LLC converter reverse current flow with secondary gate drive

HI,

I am using ICE2HS01 LLC controller for converting 300V to 50V @30A using half bridge LLC converter.
I am driving the secondary FET s with the gate driver for channel conduction to reduce the power dissipation.
IC has secondary gate pulse signals which is not giving the control signals at higher switching frequency of 1MHz.So,I given primary gate drive control to both primary and secondary FETs through separate isolated gate driver.
We had a failure in Primary top and bottom FET.
During start up,while probing the resonant current,the current waveform is not as intended and the SW node waveform resembles the wave form of capacitive region.
My converter will never operate in capacitive region because my minimum switching frequency limitation is kept above the resonant frequency.
Due to the channel conduction of secondary without sensing its diode current.I thought that it can be due to current from the secondary side which is changing my resonant current waveform.
So,I disabled the channel conduction and made it operate in diode conduction alone and I verified the resonant current waveform.
Now the waveform looks proper and the rise time is faster(50us) than later(200us) at no load.
Can anyone tell me the reason for current reversal from secondary to primary during start up.

With regards,
P.Selvam
P.Selvam.

Selvam_Perumal said:

We had a failure in Primary top and bottom FET.

Click to expand...

You're probably aware that inductors often generate high voltage spikes when their current is shut off abruptly. Your power level is upwards of 2kW. Do you have a snubber network designed to absorb spikes at such power?

LLC converters are a topic in several threads here. You may discover a schematic of a successful design which operates at your specified power levels. Try a forum search on:
'resonant LLC'
'LLC converter'
'LLC smps'

Your 1MHz switching frequency sounds too high.
You will get too much switching loss, because LLC still has some switching loss....due to the magnetising current being curtailed at switch off.

- - - Updated - - -

You say you are getting power reversal, and this sounds like your secondary sync rectifiers are not sync'd properly....and at 1MHz i am not surprised that maybe they are out of sync enough to cause you problems.

I find it unlikely that its appropriate to drive the secondary fets with gate drive signals from the primary.

And the fact that it works in diode mode alone is evidence of that.

In LLC half bridge resonant resonant converter,During start up,the magnetising inductance will not come into resonance due to short or low impedance load.
During above condition,Will the resonant current change its direction during the dead time since there is no magnetising inductance to clamp the current to zero or less.

asdf44 is correct, secondary timing is everything especially at light loads, start up and overloads, it is not straightforward or that easy to get sec side sync rect on an LLC - there is considerable phase delay between pri mosfet signals and those needed on sec side and this changes with load ...