Mosfet "Max safe operating area" question

"Marky"

I reckon your fan is way cool.

Pun intended.

........ Phil

Hi Phil, I thought the power dissipation will be only I^2R. i.e. 0.175^2*0.120 =3.7 mW Or may be I am completely wrong?

Thanks for the relpy. very helpful.

mark

Phil Allis> "Marky"

"Marky"

If V = 12 and I = 0.175 ......

SOA is a separate issue from ON resistance dissipation.

....... Phil

Looks like about 200mA @ 12V to me from the graph. From Pd it should actually be

Bear in mind this figure is a 'theoretical' figure based on a case temp of 12C.

Graham

Yes. At 12V and 0,175A the effective device resistance is ~ 68 ohms.

Graham

What Phil means is that this high starting current shows the presence of a capacitor being rapidly charged.

The best way to understand transient-power-handling capability in a MOSFET is with its Effective Transient Thermal Impedance curves, or Thermal Response curves, as ON Semi calls them, see figure 13 for the ntf3055L. These are the manufacturer's way of presenting the thermal-mass of the MOSFET's die, where the heat is first generated, and the thermal resistance and mass of the metal frame the die is mounted on, where the heat travels on its way to your ambient. Personally, I don't find the SOA plots very useful for power MOSFETs, they're largely a holdover from power BJTs, where one ignores them at his peril.

When you first try to turn on the MOSFET it has the full supply voltage across the drain-source, even as it may be conducting extraordinarily-high currents from the drain-node capacitance. This means the instantaneous power dissipation may be very high. To use the Thermal Response curve you pick the time duration of the thermal event, for example 10us. For your ntf3055L108 the single-pulse curve goes through at a normalized value of 0.002, which we can multiply by the Thermal Resistance = 72C/W to get a value of 0.145 C/W for a 10us pulse. If we allow a junction- temperature rise of 150C (we'll assume a cool room-temp start), the maximum allowed power dissipation will be 150/0.145 = 1040 watts. This tells us 1kW for 10us would leave the junction at +175C, which is the maximum allowed, and we'd have to avoid any further power dissipation until the junction cooled down some.

Considering your measured 8A, which if at 12V, is 96 watts. The extrapolated single-pulse curve tells us you're allowed 3.2kW for 3.5us. Used another way, it tells us your junction temperature will increase by 96 * 72.3 * 0.002 / sqrt 10/3.5 = 8.2C, which is pretty safe!

As for Phil's at 9A for 100us, we get 9 * 12 * 72.3 * 0.0065 = 50C rise, which tells us we'd be still well below the maximum.

That fet will be fine.

John

Actually, no, ugly trolls always make that sound before they speak.

John

Now John, you're often correctly admonishing the rest of us to avoid needlessly taunting people (even if they need it - sorry, couldn't help myself).

Well, yeah, you're right about that, but this is *Phil*

John

As per my knowledge, Vds on the SOA is VDS ON (Voltage when the FET is ON).