1) Could someone give me a sanity check on the switching loss of M1 w/o the aux circuit? I'm getting around 30W. Is that in the ballpark?
2) In ECEN 5807 there normally was no switching loss in the diode. The switching loss showed up in the primary switch. In this case, during time tb, there is an increasing voltage across the diode as there is still current decaying to zero. It looks like there must be some power loss in the diode during time tb. Am I correct?
Subscribe to:
Post Comments (Atom)
16 comments:
I agree with your comment to (2). I'm currently working on the calculation for the losses now.
For the transistor, I'm struggling a bit. I'm not sure how the voltage profile on the transistor looks as it decays from Vg to zero (this occurs in the second time interval between ta marker and tb marker)? Or does Vg drop straight to zero without an exponential decay?
I get around 6W for the diode.
When I calculated the loss in the transistor I used a constant 400V for Vds until the charge is removed from the diode. This approach is similar to the buck converter example on pages 96 and 97 of the text. I believe the 400V, across C at the output, will be present across M so long as current is flowing through D. It seems to me that so long as D is on, there will be 400V between the drain of M and ground. Does anyone disagree?
Brandon, what approach did you use to calculate the diode loss?
In the tb interval I am getting a linear equation for i(t) and V(t) and the loss will be intgral of the i(t)*v(t) over time interval tb, but that integral gets kind of complicated to solve....am I in the right direction?
Hello Yusuf,
If both v(t) and i(t) are linear function, that
sounds reasonable to me.
However, during tb voltage tapered down as linear function (from problem statement), but, current seems like exponential function (from figure 2).
I'm not sure yet how we should treat the current part.
Tanto,
If you are referring to reverse recovery current during tb, we can safely approximate as a linear fn. rather than exponential, I think.
I get switching loss in the order of hundreds of watts. Does this seems correct??
My original calculation for the transistor switching loss was incorrect since I assumed that Vds was a constant 400V throughout trr. Vds is 400V up until time tb where it linearly decays due to the diode shutting off. The tricky part is calculating the loss for time tb, as some of you guys have already discussed. I've tried multiplying the two linear functions for i(t) and v(t) and then integrating over tb but that turns into some integration by parts practice followed by some terms going to zero causing an implosion.
Is solving the integral the way to go or is there some other approach?
The way I did it is when yo expand the v(t)*i(t) you get a quadratic the integration of which gives you cube and square terms which being very small I ignored and only used the t term.
Post a Comment