There have been a number from posts of folks having issues trying to power some substantial load with an IRF520 MosFet.
Assumption: The below addresses the situation when the IRF520 is being operated directly from an Arduino output pin. i.e. Arduino output to the gate of the IRF520.
It turns out the IRF520 is a poor choice for use with a 5V arduino, forget even thinking about a 3.3V arduino.
The issue is the Arduino output pins does not provide enough voltage to the gate of the IRF520 to adequately turn it on. The Arduino outputs less than 5V even with no load, I've measured a Pro Mini at 4.5V out. This is not enough voltage on the gate of the IRF520 to operate reliably and that is barely turned on.
From the Vishay specification:
PARAMETER SYMBOL TEST CONDITIONS MIN. TYP. MAX. UNIT Gate-Source Threshold Voltage VGS(th) VDS = VGS, ID = 250 μA 2.0 - 4.0 V
This says the IRF520will just barely start to turn on at between 2 and 4 volts. The current the IRF520 will pass is 250 µS (or a quarter of a milliamp. This is not enough current to make an LED even visible)
The next important specification is the front sheet rated resistance. For the IRF520 RDS(on) = 0.27 Ω. However to attain this "ON" resistance the gate must be at 10V. The Arduino output of 4.5 V will result in the IRF520 being partially turned on and at a very unstable operating point where small changes can cause large variations in how the IRF520 might be operating in your circuit.
In the above explanation I've taken some liberties to make it easy to read, however the fact remains this MosFet is not one the casual hobbyist wants to use in their Arduino design.
Threshold voltage commonly confuses people - it is to be ignored completely, ditto max current.
The only important specs are the max Vds, the max Vgs (some logic-level devices are only +/-8V or so),
the on-resistance as mentioned, and the input capacitance (or total gate charge which is another way
of saying the same thing).
If you need to switch 10A, picking a 10A MOSFET will likely be a disaster. Choose MOSFETs by the power
dissipation using the Rds(on) and the formula power = I-squared-R
For instance a device with 4 milliohm on resistance will dissipate 400mW at 10A (and drop 40mV).
This is a good choice, little heatsinking will be needed, and it will tolerate brief overloads well.
Picking a ~10A device will require a large heatsink and lose maybe upto 1V across the device too.
For nearly every device I've seen the max current is just the theoretical max power rating in disguise, ie it only
applies for infinite heatsink (liquid cooling!) and the junction temperature at maximum (150C or so). This
is bad news for your bill of materials and long-term reliability.
Typically you'll find a realistic choice of MOSFET for 10A will have a nominal current rating of 50A or more.
Unless you intend to have a bulky cooling system, which should only be necessary for quite large currents
(20A or more tends to push you into that territory).
Many starter kits include a IRF520, probably because they can be bought in bulk for a few dollars per thousand. With 5V on the gate they can pass enough current to light a few LEDs or run the tiny motor that comes in the kits, it's usually when the noobies try to run a 5 meter string of NeoPixles or a motor that draws 2 or 3 Amps that they totally fail and the noobs come here for help. Putting them in a "starter kit" is doing beginners a disservice.
One reason - the other being the inclusion of a UNO where a Nano is really the more practical unless you have a shield which will concisely execute the required application - why I baulk at the suggestion of "starter kits".
I think you are better off to do the research and decide which "experiments" you intend to pursue and order specifically the modules for those as well as a jumper wire kit, a Nano with pins (fitted), the Mini USB cable and a solderless breadboard.
OK, some of these parts can be dodgy and a testing process is in order, but this will apply equally to "starter kits" unless you bite the bullet and buy them from Arduino itself (and then?).
As regards the FET switch, I am suggesting this Aliexpress module. I have purchased a couple but not tested them as yet. They supposedly contain (two of) the correct FET. While MarkT claims they are fake, I would like to know before I have the chance to test mine, has anyone here actually tried them and found either way?
The problem is that nowadays very many are fake, you can't tell without buying, and of course the fakes
undercut the genuine very rapidly in the market without a way to get back at the criminals
involved. Best to only buy from reputable suppliers who'll try to avoid the fakes and reimburse
you if any slip through.
Many fakes sort of work, but are cheaper less performing devices re-marked as the more expensive ones,
or simply testing rejects (which are supposed to be destroyed) that someone's then packaged anyway.
for instance its a pretty common experience on the diyAudio formums that buying a 200V 16A
audio output transistor gets you something more like a cheap 60V 4A part in a faked package.
Consequence: they tend to explode in circuit.
