[SOLVED] MOSFET Dynamics Drives me Crazy !!

Status
Not open for further replies.

bio_man

Full Member level 2
Joined
Mar 30, 2010
Messages
144
Helped
2
Reputation
4
Reaction score
4
Trophy points
1,298
Activity points
2,698
Hello folks,

I am using MOSFET as switch and I want to find the switching power loss.
Looking to the dynamic characteristics in dataSheet of any MOSFET, there is these four timing parameters:

td(on) == time to charge Ciss and pull Vgs up to Driving signal level. (Vds does not change, Ids=0)
tr == MOSFET now is conducting, Ids starts to rise and Vds starts to fall to zero.
td(off) == time to discharge Ciss ( Ids stays the same, Vds=0)
tf == MOSFET is turning OFF, Vds start rising and Ids falls to zero.

considering these states, it means we have only two states for switching Power Loss:

1) Psw(on) == during tr only (I assume tr is meant for current rise time, because usually it is much higher than tf)
1) Psw(off) == during tf only

Does my assumptions make sense? because I saw some references found Ton and Toff which are,
Ton=Td(on)+tr
Toff=Td(off)+tf
and then found Psw(on) and Psw(off) based on these timings, which I believe incorrect because in both, Td(on) and Td(off), one parameter is zero, Vds or Ids, so there is NO power loss!.

what do you think guys?
 

You are only dissipating power in the mosfet when it's conducting, and it's not conducting (by definition) during the td(on/off)times. So, I agree with you.
 
Last edited by a moderator:
I am using MOSFET as switch and I want to find the switching power loss
thanks, what is the actual topology you are using, buck?, boost?,LLC?,BLDC inverter? DCM?, CCM?
 
..you should read the article on high speed mosfet drive by laszlo balogh.....

- - - Updated - - -

https://reipooom.files.wordpress.com/2011/08/mosfet-must-read-2-slup169.pdf
...pages 4 to about page 10 or so

this really for good reference! thanks treez, I really appreciate sharing it !

I feel now more confident, and I like Laszlo's conclusion, he said 'Calculating exact switching loss is impossible due to the parasitics and how they are affected by the load'

I would say, always take the worst case of assuming hard load (capacitive and inductive), then Psw=(1/2)*Vds*Ids*fs(tr+tf)
which is almost 3 times Psw if resistive load is considered (Psw(resistive Load)= (1/6)*Vds*Ids*fs(tr+tf)

Thanks all for your feedback and take care guys!

- - - Updated - - -

thanks, what is the actual topology you are using, buck?, boost?,LLC?,BLDC inverter? DCM?, CCM?

Switched-Capacitor converter (No inductor involved), it is usually CCM,
 
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Switched-Capacitor converter (No inductor involved), it is usually CCM,
aaah..ok, just remember that the Balogh article refers to "clamped inductive switching" as it says on page 5
 

aaah..ok, just remember that the Balogh article refers to "clamped inductive switching" as it says on page 5

Yep, you are right. Actually, I figured it out because he assumes hard load switching.. thanks though for your comment and feedback.
 

Status
Not open for further replies.

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…