I built up a circuit which did not work as expected. I reduced the problem to this circuit:

enter image description here

I expect to have a current of ~2.4 A at gate voltages below 10 V, but actually the transistor does not fully switch at 100 V. What is the Vgs threshold voltage I should assume for the standard nmos? I expected it to work at any value as it is an idealized model, isn't it? What I'm I doing wrong?! I tried adding a pulldown at the gat, even though I didn't expect this to enhance the situation...

Info: I'm using LTspice XVII from Nov 26 2018 under linux with wine.

  • \$\begingroup\$ Standard switching NMOSFETs are meant to be driven with 10V-15V, but can handle as high as 20V, sometimes 30V, at the gate Note that this is higher than and not the same as the Vgs threshold). Logical level NMOS are expected to be driven with 5V at the gate. Sometimes even 3.3V or 2.7V. \$\endgroup\$ – DKNguyen Feb 28 '19 at 19:45
  • \$\begingroup\$ Yes, I know that. But in my simulation it doesn't switch at 100V, that's what I wonder. \$\endgroup\$ – Sim Son Feb 28 '19 at 19:46
  • \$\begingroup\$ Hit F1 on LTspice. In the search box enter 'M' (without quotes) and search. Select the "M. MOSFET" entry and read it. There is a default column that tells you all the details used, by default, when you don't otherwise specify a model. The default MOSFET level is "1". \$\endgroup\$ – jonk Feb 28 '19 at 19:48
  • \$\begingroup\$ Well, 100V at the gate would blow the MOSFET in real life. I'm not sure how simulations deal with hardware failures. I don't know if they just ignore it and simulate through it as if it worked or if they spit out garbage. \$\endgroup\$ – DKNguyen Feb 28 '19 at 19:48
  • 1
    \$\begingroup\$ @Toor I'm not sure how simulations deal with hardware failures. Depends on the MOSFET's model and features. I have not yet seen very sophisticated models in LTSpice that would complain about exceeding Vgs_max for example. In more sophisticated simulators (Cadence Spectre) with appropriate models you might get a "Warning: junction melting" (yes, that is funny) and/or over voltage/over current messages. \$\endgroup\$ – Bimpelrekkie Feb 28 '19 at 20:12

You expected Amps instead of the mAmps you get.

I am quite sure that this has to do with the model used for the NMOS. You very likely didn't select a specific model nor did you specify a certain W/L. So LTSpice uses a default model which cannot conduct much current as it is likely more suited to model a small signal NMOS. It is somewhat similar to an NMOS you'd find inside an IC: quite small and for small currents only.

What happens now is that the NMOS determines the current and since it has a limited W/L it will not conduct much current.

What you need is a "really big" NMOS, with a very large W/L, one that is intended for switching high currents.

There is a library with discrete components in LTSpice. You need to see what is available there, then look in the datasheet of that MOSFET (type the model number in Google) and see what its maximum Drain current is.

After some Googling I found that maybe the IRF7401 is available in LTSpice, if it is there try it. It has an Rds_on of 0.022 ohm so that should work.

  • \$\begingroup\$ Thanks for your research. Even though Id idn't find the IRF7401, thanks to your answer I found the "select MOSFET" button in the mosfet properties. I've been looking for a specific mosfet before, but didn't find a library with discrete components \$\endgroup\$ – Sim Son Feb 28 '19 at 20:26
  • \$\begingroup\$ @SimSon OK, indeed these libraries might not be included but you can find some links here: forum.allaboutcircuits.com/threads/… \$\endgroup\$ – Bimpelrekkie Feb 28 '19 at 21:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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