I'm looking for a power MOSFET that will handle 48 Volts and about 42 Amps continuous. The Fairchild RFP50N06 seems like it will work. The datasheet is here: www.fairchildsemi.com/ds/RF/RFP50N06.pdf
Two questions:
(1) What does the power dissipation max mean? It says 131 Watts, but I'm not sure what that means.
(2) It looks like the switching voltage maximum and minimum are + and - 20 Volts, but the threshold max is 4 Volts. Does that mean I can use an Arduino output to drive it, with 0 Volts turning it off and 5 Volts turning it on?
The power rating is what the FET can burn with a sutiable heat sink. It burns the current dropped across the drain / source. That in turn is determined by drain source resistance hence giving a voltage drop to multiply by the current for the power.
The threshold voltage is just when the FET starts to turn on, not when it is on fully, so no you can't use it connected directly to an arduino.
Power dissipation isnrelated to the on resistanceRdsON. Which is 0.022 ohm. This is the internal resistance of the Fet when fully turned ON..if you ate conducting 20 A then you can calculate the power. Then you have to take care of the heatsinking..
The max voltage is 60v meaning that is the max the device can withstand in the reverse direction before it breaks down, like a diode..
It can be used directly connected to the Arduino (5V Vgs), but the resistance will be much higher than the .022 ohms spec'd at 10V Vgs and the current will be around 20A (per figure 8 ). It won't be able to meet its full capabilities unless it is fed 10V.
Try FDP8030L and compare its current vs. gate voltage graph (figure 5) to the RFP50N06. Note that the FDP8030L is limited to 30V and doesn't meet your requirements; it's just for comparison.
Regarding your need for a mosfet, are you planning on driving it with PWM? Don't skimp on details.
If the MOSFET must switch on and off quickly, for instance as part of a PWM circuit, you also need to take into account the power burnt while the MOSFET is switching. That is going to be a factor of the driving circuitry and the required gate charge.
Ok, thanks Chargin I compared the datasheets examples.
I was wondering because I've been using these MOSFETs with circuit #5 of this instructable.
I've made a few projects with that combo (using Arduino w/PWM) and didn't see a reason to switch to the FQP50N06L, I haven't experienced any heat issues and the efficiency is around 75% (3 x 3W LEDs in series).
Would efficiency increase enough to make a switch to FQP50N06L worth it?
By using the RFP50N06 am I setting myself up for problems e.g. magic smoke show?
Thanks to everyone for the information. Your comments helped me better understand some of the things I was reading. That helps a lot.
A few more details about the project. I read about a permanent magnet motor with several stator poles that are electrically independent and each pole powered by its own driver. I'd like to build one and test the concept. I may be biting off more than I can chew, but I figure I might as well bite first and then try to chew.
To that end, I'm looking at using just a MOSFET to control each stator pole. I do plan to use PWM to drive the MOSFET. My tentative design has 17 poles on the stator and 19 permanent magnets on the rotor. Each pole will be powered by the same 48-Volt power supply, and draw up to 42 Amps for a maximum power of 2 kW. But the test-of-concept motor I actually build will probably just have 3 poles and 5 magnets.
As I say, I picked out the RFP50N06 as a possibility. Any other suggestions? And any suggestions as to a driver chip or circuit to switch the MOSFET? It seems like Microchip and others sell chips that are specifically designed to drive MOSFETS, but I could not find one that seemed like it would work with the RFP50N06.
These are not logic-level mosfets, so they can't be driven direct from an Arduino. However as you will be using PWM and switching such a large amount of power, and mosfets for this amount of power have high gate charge, you will need to use a gate driver chip anyway(for example MCP1407) and provide around 10 to 12V supply to it.
What bridge topology are you considering, I'm suspicious you think only one MOSFET
is needed per winding, which would be unipolar drive, which is unusual outside of
low speed stepper motors.
Thanks for the suggestions, dc42. Those do look much better.
MarkT:
What bridge topology are you considering, I'm suspicious you think only one MOSFET
is needed per winding, which would be unipolar drive, which is unusual outside of
low speed stepper motors.
MarkT, I'm embarrassed to say that I forgot all about the bridge topology. I will need current to flow both ways in the winding, so I guess I will need to do a full H-bridge.
What I would really like to do is find some off-the-shelf motor drivers that I could buy to test the concept. Something like the Pololu High-Power Motor Driver 36v20 CS, shown at www.pololu.com/product/1457. But Pololu said that the 50 Volt maximum is an absolute hard limit, so using a battery 48 Volt supply would not work. Also they said even with heatsink the maximum current is about 30 Amps.
[Interestingly, Pololu uses a IPD048N06 MOSFET on that board, and the MOSFET's datasheet says it can handle 60 Volts and 90 Amps: www.pololu.com/file/0J388/IPD048N06L3_Rev2.0.pdf. According to Pololu, the 50 Volt limit comes from the MOSFET driver on the board. But I can't tell which chip that is.]
Pololu was unaware of any motor drivers like theirs that could handle my 48 Volt, 42 Amp requirement. If anyone knows of one, please let me know. Otherwise, I guess I'm on my own.