0
\$\begingroup\$

I would like to calculate the response time of a PMOS circuit and of the circuit

Could you tell me how to do it?

enter image description here

I'm using this transitor. https://www.infineon.com/dgdl/Infineon-IRFR6215-DS-v01_02-EN.pdf?fileId=5546d462533600a40153563595592114

I tried to use the rise time of the PMOS to find it can switch (and i don't know if that's the good way), but I don't now the results with the load. CL en RL. I assume CL at t=0ms is equal to 0V. I tried to use dv/dt, but I failed VIN1 = 28V. VCMD is a command

Thanks.

edit: I made a mistake. The Zener is also connected to the source. Sorry

\$\endgroup\$
1
  • \$\begingroup\$ What's the zener voltage? That's going to short your Vcmd signal of it's lower than Vcmd. Having those two unrestricted components in parallel with your gate means they'll have an impact on the response time, we're talking nanoseconds (0.000000001 secs) here so you're asking for incredible detail. The biggest factor might actually be the Vin voltage and size of R1 \$\endgroup\$ Commented Jul 15, 2019 at 22:06

1 Answer 1

1
\$\begingroup\$

You've got two things happening when you pull VCMD low to get the FET to conduct: you're discharging C1 and discharging the gate capacitance of the FET. I would also recommend using a resistor directly on the FET gate to control your current flow to/from VCMD. The datasheet gives the values for a general use case with the time delay on/off and the rise/fall times at the bottom of the "Electrical Characteristics" table.

It looks like it takes 50ns to conduct given a -10V drive signal through a 6.8 Ohm gate resistor with a 12 Ohm load to a -75V supply. Alternatively with the same setup and a 0V drive signal it takes 90ns to turn off. This isn't exactly your use case but it may be close enough if you mimic the gate resistance and drive voltages (Figure 10 in your datasheet shows the schematic of the test setup).

For a more tailored answer to your application, MOSFET gates act like capacitors that you have to fill/empty to get past the threshold voltage to switch from conduction to non-conduction or visa versa. MOSFET drivers exist in high power applications to really slam the capacitors around to get the fastest on/off switching times possible by providing high currents to quickly source/sink charge. This limits the time the MOSFET spends in the linear conduction region where there's a voltage drop between the source and drain, therefore burning power (conduction losses). On the other hand, resistors are used on the gates to prevent infinite currents gate charge/discharge cycles from burning out control circuitry - this increases conduction losses because it slows switching speeds. So for driving FETs, that's an engineering tradeoff because higher current drivers adds cost, complexity and electrical noise.

Check this application note from Vishay for more detailed calculations to get switching time from the datasheet values for your particular use case: https://www.vishay.com/docs/73217/an608a.pdf

FYI just because you don't have a gate capacitor doesn't mean that it will respond infinitely fast, your microcontroller GPIO pin or other gate driver will have a Thevenin equivalent that will tell you what your effective "gate resistance" is at the Thevenin equivalent voltage, this is a way of looking at your gate driver's source impedance (which is critical to know how fast you can switch the FET).

\$\endgroup\$
4
  • \$\begingroup\$ Thank you for your answer. You're right, a resistor connected to the gate is missing. So, is the switching time equal to 50ns? Or I should calculate RC response time of the gate? Thanks \$\endgroup\$
    – Tack
    Commented Jul 13, 2019 at 23:04
  • 1
    \$\begingroup\$ Most likely not because that's almost best case scenario from the datasheet. I don't know how detailed you need to get, but you've got multiple stages: 1) How much current can VCMD source and at what voltage? 2) get rid of D1 and C1 because they're adding capacitance that your VCMD signal also has to drain while trying to turn the FET on 3) find the FET rise time (you haven't provided numbers but I bet its way worse than 50ns) 4) find RC time constant but FYI the FET is partially conducting so RC is not being charged from a constant voltage source... what are you trying to accomplish? \$\endgroup\$ Commented Jul 14, 2019 at 0:07
  • \$\begingroup\$ Thank you for your comment. I get what you mean. 1) VCMD is equal to 12V, and can provide until 2A. 2) ok 3) Isn't that 90ns from the datasheet? 4) I should find the RC on the gate? I'm trying to find the response time of the circuit when it is loaded. \$\endgroup\$
    – Tack
    Commented Jul 14, 2019 at 8:39
  • \$\begingroup\$ I've already given you the calculations that you asked for in your duplicate post. You also say that you just need the calculation but there's so many moving pieces that it doesn't seem like you know what you're asking for. For example you're asking about the RC on the gate... Well there's a big C1 and there's smaller C's within the gate that have unique charge/discharge curves (hence why this question does not have an easy answer), but you haven't provided a gate R. Instead you're asking us to guess the Thevenin resistance of your source based on an uncertain Vcmd signal that can supply 2A. \$\endgroup\$ Commented Jul 15, 2019 at 22:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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