installing a potentiometer to adjust the speed of a CPU fan

[kq702]

Hi I am trying to make a magnetic stir plate, basically what it does is it spins a strong magnet in circles, and you place a magnetic stir bar in a container filled with liquid and you put the container just above the spinning magnet. What happens is the stir bar spins in circles, which causes the liquid to be stirred nicely (over time).


So I took a DC powered CPU fan and mounted a strong magnet on the top middle area, and wired it to an old AC cell phone charger. The fan specs are 12 volts, 0.15 amps. I tested it with a charger that puts out 5 volts and 0.7 amps but the spinning motion is too fast, it causes the stir bar to get tossed around rather than spinning in circles as it should. Also I have several different sized stir bars, so the speed needs to be adjusted based on how big the containers size, the ammount of fluid, and the size of the stir bar. So I went in my scrap pile and looked for an old potentiometer, and found one mounted to a circuit board so I removed it, and hooked the positive end of the power supply to the potentiometer, and another lead from the other end of the potentiometer to the positive terminal on the fan. Then I hooked up the negative lead from the power supply directly to the fan.

I plug the power supply into the wall and spin the potentiometer dial, and nothing happens. I turn it to the right and get 0 rpms on the fan, then turn it to the left and get 0 rpms on the fan. What am I doing wrong? Or maybe the potentiometer is out of range for this project, But I thought any potentiometer would work, according to my research.

[alm]

Are you using the correct pins? Most pots have three pins, the middle being the wiper. If the pot has a high resistance (you would probably want to be in the tens of Ohms here), then the fan may be unable to start at all but the lowest value. You could also use a silicon diode to drop some voltage. DC fans won't start below a certain voltage, this voltage may very well be close to 5V. You'll need pulse width modulation of a higher voltage signal to further lower the speed.

[Psi]

To use a potentiometer you need to connect the center pin and one of the outside pins. It sounds like you wired up both the outside pins which means you just have a resistor and have bypassed the wiper pin completely.

This isn't the best way to control the speed of a fan, you will probably find that the potentiometer gets hot.
Maybe try a bigger PC fan, like a 10 or 12mm, They spin slower.

[notsob]

Silicon Chip covered a magnetic stirrer in the Dec 2011 issue, you can get the 'fleas'  from amazon, http://www.amazon.com/dp/B007EZ399I/ref=biss_dp_sa1

[alm]

Keep in mind that the rated power of a pot is for the full resistance. A 1 W pot won't be able to take 1 W if you set it to 10% of full scale, then it might only be able to take 0.1 W since power is dissipated over a much smaller surface. A random volume control pot will be designed for much less power, probably less than 0.1 W full scale.

[mariush]

CPU fans, the ones with 4 wires, use PWM to adjust the speed.

Normal fans with 3 wires will accept varying voltages more easily... HOWEVER, they usually don't like to run very slow, so you would be better off using one of those tiny motors usually found inside remote controlled cars.

You can find a ton of them on eBay and in serious stores, for example there's 1700+ here: http://www.digikey.com/product-search/en/motors-solenoids-driver-boards-modules/motors-ac-dc/983125

From 3v to whatever you want, pick something with a reasonable rotation.. adjust voltage to get different speeds... for example this looks nice 12v and 3000 rpm :

http://www.digikey.com/product-detail/en/MXN12FB12F/P14348-ND/2417072

but there's motors from 3$ and lower voltages.

Put something with a circular channel on the motor, and now you can adjust the rotation of the base by changing the diameter of the base of your magnetic thing by adjusting the circle the rubber band is wrapped around... I don't know if I explain it well... see if this helps:



The one on the left is your motor, the one on the right is under the base... imagine a round plastic/hard paper put over the black wheel in that video, and the glass container a few mm above the whole thing.

Your fan will likely start vibrating and eventually die from the imbalance, they're not built well enough to handle heavy things.

[kq702]

I like the idea with using a motor to pull an elestic, I may switch over to something like that when I have more spare time to build it.

Turns out my problem was that the voltage was too low. I put a power supply in that says output 12 volts 0.3 amps and it works now! Also I gave the wrong info in my orig. post, I meant to say I am using a computer fan, not a CPU fan. Thanks for the info everyone!


PS- I tested the voltage with a meter and it says its 17 volts, not 12 as the label says. Is that bad? I am not concerned with burning the motor out, more concerned about a potential fire risk.

[TorqueRanger]

I meant to say I am using a computer fan, not a CPU fan.

What's the difference ?

[Jimmy the Squid]

PS- I tested the voltage with a meter and it says its 17 volts, not 12 as the label says. Is that bad? I am not concerned with burning the motor out, more concerned about a potential fire risk.

Most class 2 wall transformers are unregulated and will put out much higher voltage than the nominal (stated) voltage with no load.  They should put out something close to the nominal voltage when put under the rated load. Not really a problem with something like a computer fan, but could be a problem with a circuit or something with a very small current draw. I, unfortunately learned this the had way.

[Jimmy the Squid]

Also, check the current rating of your fan, they usually have a sticker in the middle of the blades. Many of those that I remember can draw more the the 300mA your power supply is rated for. It probably won't catch fire or anything, but it might contribute to control problems later on as you complete your build.

[kq702]

The fan is rated for 0.15 amps (150 mA) so I dont think I shoud use anything higher then the 300 mA that I am using now.

TorqueRanger- the CPU fan goes real fast, is real small (usually just covers the heatsync) and has 4 wires connected to it. The regular computer fan goes a bit slower, is usually larger then the CPU fan, and only has 2 wires connected. Typically they mount on the back of the computer (toward the middle area, under the power supply) to help with the ventilation.

[T4P]

17V will kill your fan in short order
No doubt the extra voltage will make your fan spin faster

[Simon]

you need to use PWM to control the speed and i think that you need to make sure the fan is powerful enough. Just because you are dragging a magnet around does not let you off from the laws of physics. You still have a load on that fan by way of the magnetic connection between the magnet and the stirrer and that the stirrer needs to be dragged thought the liquid that will oppose a resistance. To ensure the stirrer goes round rather than dither in the middle you probably need the magnet dragging it on a reasonable radius, not just going around on the middle of the fan unless it is a very strong magnet.

[ElectronicTonic]

I made this video about PWM computer fans ( ), then I wondered if anyone on the forum could have used the info in a previous thread. I hope I'm not too late. I experimented with PWMing a regular 2 or 3 wire fan, and it does not work because it is a brushless DC motor with driving electronics that are not designed to be switched on and off at high speed. If you want to finely control the speed of a computer fan, it's bast to use the 4 wire fan that can be found in more recent computers. Then you can control the speed of the fan with Dave's 555 PWM circuit that he recently made for his microscope LED lights. http://www.eevblog.com/2012/11/29/eevblog-392-555-led-pwm-hack/

[T4P]

With PC fans you're not supposed to PWM them if you don't have a PWM signal wire (PWM SIGNAL!)
Just a reduced voltage like a filtered PWM aka a adjustable switchmode will do instead of a LM317

[alm]

What's wrong with driving a regular fan with PWM? The four wire fans with PWM wire allow for a much lower drive current and higher frequency, but nothing prevents you from driving a normal fan with PWM. People have been doing it for years, works just fine. Switching on and off at high speed suggests that the PWM frequency might be too high.

[dr_p]

To ensure the stirrer goes round rather than dither in the middle you probably need the magnet dragging it on a reasonable radius, not just going around on the middle of the fan unless it is a very strong magnet.

both magnets are bar-shaped and mounted in the center. They eventually synchronize.

[ElectronicTonic]

What's wrong with driving a regular fan with PWM? The four wire fans with PWM wire allow for a much lower drive current and higher frequency, but nothing prevents you from driving a normal fan with PWM. People have been doing it for years, works just fine. Switching on and off at high speed suggests that the PWM frequency might be too high.

I tried two different fans, big and small. PWMing the 12V power may work if you're lucky, but I wouldn't trust it. These fans are not driven by simple DC motors with commutators that certainly can be PWM driven. They are brushless DC, meaning each fan has a couple transistors, passives, and hall effect sensors working together to drive the coils which propel the spinning magnet cylinder inside the fan blade. That whole circuit was designed to be powered constantly. Even if I gave it a low frequency PWM of 10 to 100 Hz, the circuit just locked up and never got the fan blades spinning unless I gave them an initial push. If it was easy to PWM a two- or three-wire brushless DC fan motor, then computer designers would have been doing it for years.

[alm]

Computer designer have done it for years. Before the advent of the four pin PWM-aware fans, this was a common way of regulating fan speed. Even some motherboards claiming to vary the voltage used a 12V PWM signal. PWM is much cheaper than voltage control  and allows a much wider range of speeds (voltage control is limited by torque required to overcome inertia). Fans would often rattle at low speeds, probably due to the low duty cycle and PWM frequency. A dedicated PWM wire is nicer of course.

Some fans 'liked' PWM more than others, which I'm sure comes down to the drive electronics. But it was hardly a lucky shot only used by a few electronics hobbyists.

[SeanB]

I had a few of those boards, i just turned the fans on to full, or used the switch function to have them run only if the CPU was over 50C.

[PhynixViking]

I don't know if anyone mentioned this before... but... a lot of computer fans have circuits in them that make them regulate their speed. So even at a slightly lower voltage they may still be very fast.
Today's most modern computer fans are 4 pins and use PWM.
How you can get around this... instead of using direct resistance to your fan, use a 555 timer and make it pulse the voltage slower/faster with out a capacitor between the load (fan) and the 555's output.
This SHOULD work fine.
You can use a 555 timer with a resistor or a capacitor to control its pulse/oscillation frequency.
So in your case place a small resistor (idk maybe 100 ohm? 200 ohm? what ever you want the "fastest spin" to be) and THEN add the ped. to the resistor (so when the ped. is at "0" it still has a resistance to the 555 timer.)
Recently there was a video on the EEV youtube that explained a 555 timer being used in this same way for his microscopes light.
Hopefully this works... I know its probably not the cheapest way but I think might be the best way.

ALSO: from my experience some fans are more complex then others.
I have a good maybe 1000 fans at the office, if you want me to, I will actually set up a bunch of them to a ped. and see if any brands/models actually end up slowing down (a lot) with just a ped. and no 555 timer. Let me know if you want me to do this (as right now I'm not busy but might be any day now, lol.)

Good luck! ^___^