AceofWWII:
My apologies Marle, I did not intend to come across aggressive.Thank you for your response PE. I don't believe heat should be a concern for this project, as it is just going to be turned on for short periods of time (1min or so) a couple of times per day. As far as the drive voltage, I should have plenty for the drive voltage (9V), and my gate voltage is going to be at 3.3V with a 3.3V Arduino. As you can tell, I have never used a FET before so I thank you for clearing up that. You said that the FDD8447 would work well assuming I had a 5V Arduino, and the minimum gate voltage would be at 2V. I'm assuming this would also work fine with a 3.3V gate voltage, would I run into any issues with that?
So, to clear up any misunderstandings, I am running a 3.3V Arduino with a 950-1050 mA LED. My gate voltage is going to be 3.3V from the arduino, and my drive voltage will be 9V from a battery (not a supply). The LED will be connected to the 9V source.
Sorry it has taken me a while to get back to you. The 3V3 could be a problem for the FDD8447. Taking a few tolerances into account it seems a little close to the edge and even 1A might be enough to pull the MOSFET out of saturation. Best if you can to bypass this concern altogether and drive the MOSFET gate to 9V.
When I say 'drive' I mean 'pull up gently' but before I explain; this thing is running on a 9V battery, yes? So power consumption is a fairly critical parameter?
If it is, then there is a simple add-on option to the following that will reduce the current a great deal.
My suggestion would be to use a small fet or BJT to short the gate and source of the power mosfet (to turn it off) and a pullup resistor from gate to +9V that will turn the power mosfet on when its gate isn't shorted to source via the small mosfet / bjt. Did that make sense? (I may have to master getting schematics into posts.) The pullup resistor should be low enough value that the transition from off to on does not generate enough heat to damage the mosfet. This is where a part in a bigger package can be real advantage. A SOT23 device has little room for error but a TO220 or better, a TO247 could switch as slow as a wet week and be ok.
This new arrangement will mean the arduino output will be active low rather than active high. And you will need another transistor, say a BSS138 or a 2N2222 as the first transistor. The 2222 you already have so lets just assume that's what you'll use. Since it is a BJT it will need a base resistor to limit the base current. The advantage of driving the gate to 9V is that you could use almost any mosfet that was good for the load current and be ok. Just plucking a few numbers from data sheets (FDD8447) and a very rough calculation, very rough, I get a switching transition time of around 3.5us after a delay of around 10 to 15us for turning the mosfet on from 9V with a 9V load and using a 10k pullup resistor on the gate. A 33k base resistor should be ample for the bjt (assuming a 2222). The transition, simplifying the load to a constant current (not proportional or resistive) the power lost in the switch (FDD8447) will be no more than 30uW per transition so given it will only switch a few times a day I would be astonished if you could even feel any warmth whatsoever from this circuit.
Let me know if you need more details or if the pullup current is going to be a problem.
PE.