Can this mosfet really switch 4A?

[xmen_xwk]

Can this mosfet really switch 4A ?

I need a mosfet to switch on/off 12V 4A load, it could stay on for minutes or hours. So I found this mosfet. And according to my calculation(which could be wrong), it can easily handle 4A, but something is bothering me, I'm not sure if I'm calculating things right.

So here is what I did
Its RDS(on) is 5.8 mOhms at 10VGS.
So the power(W) it will waste(?) will be P = I^2R, thats 0.0928W
and even if I take the smallest pad(Steady State (Note 5)) that is 164.6 C/W.
Thats only 15.27C.
15.27C + 25C ambient, its 41C temp of Mosfet. Is this real ?

 

[KlausST]

Re: Can this mosfet really switch 4A ?

Hi,

Your calculation seems to be correct.
This is for steady state only.

But you need to take care during switching also. Switching needs to be fast enough not to dissipate too much energy. And the switching current needs to be within SOA.

Klaus

 

[Eric Best]

Re: Can this mosfet really switch 4A ?

Your calculation (for steady state conditions) seems to be ok. However, you haven't specified the load to be switched. Is it pure resistive? In case of a load with significant inductive component you have to use some transient suppressor (diode, transil, etc., across the load) to remove the voltage spike (its peak voltage can be several times higher than the supply voltage 12V, for instance 200V) created during switching the load off else it will damage the MOSFET. Its maximum drain-to-source voltage is 30V. Also the control circuit can be influenced or even damaged by that.

 

[asdf44]

Re: Can this mosfet really switch 4A ?

Yes that's quit easy for a modern mosfet. As you may have seen, for "load switching" applications in the 12-24V range Rdson goes down to <1mOhm so you can switch quite a lot of current easily.

Off the top of my head if size is important that seems like a cost effective choice. Those leadless packages are optimal from a performance/size point of view but you could probably go down in cost and/or increase flexibility by going to a more standard package like SOT-223 or maybe SOT-89.

 

[xmen_xwk]

Re: Can this mosfet really switch 4A ?

Your calculation (for steady state conditions) seems to be ok. However, you haven't specified the load to be switched. Is it pure resistive? In case of a load with significant inductive component you have to use some transient suppressor (diode, transil, etc., across the load) to remove the voltage spike (its peak voltage can be several times higher than the supply voltage 12V, for instance 200V) created during switching the load off else it will damage the MOSFET. Its maximum drain-to-source voltage is 30V. Also the control circuit can be influenced or even damaged by that.

Its an inductive load, I believe a diode will take care of spikes. Thats not the problem here though. I believe it needs more things to be calculated.

 

[Eric Best]

Re: Can this mosfet really switch 4A ?

...I believe it needs more things to be calculated.

Well, would you tell us what particular things you mean?

 

[CataM]

Re: Can this mosfet really switch 4A ?

I believe it needs more things to be calculated.

As you have said. You have calculated the steady state. Transients assumed to be 0. Real mosfet has a turn on/off time. If you think it is fast enough for your switching frequency, then, should not be any problem.

 

[Eric Best]

Re: Can this mosfet really switch 4A ?

CataM, I agree, xmen_xwk wrote

"...it could stay on for minutes or hours..."

so the switching losses shouldn't play any role in this case.

 

[xmen_xwk]

Re: Can this mosfet really switch 4A ?

Well, would you tell us what particular things you mean?

As its being discussed, I only calculated for steady state, I need to calculate for switching and I dont know what more yet, maybe long time ON state effects or something.

 

[Eric Best]

Re: Can this mosfet really switch 4A ?

xmen_xwk, if the switching itself is fast enough (short ton/toff times) and within the range of minutes or even hours as you wrote, then I'm convinced there shouldn't be any problems with it. I don't know anything like "long time ON state effects". Do you do?