1

From my understanding, the SOA lists the drain current for a given VDS in the ohmic and saturation regions where the FET can function without damage.

I read a thread that says FET SOA graphs with short pulses and no line for DC operation are not rated for DC in the linear mode/saturation region.

Thread in question: What's the difference between field-effect transistors (FETs) marketed as switches vs. amplifiers?

Can these FETS operate with DC in the ohmic region without an SOA line for DC operation?

I don't have a specific FET in mind but I plan on using a FET driving a relay that will remain closed for an extended period of time while DC flows.

Here is my schematic(I forgot the gate resistor):

enter image description here

Related thread: Undestanding the SOA of MOSFET from the datasheet?

ocrdu
  • 8,705
  • 21
  • 30
  • 42
AllHailTheMilkman
  • 11
  • 2
  • Linear and ohmic regions are not the same. – Hearth Dec 30 '20 at 21:55
  • Yes I know they're different. – AllHailTheMilkman Dec 30 '20 at 21:56
  • There is no such thing as a MOSFET that can't be operated continuously as a switch (as you intend). OK, probably there is somewhere such a thing, but it will be some oddball special purpose MOSFET. Not the one you will find when you do a simple parametric search on mouser or digikey. The oldest common MOSFETS are 2N7000, 2N7002, BSS138 and any of them could work in your application assuming they can handle the DC power dissipation. Of course any newer MOSFET will also work and have lower Rds(on) for a given V(gs). – user57037 Jan 01 '21 at 19:20

1 Answers1

1

When using a FET as a switch it makes sense to choose a FET, that is designed for switchig. These (MOS)FETs typically include a parameter called RDS(on) in their parameter list. This is the remaining DC resistance, when the FET is switched through.

Typically, a FETs DC capability is limited in the following ways:

  • Maximum Drain current.
  • Maximum power dissipation or junction temperature.
  • Maximum breakdown voltages on Drain and Gate pins.

As long as you don't overload any of these parameters, it is, generally speaking, fine.

FETs for amplification (operation in saturation) are a little different by design. For an amplifier FET it is usually not that important to have a low RDS(on). Instead parameters like the Drain-Source parasitic conductance, noise, linearity, parasitic capacitances etc. are optimized. These devices are just not made to be operated in the linear region and can perform worse than a "digital" switching FET.

However, as long as you don't overload the FET with regards to the above constraints, it will work. But again: DC performance can be a lot worse than a FET that is made for switching.

GNA
  • 1,471
  • 13
  • 18
  • If I have a switching FET I think it's required to check if the FET's VDS and ID is within the SOA boundaries to determine the FETs DC capability . To be specific I mean looking at the SOA for switching FET operation when it's transitioning between on and off ( in saturation/linear mode) and when it's turned on in the ohmic region. – AllHailTheMilkman Jan 02 '21 at 22:21
  • Yes. You have to look at the dynamic behavior, too. However, switching performance vastly depends on your load and driver. You have to take them into account. During switching none of my limits must be violated. It is possible to calculate (at least to some degree) the switching loss for rising and falling transition. Add both together and multiply them by the switching frequency and you got your dynamic switching loss, which must not violate the maximum dissipated power of your transsistor+heatsink combination. – GNA Jan 03 '21 at 18:27