So the mosfets are voltage controlled devices right? If I have a logic level mosfet (1-2V Gths) then with 3.3V or 5V I should be able to turn it on. BUT how about current? I am running on batteries so I dont wanna waste current. Nobody nowhere (I searched trough the datasheet but found nothing) talks about the amps needed to turn ON a mosfet. SO If I connect 1Mohm resistor between the gate and the Arduino pin, then would 5/1000000= 5uA be enough to turn it ON? If not, then please tell me when am I wrong and explain me what should I do.
I'll give karma points to the 3 most helpful comments 'cuz this is important for me.
The MOSFET gate can be modeled as a capacitor between the gate and the source. That is, essentially no current flows to or from the gate except when it is being switched on or off - and when you're switching it, you generally want the peak current to be as high as whatever is driving it can supply (meaning that the pulse of current is shorter) in order to minimize the time that the MOSFET spends between the on and off states (as during that half-on state, it presents a higher resistance between the drain and source, so some current will flow, but the high resistance of the MOSFET will cause it to heat up).
So - as mentioned above, you want to limit the current into or out of the gate, but only to the maximum that the device driving it with can safely supply - you won't save power by using a higher value resistor, and it will inhibit switching. It's common to put a 100~220 ohm resistor between the output pin of an Arduino and the gate of a MOSFET to limit the peak current (this is more important with big beefy fets - it's common to not use a resistor when switching little ones. While that's not best practice, empirically, it seems to work fine; I have devices that directly drive small FETs from the pin of an AVR and have been working fine for years, even PWMing the gate of the FET).
If you need to do PWM on the gate of a big FET, a gate driver may be appropriate. These are little IC's with a high impedance input, which can output very high peak currents to quickly switch the gate of a MOSFET. An example is the FL3100 (hey, I happen to sell these mounted on a breakout board!)
Thank you very much, I gave you 1 karma.
I forgot to mention the MOSFET itself- IRLZ44N.
From what you have written above the MOSFET does not suck current when ON or OFF only when at the moment it is switched? Can you please tell me what amount of current does it take and for how long (the duration of the pulse)?
Hi,
Because the gate to source connection is capacitive, if you are pulsing the gate at high frequency there will be some AC current, this is usually limited with a 470R resistor.
If you are just switching the gate ON and OFF at say 1Hz or less then you can consider the gate current as DC.
This spec tells you the gate current for your MOSFET.
Igss forward = 100nA... not much really 0.0000001A
Tom... 
So if i am switching it at 1Hz or less (1=< sec) then u dont need resistor, if the delay is less than 1s then i have to use resistor?
Hi,
There is debate about when to and not to use a 470R in the gate.
I use 470R or 220R in ALL my MOSFET circuits as I have seen what a MOSFET failure can do to the driving circuitry/ controller without a gate resistor.
The resistor will not affect the performance of your circuit.
The 10K gate to source is very important and should not be left out, 100K if you are trying to keep battery current down.
Tom...
Qg is the total charge on the gate, and will determine for how long current is drawn during switching; that is spec'ed in writing for specific conditions and also often with a graph (of Vgs vs Qg) - you can calculate how much current is drawn for how long when switching from that; that said, if you're not PWMing it (and maybe even if you are, depending on specifics), the impact of this on the current the project draws will be negligible.
Oh - and you can't use the Vgs(th) to determine what voltage you need to drive the gate with; instead, look for the lowest gate voltage that they give you an Rds(on) spec for.
TomGeorge - depending on what speed you're PWMing the gate at, and the specifics of the load, the resistor in series with the gate absolutely can impact performance. I am aware of situations (from this forum) where this was not accounted for that resulted in MOSFET overheat and failure.
So if i am switching it at 1Hz or less (1=< sec) then u dont need resistor, if the delay is less than 1s then i have to use resistor?
You are still (slightly) "abusing" the Arduino by directly driving a (small) capacitance, but you're only abusing it for a few microseconds once per second. Under those conditions there isn't time for the Arduino to overheat and the odds are good that there will be no damage and no problems.
And as TomGeorge says, the resistor also provides some protection to the Arduino from a shorted/blown MOSFET (by limiting the current).