0
\$\begingroup\$

I am planning to build a circuit that could convert a DC voltage (i.e. 18V) to an AC voltage with frequency f=1kHz and Vpp=36V. Here is my schematic in Orcad Capture Lite. The main idea is to use an H-bridge with 4 MOSFET and turn them on/off by applying specific PWM wave on the gate. So that the output between A and B will be an AC voltage.

enter image description here

Component

  • I use 2 N-channel MOSFET at the low end(IRF3205), and 2 P-channel MOSFET(IRF4905) at the high end. Also, there are two 2N2222 BJT to provide the gate voltage of the high end.
  • V2 and V3 are PWM signals which are out of phase and with 20us overlap. (see waveform below.)enter image description here

Porblem

When I run the PSpice Simulation, the results are shown in the figure. The top one is the current of the V1 source. The bottom one is the voltage between A and B V_ab, and source-gate voltage of M5 Vsg. enter image description hereenter image description here

The problem appears at the time of 0.5ms, at this time both P-MOS and N-MOS are turned on and causing large short current.

How to address this issue?


Update 1:

I increase the overlap time of both PWM from 20us to 100us. The result is followed. The large current happens when the rising edge of any PWM wave. Look closely at the V_SG of M5 (yellow curve), the value of V_SG will have a spike even it was at a low level. Is it because of the parasite capacitance? enter image description here

Update 2:

I change the PWM's peak voltage from 5V to 4V, the result is shown below. The short current is much lower, around 2A. I could add resistors on PWM to lower the 5V to 4V. enter image description here

Or someone has a better idea?

\$\endgroup\$
  • \$\begingroup\$ Overlap? You mean dead time? \$\endgroup\$ – winny Apr 17 at 22:31
  • \$\begingroup\$ Have you tried increasing your deadband? The MOSFETs may be turning off slower than you think. \$\endgroup\$ – evildemonic Apr 17 at 22:34
  • 1
    \$\begingroup\$ not related to your question, but the data sheet suggests that those mosfets perform better with 10V on the gate, and a real the 2n2222 would be most unhappy with 5V between base and emitter (use a resistor on the base). \$\endgroup\$ – Jasen Apr 18 at 1:59
  • 1
    \$\begingroup\$ What;s the purpose of R5 and R6 (18v is allowable on the gate) , why is R7 different to R8 ? \$\endgroup\$ – Jasen Apr 18 at 2:02
  • \$\begingroup\$ @Jasen The purpose of R5 and R6 is to avoid large V_sg on P-MOSFET. I corrected R8=R7 and R6=R5, still the same issue. \$\endgroup\$ – Libo Wu Apr 19 at 1:59

Your Answer

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

Browse other questions tagged or ask your own question.