DC DC converter using H bridge topology

I have designed a DC DC converter ,
Specification :
input voltage : 10v to 14v
output voltage : 311v
power :400 watts
topology : H bridge
frequency 50KHz
core: ETD44
for driver circuitry i have used sg3525+IR2110(bootstap)
, the problem is when i connect a 100 watt bulb at the output , the output voltage decreases to 180v and the low side mosfet of left leg gets too hot n after 30 second its get damaged , ir2110 also gets damage,
if i put a 15 watt bulb on the output , the output voltage is regulated @311v , in this case also low side mosfet of left leg is getting hot(but not too much). i have replaced the mosfet and ir2110 , but the problem didn't resolve,
the other 3 mosfet of H bridge are not getting hot .why only one mosfet is getting too hot ?

Sounds like a hidden problem in your circuit that can't be understood from the general description. There's apparently some kind of asymmetry, either in the layout or the actual component values.

FvM, yes there is some asymmetry in the design , so do u think that heating of mosfet is due to asymmetric layout ? i am attaching the image of the layout , plz have a look n tell me if any layout design rule is voilated. Mosfet M3 is getting too hot n gets damage.FvM i have generally seen that there is a capacitor connected in series with primary winding of transformer , which is then connected to H bridge, i didn't use that capacitor, what is the purpose of that capacitor ?


one more question : i will implement the push pull topology to get 311vdc from 12vdc , but the problem is flux imbalance in the core which might result in staircase saturation, so i wanna ask how to overcome this problem , i have read about current control mode , n was planning to use sg3525 , so can SG3525 be used in current mode? if no, plz tell me any other IC which can be used for push pull and also easily available?

I don't recognize a particular reason for overheating of one transistor in the layout. There might be an effect like high frequent self oscillation of a MOSFET or a hidden component fault. Thoroughly monitoring voltages of each circuit node in operation with an oscilloscope is probably the best way to understand the problems.

A symmetrical H-bridge won't show transformer saturation without additional causes, I think.

A series capacitorcan be used to setup a resonant circuit or to block DC currents (preventing saturation in case of asymmetrical duty cycle). It's not practical for a low voltage winding.

Did you put gate to source resistor 1k .If you have done so your bootstrap capacitor value for IR2110 should be increased accordingly

Dear,

Have you verified the gate pulses of all the mosfets .i.e have you verified the boostrap circuit? The leakage inductance of M3 mosfets seems to be higher that's the reason it fails more often. plan with a series capacitor as you mentioned. and also the layout need to be modified for better results.

--BawA-- said:

i have read about current control mode , n was planning to use sg3525 , so can SG3525 be used in current mode? if no, plz tell me any other IC which can be used for push pull and also easily available?

Click to expand...

For currentmode control use SG1846, SG2846, SG3846