0

Sorry, a very noobish question here. I'm trying to understand if MOSFET ratings work on the principle that current and voltage are proportional (like Watts) or whether each rating is separate and shouldn't be combined.

Take for example a 2N7000 MOSFET rated at 60V DS and 0.2A Drain Current.

If I only ever run 30V through it, can it safely pass 0.4A of current? Or is it 0.2A all day?

(application: I want to use a MOSFET in a voltage boosting circuit from 5V to 12V but need up to 6A to flow through and am wondering whether I can use these 2N7000's I have or whether I should source a beefier part).

Tom Auger
  • 265
  • 1
  • 9
  • 1
    Short answer: No. This [Q&A](https://electronics.stackexchange.com/questions/137513/why-pulsed-drain-current-is-higher-than-continuous-drain-current-in-mosfets?rq=1) will help you understand. – Long Pham Jan 08 '19 at 16:32
  • You *could* double the drain current, but pulse width would be limited to 10uS at a 1% duty cycle. It is a matter of how hot the drain junction gets. Too hot and the MOSFET is destroyed. –  Jan 08 '19 at 16:42

3 Answers3

2

It's actually quite a bit more complicated. You must consider the current, the voltage, and the pulse width (how long the device is turned on for, being allowed to cool completely between pulses). This is specified in the "safe operating area" (SOA) chart in many MOSFET datasheets.

Here's the SOA chart for one version of the 2N7000:

enter image description here

This shows that the 0.2 A limit is only valid when \$V_{ds} < 2\ {\rm V}\$. For DC applications with higher \$V_{ds}\$ you must de-rate the maximum current.

However, in a pulsed applications it can carry higher current for a brief period.

Note: I found 2 different 2N7000 datasheets on the On Semi website. One is probably for the version made by the division that was formerly Fairchild Semiconductor, and one for part that's been made by On Semi and Motorola all along. One of these datasheets has no SOA plot at all. If you want to rely on the SOA curve it would probably be a good idea to be sure you are sourcing your parts from the product line that the more complete datasheet applies to.

The Photon
  • 126,425
  • 3
  • 159
  • 304
1

From the datasheet there's a safe operating area defined, that you need to stay within. enter image description here

There's a few mechanisms that can cause failure, and for DC operation, the heating in the FET is dependent only on the current, and the voltage developed across the FET - which varies dependent on the gate voltage and its effect on the state of the conduction channel. The maximum voltage is the other wall of the SOA graph, and is separate from the thermal limitations.

There's a maximum value for pulsed current, that is usually limited by the connections to the die, and a series of curves that define what can be pushed through for shorter periods - this is dependent on thermal mass of the die (and its mounting) and thermal impedance of the package.

Limit on this FET is up to 0.5A for <100ms pulses, and 0.2A for steady state.

Phil G
  • 5,609
  • 1
  • 10
  • 18
0

What do all those pulse widths mean?

1 cubic meter of silicon has 11,600 seconds thermal timeconstant.

10cm cube (1,000 cubic cm) of silicon is 100X faster tau, at 116 seconds.

1cm cube of silicon is 100X faster, at 1.16 seconds.

1mm cube of silicon is 100X faster, at 0.0116 seconds (11 milliSec)

100 microns of silicon (about the thickness of back-grind silicon wafers), in a cube, is 100X faster, at 116 microseconds.

Thus what you are seeing, in these very faster pulses, is the self-heating of the silicon die.

analogsystemsrf
  • 33,703
  • 2
  • 18
  • 46