While simulating a flyback converter, I find that even in saturation mode the MOSFET, in spite of having Rds(on) around 0.74 ohms and the Ids = 5.75mA, is dropping almost the entire drain supply input voltage. The snubber circuit and the sense resistor also consumes very little power.

Is this high Vds in saturation mode expected? If yes, why? Is it because the power is a pulsating? Can a floating gate destroy a MOSFET? Practically I found Vds = 1.4V, does that mean the MOSFET is bad?

DC Analysis

DC analysis

Transient Analysis

Analysis using oscilloscope

  • \$\begingroup\$ You may wish to change W and H for the MOS to some more appropiate numbers... or just choose a model for it, say IRF840 (some quick, random example). \$\endgroup\$ Jul 29, 2018 at 5:52
  • \$\begingroup\$ @aconcernedcitizen : I have tried with many models, almost same result. I will give a go tho. \$\endgroup\$
    – seccpur
    Jul 29, 2018 at 5:55
  • \$\begingroup\$ If the VM3 trace is to be believed, the FET is on all the time and you're simply modulating its saturation current. In a flyback, you turn it on, and off. \$\endgroup\$
    – Neil_UK
    Jul 29, 2018 at 6:36
  • \$\begingroup\$ @neil_uk: VM3 is pulse train going from 12V - 0V although the plot due multiple signals don't look like one. \$\endgroup\$
    – seccpur
    Jul 29, 2018 at 8:04
  • \$\begingroup\$ Have you tried adding an earth node to the circuit? \$\endgroup\$
    – Andy aka
    Jul 29, 2018 at 10:12

3 Answers 3


it seems that VM3 is not pulled to ground from the graph, thus the mosfet is not able to turn off. can you plot it seperately?

  • \$\begingroup\$ Sorry for the late response. The VG1 pulse train, hence VM3, runs from 0v to 12V, but the graph plot looks messed up may be because of the multiple graphs. \$\endgroup\$
    – seccpur
    Jul 30, 2018 at 1:24

I'm not sure why your mosfet isn't turning on in sim but if your simulating a flyback you should add a diode and cap to your output. Also, the phasing on your transformer is wrong. These things are actually really important to how a flyback operates.


T1 is a relatively small MOS device that you would find inside of a chip. It's a model for simulating integrated circuits - I don't think you're doing that. Instead, use a model of a discrete MOSFET.

Rds(on) around 0.74

That's true as long as the mosfet is not acting as a current limiter.

is this high Vds in saturation mode expected

The tiny MOS device can't source any more current and acts as a current limiter. Also, it's not realistic to expects a 10um square channel integrated MOS device to act as a switch for 300V. You're not designing a chip.

Also, without a ground node, the SPICE simulator is probably choosing a ground node for you, and that may not be the node you want.

The simple fix to get the simulation to at least represent a circuit you can build would be to use an actual discrete power NMOS model, and to choose a ground node explicitly.


Not the answer you're looking for? Browse other questions tagged or ask your own question.