Driving a MOSFET

akash laturia said:

Can I  drive Mosfet IRFP460 using MSP430G2553 since the minimum gate threshold voltage of IRFP460 is 2v

Drive it directly? - You can drive, but it won't open. When selecting MOSFET, you shall look at Gate-source/Drain_current graph. For particular MOSFET to reach 10A drain current, gate-source voltage shall be at least 5V. Nominal Gate-Source voltage of IRFP460 is 10V, not 2V ;)

akash laturia said:

Can I  drive Mosfet IRFP460 using MSP430G2553 since the minimum gate threshold voltage of IRFP460 is 2v

You can, But you won't be able to use the full current. Assuming the MSP is running on 3.3V, this is the gate voltage (unless you have source on a negative voltage between GND and -2V or raise MSP GND to 2V and MSP VCC to 5.3V).

On V_GS = 3.3V, the typical current will be about 100mA only. Note that 'minimum gate threshold voltage' doesn't mean that the FET will start conducting on >2V, it means the point where it startts is somewhere between minimum and maximum, and can well be 4V instead, which isn't reachable with MSP without tricks. On 25°C, it is likely to be ~3V

So I suggest picking a different FET or use an intermediate fet for higher V_GS.

In theory there is something that is possible but can be dodgy by putting a bit too much on the processor, but it had worked for me before. 
A Open drain configured output pin and an external pull up to 5V could do the trick for you, i would say 1Kohm or so.
If the pin does not support open drain configuration, then you will have to alternate between setting the pin configurations to output with zero or input Hi-Z.

Something like that, to use port 1 pin 1:

#define PIN BIT1

#define PIN_ON  { P1DIR &= ~PIN;}

#defrine PIN_OFF { P1DIR |= PIN; P1OUT |= PIN; }

void main (void)

{

 unsigned int counter=0;

  P1REN &=~PIN;

  for(;;)

  {

     PIN_ON;

    counter=1000;

     while(counter--){  _NOP();}

     PIN_OFF;

    counter=1000;

     while(counter--){  _NOP();}

  }

}

Hope it helps :)

ricardo marinheiro said:

A Open drain configured output pin and an external pull up to 5V could do the trick for you, i would say 1Kohm or so.

A bit more than 1k.
(5V-Vcc)/R must be <2mA.

This is a simple way to make input 5V tolerant and would also solve the problem of driving teh FET gate with 5V.

However, this solution would cause a permanent current of 1-2mA through the MSP pins clamp diodes to Vcc, while the FET is off. Which is way worse than any other proposal here. Remember, this shall be a low-power solution.

Jens-Michael

I agree with you, but like in any case, we choose a few key features and tolerate their consequences.

If its low power, most of the cases all components will be fitted for that.

If its cheep, the least components or the tricks like that one to make it cheap  will do.

If low power is the case i only have one solution for the problem, use a logic mosfet.
If low power and high current, TI have some SUPER ones called NexFETs, they will do the work, can be driven from a uC and at 3V you can already have more than 160A.
Everything depends. A LI~ion can be cheap if linear or efficient if switched, can be even intelligent and efficient if done with an MSP.

I've just finished a remote controlled solar power managing system proto, using a couple of those connected to a gecko and they work like a charm.

Hope it helps.

ricardo marinheiro said:

I agree with you, but like in any case, we choose a few key features and tolerate their consequences.

Sure. But in this thread, less power-connsuming solutions have been rejected not because of their complexity but because on not being low-power enough. So a higher-power solution, even is simpler, is likely to be rejected as well by the OP.

On this case then i don't see a better solution than a Logic Fet as a different fet, once using another fet to drive the first one, will probably:

- still have higher Rds plus the already in fet Rds

- twice the components

-  will have to be a P channel otherwise you wouldn't be driving anything, leading to even need once again to use an open drain, again, not directly available in the lower lines of controller.

- twice the price, in these days, logic fets, are +/- same price and TI NexFets can have about 90% efficiency.

Anyway, what is the application for the Low power uC pin mosfet driven?

 

Jens-Michael Gross said:

Remember, this shall be a low-power solution.

I don't see where that was mentioned as a requirement.

That being said, I would reject the solution based on the fact that the clamp diodes are being used. That's not what they are there for, and is an abuse of the silicon. But then again, long term reliability may not be high on the OP's requirement list either...

Brian Boorman said:

I don't see where that was mentioned as a requirement.

Nowhere. Sorry, my fault. I confused this thread with another active thread with almost the same topic.