roughly equal the PWM period.

The apparent fan resistance of the fan motor / BLDC motor controller is 100 ohms (.12A @ 12V). For a 500 hz PWM rate, the capacitor should be approximately 20uF. This value could be reduced by increasing the PWM frequency.

 

[mcgyvr]

Thanks guys. I've changed directions to get around this.
I did try a cap across the fan but I didn't use one large enough (didn't run the math..just threw a 1uF cap out of my bin on it). Might find a 20uF and try that again one more time before I switch directions as below.

Now I'm planning on switching the fan to be fed from an LM317. LM317 output will be digitally set to have 3 different output voltages. 2 digital outs from my micro feeding the base of a transistor to drop in a parallel resistor to vary the LM317 output. Like figure 11 of this datasheet

http://www.st.com/stonline/books/pdf/docs/2154.pdf

I really only wanted/need 3 fan speeds anyways and not completely variable.
Something like..
(both digitals off=12vdc from LM317) (dig1=high, dig2=low =8vdc from LM317) (dig1=high, dig2=high =6vdc from LM317)
That should get the job done.. It's a side project for home anyways.

 

[IRstuff]

Do make sure that your 317 is adequately heatsunk, particularly if the input side is a relatively high voltage. If you're starting with, say, 20VDC and running at 8V out at 0.1A, that's a watt lost in the 317, and if it were unsunk, it would get a bit toasty.

TTFN

FAQ731-376

 

[mcgyvr]

yes it has a decent sink on it. thanks for the reminder though.

 

[thruthefence]

While this doesn't really speak to the OP's control voltage issues; I'll share my 'bodge';

I have a large Quadraphonic receiver/amp (remember those?)in my machine shop that someone gave me,and the internal fan shot craps.So, I have a 12V pancake fan sitting on top of the amp, powered by a 12V 1.0A "wall wart" transformer, which is in turn plugged into a variac. By varying AC to the transformer, I can reduce the fan speed (and noise), but still keep the amp from overheating.

not sure where this places my 'carbon footprint'.