I am desiging a relay circuit using MOSFET.The MOSFET is controlled by a microcontroller.Most probably it will be STM32.It's gpio output voltage is 3.3V and can provide 15mA of current.

The relay coil voltage is 12V and coil current is 150mA. Typical circuit is shown below.I know how to design this circuit with BJT.The mosfet I am planning to use is AOSS32334C. The Vgs(th) of this mosfet is 1.8V


GPIO voltage is 3.3V it can provide 15mA maximum.I set Rlimit such that MOSFET will draw current less than 15mA.
I want to draw only 5mA.

Rlimit = 3.3/5mA = 660 Ohm. I will set R1=10times of Rlimit

May I know is this approach correct or not.

 

That 1.8 volt is at a current of only 250ua.
Meaning the mosfet is just starting to conduct.
3.3 volts will probably work.
Would set R1 at 10K and Rlimit at 100 ohms.
How much current does the relay require?

 

Hello,

@sghioto , the topic starter has stated that the relais coil needs 150 mA.

@hoyyoth , a mosfet only takes a short current pulse at the moment the gate voltage is applied.

Bertus

 

That 1.8 volt is at a current of only 250ua.
Meaning the mosfet is just starting to conduct.
3.3 volts will probably work.
Would set R1 at 10K and Rlimit at 100 ohms.
How much current does the relay require?

Coil current is 150mA

 

GPIO voltage is 3.3V it can provide 15mA maximum.I set Rlimit such that MOSFET will draw current less than 15mA.
I want to draw only 5mA.

I'd eliminate Rlimit and not worry about the gate charging current as gate cap isn't excessive. The value of R1 isn't critical - 10k or larger will be fine.

 

I'd eliminate Rlimit and not worry about the gate charging current as gate cap isn't excessive. The value of R1 isn't critical - 10k or larger will be fine.

Is it not wise to keep Rlimit to protect the gpio pin against possible transient (from inductive load or otherwise) voltage passing through the capacitive gate? (And if so, should it be larger than R1?)

 

Why you think Rlimit should be larger than R1.

 

Why you think Rlimit should be larger than R1.

While it would become tricky to get a reasonable voltage at the gate from the GPIO, and that is why another option should be considered than what is used as it is now, to protect against a possible transient surge, the same voltage divider is the first (and only) line of defense against short pulse, in the proposed circuit.

I admit that it is not a clever idea to just incread Rlimit though. In the end, it creates more problems than the hypothetical one it tries to avoid.

 

Is it not wise to keep Rlimit to protect the gpio pin against possible transient (from inductive load

If there was an inductive load! A MOSFET gate is a capacitive load. The only reason to use a series resistor is to avoid ringing when switching at high frequencies. And 10 Ohms is typical.