MOSFET choice in switching regulator

[findAround]

I'm trying to improve an old design that incorporates this battery charging circuit. Basic switching regulator, takes a PWM signal from a microcontroller, duty cycle is changed to achieve desired charging current.

I need to replace the MOSFET with one with a higher gate-source voltage threshold, since VCC will be increasing to 25V.

Why was a P-channel MOSFET used? These seem more expensive for the same performance.
What would need to change in the circuit for a N-channel MOSFET to be used instead? The PWM waveform can be inverted if required.

Any other thoughts on the circuit are welcome as well.

Cheers

[inse]

You want to feed the buck converter with a higher voltage than it’s designed for?
The Zener diode will disable the switching when you exceed ~12V supply.
You would need to check whether the inductance is suitable for 25V operation or the PWM can be adapted accordingly.
Regarding the MOSFET, you could limit the gate-source voltage with a zener diode, dissipating a little power.
Re-designing for a N-channel MOSFET would require a gate drive voltage above the supply voltage, generated by a bootstrap circuit.
That’s why a P-channel type has been used.

[findAround]

You want to feed the buck converter with a higher voltage than it’s designed for?
The Zener diode will disable the switching when you exceed ~12V supply.
You would need to check whether the inductance is suitable for 25V operation or the PWM can be adapted accordingly.
Regarding the MOSFET, you could limit the gate-source voltage with a zener diode, dissipating a little power.
Re-designing for a N-channel MOSFET would require a gate drive voltage above the supply voltage, generated by a bootstrap circuit.
That’s why a P-channel type has been used.

The current supply is 18V and it functions well, the zener just limits the negative swing of Vgs. I want to update the design so that it works with 25V. Thanks for the info about the MOSFET, I understand now why a P-channel one should be used. I'll look into using a second zener to limit the gate-source voltage, hopefully it doesn't dissipate too much power.

[Benta]

No reason to change anything. The circuit will work fine at 25 V as well.

[findAround]

No reason to change anything. The circuit will work fine at 25 V as well.

Simulating it in LTSpice, the gate-source voltage (green trace) across the MOSFET briefly spikes to VCC every time the input signal (blue trace) swings low. This would exceed the rated Vgs of the MOSFET if VCC were to -- just realised I mixed up source and drain.

The ACTUAL Vgs waveform is neatly limited by the existing zener diode, so the existing circuit will work fine at 25V, you're right Benta.
Thank you!

[T3sl4co1l]

Given the lack of sense circuitry in the section shown, I suspect it may be an open loop control, in which case be very careful that it's able to account for the higher Vin.  Vout is proportional to Vin * PWM, you're increasing the gain so the control (if closed loop) may go unstable.  It's not a big increase, unlikely it's that marginal -- but another thing to check.

Tim

[findAround]

Given the lack of sense circuitry in the section shown, I suspect it may be an open loop control, in which case be very careful that it's able to account for the higher Vin.  Vout is proportional to Vin * PWM, you're increasing the gain so the control (if closed loop) may go unstable.  It's not a big increase, unlikely it's that marginal -- but another thing to check.

Tim

Thanks Tim, there's current feedback via a shunt resistor fed back to the microcontroller, so this shouldn't be an issue.