knut_ny:
Datasheet says that its fully open at max 4V, so the Arduino will open it properly.
I guess your next challange it to speed up PWM to at least 20kHz .. then some filtering.
NO NO NO!!
Firstly in English "open" for a switch means "off" or "no connection" as in "open circuit",
its not like a water tap at all. I presume you mean "fully on at max 4V".
That MOSFET is not logic level and certainly is not switched on at 4V.
Perhaps you think threshold voltage is when the device is on? Its not, its the point
at which the device switches OFF, or more precisely conducts only a few microamps.
The voltage at which the device is guaranteed to be fully on is given in the Rds(on)
spec as the Vgs value - here its given as 10V for a Rds(on) of 32 milliohms max.
That means you need at least 10V between gate and source to guarantee the on resistance
is at or below 32 milliohms.
In general a logic level device will have Rds(on) quoted for Vgs = 4.5V. It will also
likely have a threshold voltage of 0.5 to 1.0V.
Anyway, back to the original question.
Yes you can use that MOSFET but you will need a MOSFET driver chip to drive the
gate and a 10 to 12V supply to power the MOSFET driver.
With high voltage high power switching you cannot safely drive the gate directly from
an Arduino anyway as the drain-gate capacitance can back-drive the gate and burn
out the Arduino - you need a beefy driver chip to hold the gate voltage steady, something
like the MIC4422 will do nicely.
Again with that big load voltage you need to think about protection circuitry really -
a TVS diode to protect the 12V rail from over voltage, a 15V zener on gate-source
to protect gate oxide layer, a 10k resistor between Arduino and MOSFET driver to
protect Arduino.
But also we need to establish why you are trying to use low-side switching in the first place,
is seems unrelated to "regulating a power supply" - that would imply high side switching
if anything, which at 120V is rather more complex.