Mosfet for variable power?

Hello.

I have a project in which I want to run a fan vary between 2 speeds depending on conditions.
The fan is PWM enabled and I have successfully used that to control the fan.
Initially in my experimentation, I was trying to play with a MOSFET to deliver the voltage to the fan.
The fan is rated at 12volts.
What was never clear to me, is it possible to use the MOSFET to vary the voltage or is the MOSFET simply an "on/off" switch?
Could you control the MOSFET with the Arduino PWM signal? Using that 0-255 setting to open the MOSFET gate "just a little", to vary the voltage passing through it?

Thanks,
Gary

Using your favorite internt search engine look up "arduino circuit mosfet pwm fan."

What was never clear to me, is it possible to use the MOSFET to vary the voltage or is the MOSFET simply an "on/off" switch?

Yes, it's possible to use MOSFET as a "linear" or "analog" device. But, it's inefficient.

Could you control the MOSFET with the Arduino PWM signal? Using that 0-255 setting to open the MOSFET gate "just a little", to vary the voltage passing through it?

No, not directly with PWM. PWM is digital. A [u]driver circuit like this[/u] "amplifies" the PWM to control motor speed (or the brightness of a lamp).

But, you CAN filter the PWM to variable DC and amplify that. (It's a terrible way to control motor speed.)

A MOSFET (or transistor) by itself does not make a good linear amplifier. It requires more components and a more complicated circuit to make something like an audio amplifier. (Typically, audio amplifiers can't put-out enough power to drive a motor and of course that's not what they are designed for.)

When used in linear mode, the MOSFET/transistor has voltage across it an current through it at the same time. That means it has to dissipate power and heat (Watts = Voltage X Current).

If you use the wrong kind of MOSFET or if you have a bad design you can end-up accidently turning the MOSFET partly-on, and that often overheats it and kills it!

When uses as a switch it either has voltage across it OR current through it at any moment in time so it (theoretically) doesn't have to dissipate heat and it doesn't waste power. (In reality, there is some resistance so there is some voltage across it when it's "on" and passing current, but it won't get nearly as hot and it can handle much more current than in linear mode.)

You’re not making very much sense. If you are already controlling the fan successfully with PWM why do you want to vary the voltage to the fan?

What will that achieve that simply changing the PWM duty cycle doesn’t already give you?

Steve

DVDdoug:
Yes, it’s possible to use MOSFET as a “linear” or “analog” device. But, it’s inefficient.(snip)

THANKS doug! Bookmarked this. Lots to chew on there.
“Someday” I’d like to spend some time just playing with mosfets to see what I can do with them.

slipstick:
You're not making very much sense. If you are already controlling the fan successfully with PWM why do you want to vary the voltage to the fan?

Steve

SCIENCE!!! :slight_smile:

Sometimes I have to just get a project DONE. Other times, like here, I like to use it as a learning experience. I've learned that often, there is more than one way to approach a problem. In this case it looks like the MOSFET is a bad way to go. However, until I looked into it and then asked, I didn't know this. I now have a better understanding of these devices and am already thinking of ways to use them instead of relays in future projects.

BTW - Idahowalker. Thanks for the search suggestion. Previous to my question, I had actually spent several days studying some videos on using MOSFET and PWM to control computer fans and had attempted to replicate their circuits but in the end, they left me with more questions that could only be answered by asking and I wasn't getting a response from the video's authors.
I will try your search phrase suggestion though, perhaps using your search terms will produce different info than my searches did.

yea you can use Mosfet but you can easily use control the speed of fans by using simple variable resistance

Jack Arsal

P.S.
Linear voltage regulators use a transistor or MOSFET in linear mode.

Switching voltage regulators, DC-DC converters, and Class-D audio amplifiers use a kind-of PWM in a feedback loop along with an inductor as part of a filter to store energy. The result is "smooth" analog output from an efficient switching design.

These switching circuits are pretty complicated but with modern electronics almost all of the circuitry (including the MOSFET/transistor) is built-into a chip so you just have to add the inductor, a couple of capacitors, and maybe some resistors.

DC-DC converters and switching regulators are almost 100% efficient so they run relatively cool, and with a step-down voltage regulator you can get more current-out than you feed-in (while getting less voltage-out than you feed-in). With linear regulators you have to feed-in slightly-more current that you get out.

These are not necessary with a motor controller. The inertia of the motor makes it respond to the average voltage so PWM works fine. The same is true with regular incandescent light bulbs. They "glow" with a brightness related to the average voltage/power.

LEDs are a little different because they respond instantly but your eye "averages-out" the brightness, so PWM works for dimming LEDs... i.e. If you "flash" an LED for 1/100th of second it will appear to be dim (if you can see it at all) even though it is flashing at full-brightness.

yea you can use Mosfet but you can easily use control the speed of fans by using simple variable resistance

Very inefficient, and depending on the power of the fan you might not find a potentiometer that can handle the current without burning-up.

Very inefficient, and depending on the power of the fan you might not find a potentiometer that can handle the current without burning-up.
[/quote]

appreciated it may case some power loss to but its the simplest way to get our goal, Moreover your point is also acute and original. Mosfet is also being used as potentiometer.

If the fan has a brushless motor its controller may not work well when you PWM the power supply.

Ah, yes! This is a common misunderstanding: lower voltage to a fan
makes it run slower. But, when you add a resistor in series with a
fan, you starve it from drawing the current it needs to perform to
specifications. (Think: I^2 * R = Watts = Power) It does run slower
but it will have much less power to move the air!
(The resistor does what it does: it resists the flow of current.)
Herb

Gary_Arduino:
The fan is PWM enabled and I have successfully used that to control the fan.

You forgot to give a link to the data sheet of the actual fan you're using.

I'm guessing you have a three wire fan: power, ground and signal (PMW). Those things are normally built with brushless motors, which means they have control electronics built in, and change the fan speed based on the PWM signal. A change of voltage is likely not having much if any effect on them (as long as it's high enough for the electronics to work and for the motor to run at all).

Voltage control is something for DC motors, which indeed will change speed based on voltage.

wvmarle:
You forgot to give a link to the data sheet of the actual fan you're using.

I'm guessing you have a three wire fan: power, ground and signal (PMW).

Sorry about that. http://www.nidecpg.com/fanpdfs/ta450dc7.pdf
It is the Model#B35502-35
4 wire actually. If I were smart enough, I could use that 4th wire to monitor the speed. While I don't know how to do that and it isn't really pertinent to my project, it would be a cool thing to know.

DVDdoug - I didn't know that was how LED dimming worked. Amazing!