1
\$\begingroup\$

I am trying to model a gate driver desaturation detection mechanism with discrete elements on LTSpice. The waveforms I have obtained now provide a constant Vdesat of about 5.6V when MOSFET is turned on. But even upon varying Vdc or Id for the MOSFET, the Vdesat does not change, hence it is unable to detect desaturation conditions. I understand beyond a limit Ddesat is supposed to get reverse biased and I1 should charge the Cdesat. Kindly please help me identify my mistake. I have added a blanking switch to pull Vdesat to ground when the MOSFET is turned off. Do I need to alter values of Rdesat/ Ddesat specifications?

Edit: Added some waveforms

enter image description here

enter image description here

\$\endgroup\$
6
  • 1
    \$\begingroup\$ Problems with LTspice schematic: 1. S1 has model MYSW, but MYSW is not defined. 2. Expression for B1 contains "?". 3. Expression for B1 has reference to LTspice defined node v(N011), this is not good practice, expressions should only contain user defined node names, since definition of N011 could change. \$\endgroup\$
    – Fabio Barone
    Commented May 31 at 12:24
  • 1
    \$\begingroup\$ Waveforms to add: 1. Voltage at drain of M1. 2. Voltage of node Vcontrol. 3. Current in Ddesat. 3. M1 drain current. \$\endgroup\$
    – Fabio Barone
    Commented May 31 at 12:26
  • \$\begingroup\$ @FabioBarone I removed mysw, was for testing in an auxilarry circuit. The ? in B1 denotes the conditional statement i.e if both conditions are satisfied it is low, else it is high. Also renamed V(n011) to V(gates_sense) \$\endgroup\$
    – Agniva
    Commented Jun 2 at 13:21
  • \$\begingroup\$ Also. added the waveforms \$\endgroup\$
    – Agniva
    Commented Jun 2 at 13:28
  • \$\begingroup\$ Have you updated the schematic? N011 still appears in the expression for B1. \$\endgroup\$
    – Fabio Barone
    Commented Jun 2 at 23:32

1 Answer 1

Reset to default
1
\$\begingroup\$

5.6 volts seems exactly right for moderate to low drain currents.

You have a current source of 5 mA flowing through a 1 kΩ resistor and a forward biased diode (when the MOSFET is conducting). This is going to produce a voltage of 5.6 volts for a MOSFET that turns hard-on to 0 volts.

If the MOSFET turned on to a looser value of (say) 1 volt then you'd see 6.6 volts on your DESAT monitor line.

In other words, I see nothing wrong with your circuit.

But even upon varying Vdc or Id for the MOSFET, the Vdesat does not change, hence it is unable to detect desaturation conditions.

Well, your MOSFET has an on resistance of around 0.4 milli ohm so, if you are pushing around 200 amps of drain current into the device, you would expect to see an increase in Vdesat from 5.6 volts to 5.68 volts.

I'd be tempted to lower the 1 kΩ resistor to maybe 100 Ω so that Vdesat rises from 0.65 volts to around 0.73 volts when there is 200 amps of drain current.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ @Agniva a 7 volt saturation voltage across a 1 milli ohm MOSFET implies a current of 7,000 amps and this sounds like you have your math wrong. The MOSFET in your question (linked to in my answer) has a maximum pulse current of 1200 amps so, I can't see how you get 7 volts. Maybe you mean that your desat monitor circuit produces 7 volts rather than "they" (the MOSFET) has a desat threshold of 7 volts. Please be clear. \$\endgroup\$
    – Andy aka
    Commented Jun 2 at 13:55
  • \$\begingroup\$ OK, you need to increase the value of the resistor in series with the diode that connects to the MOSFET drain. You call it Rdesat in your original diagram @Agniva ||| I don't understand the relevance of the ADuM4146 MOSFET driver circuit. \$\endgroup\$
    – Andy aka
    Commented Jun 2 at 15:05
  • \$\begingroup\$ @Agniva what does this mean --> Vdc increases around 40us ?? Please also note that evolving questions will usually not be tolerated that well so, be clear in your comment and explain what it is that you are having a problem with. Double check your comment to make sure it makes sense. \$\endgroup\$
    – Andy aka
    Commented Jun 2 at 15:45
  • 1
    \$\begingroup\$ I think your have fallen into the trap of not giving the full picture when you wrote the first revision of your question then, I answered that 1st revision but gradually, you are trying to draw me into doing a full analysis of the ADuM4146 and, I'm not prepared to go that far. We have a name for this type of question; a chameleon question and so, I'm advising you that I'm not prepared to support this evolving question any more. This is a Q and A site with strict rules and an evolving question like this is breaking the rules in all fairness. \$\endgroup\$
    – Andy aka
    Commented Jun 2 at 18:38
  • 1
    \$\begingroup\$ Sorry for making it confusing. Thanks for being patient with and answering my questions. I'll try to figure out the rest. \$\endgroup\$
    – Agniva
    Commented Jun 2 at 18:46

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