Forward converter FET issue.

[bowman1710]

Hi Guys,

Im am having issues with a forward converter with active clamp that i have modified from an existing linear demo board. The link is below of the board i have modified:

http://cds.linear.com/docs/en/demo-board-schematic/DC1929A-5-SCH.PDF

My spec is as follows:

Vin: 250-370V
Vout:24V
Power out:30W
Switching F:333Khz

The changed i have made made:

Transformer:
Core: EFD-20-3F3
Turns ratio:6:1
Primary:104T
Seconardy: 17T

Changes to the UVLO- for 250V turn on instead of 150V

Changed Current sense resistor to 25mOhm

Tied the FB pin to ground and comp pin to isolated compensation (see attached)

Completely removed all of secondary sync components and secondary side components, have added some rectification with snubbers , Lout =68uH and Cout =99uF

My issue is when i have the secondary side open (so no rectification, Inductor, Caps and isolated FB) i get 35V out at 250V in. But as soon as i connect the secondary side circuitry both FETs blow.

Any ideas why, I'm assuming possibly something wrong with my feedback loop?



- - - Updated - - -

Sorry the 10K should read to output as well

 

[BradtheRad]

As you connect the secondary, can you confirm what waveform the mosfets produce at that moment?

Is it narrow pulses (narrow duty cycle)? Or do they conduct continually so that they burn up?

 

[bowman1710]

I haven't managed to obtain the waveforms on the primary side MOSFETS, at the moment I am now winding a new transformer to produce a higher voltage on the secondary(to lower the duty cycle to see if that makes a difference), I will see if I can get the waveforms for next week, I had a feeling that both FETS might be on at the same time, but wasn't sure the reason behind it whether i was pushing the duty cycle to high or whether it was something to do with my feedback circuity.

 

[bowman1710]

Just an update Brad the Rad, I have re-worked the layout for the feedback circuitry and the the FETs have stopped blowing, I have got some measurements of the MOSFETS and they are fine now. The active clamp has enough dead time from the main switching MOSFET and the duty cycle is around 20%. The only issue I have now is that when I turn it on I get several switching cycles, the output ramps up to about 11V, then it stops switching. Any ideas?

 

[Velkarn]

where did you get such values of compensation network 68p 100k?
in active clamp forward working at 100 kHz i have compensation network 8k 33n.
try to increas capacitance. it is not dangerous - u just make system slowly.
one more advice. maybe try to supply pwm controller ic from separate lab supply, and than increase input voltage from 0 to nominal. but you have to change UVLO level (if it is activated in your shem).

 

[BradtheRad]

The only issue I have now is that when I turn it on I get several switching cycles, the output ramps up to about 11V, then it stops switching. Any ideas?

Try playing with the frequency, to find a range where the transformer provides the spec output. You may need different frequencies for your various input voltage.

Also, in my simulation experiments with flybacks, I have tried putting a full diode bridge on the secondary output. This is because current might try to flow during both halves of the cycle, rather than the second half. If it tries to flow during the first half, then you know something is not optimal. It might be:
* duty cycle
* turns ratio
* Henry value