3

I recently saw that the input parasitic capacitance of a MOSFET (Cgs, etc.) changes with the applied gate source voltage? Why is that?

I always thought that it was fixed since the parasitic capacitances are due to overlap capacitances or due to the gate channel (with oxide as dielectric) capacitor itself?

How can a physical overlap area or channel area (hence capacitance) change with the applied voltage???

AlfroJang80
  • 3,025
  • 2
  • 25
  • 50
  • 3
    There's a lot of junk going on the "plate" on the channel-side of the capacitor. It's not just a normal conductor. https://electronics.stackexchange.com/questions/66660/why-is-the-gate-charge-curve-miller-plateau-of-mosfets-dependent-on-vds – DKNguyen May 20 '20 at 04:58
  • 3
    In outline : remember that gate voltage affects the conductivity of the channel ... when it's conducting, there is much more conductive material close to the gate. So you have moved the two plates of the capacitor closer together... –  May 20 '20 at 07:31
  • Voting close for insufficient information -- different types of MOSFETs behave differently: some depend on Vgs, some depend on Vds. – Tim Williams Aug 05 '23 at 04:15
  • @TimWilliams can you give an example of a MOSFET that does NOT change its Cgs (a great part of the input capacitance) by changing Vgs? I can't think of any... – Designalog Aug 05 '23 at 09:10
  • @ErnestoG Any VDMOS I've tested -- basically any power transistor. It depends heavily on Vds, and very little or not at all on Vgs. At least over the range I tested. There should still be a MOS capacitance effect, but it wasn't perceptible at the accuracy and voltage range I tested (a few percent, so I might've missed it). Or it pulls in outside the tested voltage range (but I wouldn't expect a sudden change at like < -10Vgs, that would be weird). – Tim Williams Aug 05 '23 at 10:05

1 Answers1

3

the spacing of the gate "plates" will vary from

  • gate-oxide thickness

to

  • gate-oxide thickness plus depth-of-channel-bulk thickness
analogsystemsrf
  • 33,703
  • 2
  • 18
  • 46
  • 1
    not sure why this was voted down....to me you were saying that the separation of the "plates" increases as the channel increases, thereby changing the input capacitance. Perhaps someone thought this was vague as a complete answer? Even so, I don't think it deserved a down vote.... – jrive Jul 31 '23 at 21:10