I am trying to bring rectified 110 volts AC down to 90 volts. Rectified 110 rms actually reaches about 155 volts, so I am operating
(on a half wave) anywhere from 0 volts to 155 volts. I have a circuit which, in simulation, does a respectable job of smoothing this and providing 90 volts with maybe +/- 3 volts ripple. In addition, I want to deliver up to 4 amps to my load.
That means this MOSFET is handling a lot of power, and power makes heat. So I find that there are graphs of Safe Operating Areas (SOA) for MOSFETs, and they seem to be premised on pulsed performance. I am doing pulsed performance here, but its a very long pulse (50% duty cycle, 60 Hz.). I don't find anything applicable in these SOA specs, so I am hoping some MOSFET guru out there can tell me whether I am setting myself up for melted MOSFET's, or can this be done with proper heat sinking?
Thanks in advance for any help.
Why not use an Switch Mode Power Supply (SMPS) instead?
SOA are a limit för bipolar transistors. High current an high voltage makes hot spots in the silicon so they have to reduce maximum power.
MOS doesn't have problem with hot spots. But read the datasheet, there you get all information.
The limits are voltage, current and power.
First try an SMPS.
That means this MOSFET is handling a lot of power, and power makes heat. So I find that there are graphs of Safe Operating Areas (SOA) for MOSFETs, and they seem to be premised on pulsed performance.
Normally the graphs show SOA at various pulse lengths and DC. You'll find the
curves are simply current, voltage and power limited (with the power limit
depending on pulse width / duty cycle.
You normally wouldn't go anywhere near the current limit.
Plot your system voltage on the I=0 axis, plot your max current on the V= 0
axis, both points should be inside the DC SOA. Draw a straight line between
them (for a resistive load), and check that if the line strays outside the DC
SOA the shortest pulse time involved. Ensure you switch the device faster
than that (in practice not a problem).
With a high voltage MOSFET you'll find the on-resistance is quite high and the
dissipation may be dominated by this. Note that the edge of the SOA, where it
represents a power limit, assumes you have infinite heatsinking (which you
don't), so derate accordingly.
If you are PWM'ing you will have to calculate switching losses carefully.
BTW If I was handling voltages this high I would definitely use a MOSFET driver
chip to control the gate voltage against dV/dt from the drain. Better still use an
IGBT which are much more robust for high voltage.