Problems with LLC resoant converter

[eem2am]

Hello,

I think i have an SMPS that cannot be done with an LLC Resonant converter.
It is as follows:

Vin = 385VDC
Vout = 80V
Iout = 4A


I tried to design an LLC Resonant converter for the above spec.......i wished to use an external "leakage" inductance.
-Therefore i used the design procedure of pages 1 to 3 of the following.....

http://www.fairchildsemi.com/an/AN/AN-4151.pdf


However, no matter what values are chosen for frequency, Cr, Lr, Lp, ........I cannot get a workable design.

The problem is that the resonant frequency "f0" occurs at the peak gain frequency.

...this is a massive problem because:-
1....You need to operate at the resonant frequency "f0"
2....You need to operate significantly below the peak gain frequency (otherwise there is a danger of going into the capacitive region where hard switching occurs)

...so now you can see the problem........i need to operate at the resonant frequency, but cannot because its the peak gain frequency.

Have i picked a load and input voltage that can't be served by a LLC Resonant frequency?


**********************************
Here is the gain vs frequency graph from equation 6, page 3 of the above...

http://i44.tinypic.com/hu15s4.jpg

.....the peak is also "f0", no matter what components i choose.

 

[mag]

What is Lmag/Lr ratio?
What is your turn ratio?
There is a very complete app note from ST for their L6599 ic which is very well done and explain everything. Check this out

 

[eem2am]

Thanks.....actually i have found some values which seem to get it not bad

Fsw = 107KHz
Lr = 100uH (not integrated)
Lm = 500uH
Ls = 96u
Vout = 80V
Iout = 4A
Vin = 385VDC
Cr = 22n

here is gain vs freq graph

However, i would still like to improve it and get the peak gain frequency a lotl lower than the "f0" resonant frequency.....for safetys sake and robustness