How is it possible to limit the reverse current of the body MOSFET diode when the VUSB voltage drops to zero?
LTSpice shows a current spike that would destroy the mosfet in a very few moments when VUSB drops to zero.

 

hi ST,
There is a reverse diode built into the body of the PMOS.
Add an equivalent diode across the LTS PMOS model/

E

 

Discharging a 4u7 through body diode shouldn’t be problem, it certainly doesn’t destroy the mosfet since Vusb doesn’t pull the Vcc rail to ground like in your simulation (it just disconnects the rail).

If you insist on backwards current limiting you need to use two mosfets with back-to-back connection. Or mosfet+schottky diode for blocking the backward current completely.

 

@Michal Podmanický

Discharging a 4u7 through body diode shouldn’t be problem, it certainly doesn’t destroy the mosfet since Vusb doesn’t pull the Vcc rail to ground like in your simulation (it just disconnects the rail).

How can I modify my simulation so that it comes closer to the real case?

 

@Michal Podmanický



How can I modify my simulation so that it comes closer to the real case?

Add some realistic values for the cable resistance and inductance, for a start.

 

This simulates the disconnection of source, because it makes the source a high impedance, not a pull to Gnd.

I placed the R1 for slow C1 discharging. It’s also needed to discharge your soft-start capacitor.

 

OK I will try and let you know.
Is it right for the diode in my board to be wired that way, or is it better to connect it with the cathode at V1 and the anode connected to the node formed by C2 and R1?