I am designing a high side switch using P-Mosfet to drive a heat element. The Mosfet is driven by PWM from MCU, which in the test is around 40Hz, 10% duty.

Here is the schematic I am testing. I am driving the AOD403 with a N-channle mosfet. The heat element is supplied with 24V. enter image description here

I am facing 2 issues here, one is that there is a unexpected amount of voltage drop between Source and Drain of the P mosfet, when the Mosfet is turned on. From the scope shot, we can see the voltage drop is around 3V. Where is the voltage drop coming from? The Rdson of the is pretty small, around 7mOhm or less. The current is maximal 6A, should the voltage drop just be maximal 28mV, right?

Yellow is the PWM from MCU; Blue is the voltage across the heat element; purple is scoped at the Gate of AOD403. Yellow is the PWM from MCU, Blue is the voltage across the heat element, purple is scoped at the Gate of AOD403

The other issue I am facing is that when switching the heating element using PWM, the circuit is runing with audio humming. Is there some way to over coming this? Here is the scope shot with the rising edge. enter image description here

Any hint is appreciated. Thanks

  • \$\begingroup\$ Your circuit looks OK for 40Hz operation. Maybe you have a damaged FET? Have you tried replacing it? The maximum Vgs for this part is 25V, so you might want to check your supply tolerance to be sure you can't exceed it. 40Hz is low enough that dimensional shift is causing your audio noise problem. You could eliminate it by increasing the PWM frequency. You might not have to go up to 20KHz, but if you did you would guarantee that you wouldn't have noise. You would have to re-design the gate drive to operate at that frequency. \$\endgroup\$ – John D Jun 17 at 18:38
  • 1
    \$\begingroup\$ How did you measure the voltage drop? I only see the voltage across R4 in your scope traces, not across M1. Scopes are not very accurate for voltage measurements, and your supply voltage might be sagging. \$\endgroup\$ – marcelm Jun 17 at 18:48
  • 2
    \$\begingroup\$ @marcelm is correct, you should measure the Vds directly, or at least make sure your supply isn't sagging. The dimensional shift is the physical change in size or position of your components that's causing the audio noise (Thermal and/or magnetostriction/magnetic). You could increase the PWM frequency to 20KHz, but you'd need a fast gate driver to avoid excessive losses in the FET. \$\endgroup\$ – John D Jun 17 at 20:02
  • 1
    \$\begingroup\$ "Will decrease the value of R1 help?" - yes. \$\endgroup\$ – Bruce Abbott Jun 18 at 0:08
  • 1
    \$\begingroup\$ Try checking your connection from the power supply and to your load. At 6 amps, the leads and connections only need to have 0.5 ohms to give you the drop you are seeing. Add some heavy gauge wire and see if it improves. \$\endgroup\$ – John Birckhead Jun 18 at 13:17

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.