You will have to re-wind or re-design and re-make your transformer for lower leakage - a larger core with a longer bobbin can help if you double up the wire, i.e. 2 in parallel, this gives longer winding(s) and closer coupling = lower leakage... I note the design is already interleaved (once) to lower leakage.

Hi Thanks for your replay ,but please let me know how to analyze the snubber issue with waveform how to read the waveform and how we need to decided which snubber components we need to use.I have used the attached datasheet from PI but it is not helping me out.

Link : https://ac-dc.power.com/sites/default/files/forum/files/PI-DG-101_ClampSizing_1.pdf

very simply the more leakage energy in the Tx at turn off (i^2.Llk) the bigger power the snubber has to be to limit the ringing produced by it. Thus the snubber cap may have to be bigger, the parallel R's may have to be a little smaller in value but with increased power capability and the blocking diode a reasonably fast type, or 2 x 600V in series. Tuning the snubber at lower powers (by substituting parts) will allow you to get to higher powers and final tuning of the snubber. You need to know that significant heating from the snubber can occur at fulll power for high leakage L and high peak currents in the Tx.

you should be aware that 10A out equates to 40A peak in the output diode when the pri side switches off, equating to 40 x 5 / 100 = 2A pk turn off on the pri side

2A into the 10n snubber cap is (i/v = dv/dt) = 200V/us, if you have pri side Llk of 30uH then it takes (V/L=di/dt) = 300nS for the current to fall on the pri side if opposed by 200V (spike volts above flyback volts) - and 300nS to rise to 40A on the sec side.

Now you begin to see how critical it is to lower Llk for good performance.

In 300nS the volts on the 10nF snubber cap will rise by ~ 50V, if it is not discharged enough between cycles, this can easily be 200+50 = 250V which is a lot like your scope shots.

100k & 10nF T = RC = 1mS so there will be not so great a discharge between cycles, so eventually the spike volts will rise until the peaks are clipped by the TVS - which looks to be happening.

So, moving to 22nF 250VDC cap, & 10k ohm parallel cap, will lower the turn off spike, the flyback voltage is ~ 6Vx100/5 = 120V, thus the dissipation in the 10k should be 1.44 watt, but is also set by the leakage energy 0.5 Llk 2^2 F = 4W for 66kHz.

So any way you look at it a lot of power is needed to be dissipated in the pri snubber for 2A peak and 30uH - if you want clean waveforms and reduced ringing at turn off.

You can also put a 10R (1206) res in series with the snubber blocking diode (adds 20V to turn off spike) to add damping to the ckt.

Or - completely redesign the Tx for lower leakage and re-do the pcb for much tighter layout... and double up the main output diode for faster turn on and lower losses...

EP.

Hi Easy Peasy,

Thanks for your replay ,

i saw a good results in snubber of my smps with this values but my doubt is can we change the transformer design if so please provide one design using piexpert and guide me what are the instruction i need to give to transformer winder to wind to get lower leakage inductance.

Glad to hear you now have "good results" perhaps you could share a picture of these with the group? In any event you say that you cannot change your transformer - you need a new longer bobbin and therefore larger Tx if you wish to lower the leakage - since you say this is not possible you are stuck with what you have.

Now we are ready to redesign the power supply,please find the attachment for the Transformer details,we are using EI35 core .
Please let me know which is the best transformer core suite for our requirement.
Please find attachment for the spec details.
If it possible please provide the one transformer design using PI software.

Dear Mr Mohan, if you could send a quick email to me, Dr Anna Moon, c / - anna.m(at)clear.net.nz then we can design a suitable transformer for your application with lower leakage, lower core losses, and lower wire losses. Kind regards.

hi easy peasy,

I have sent mail to you please check and provide a proper solution for transformer leakage inductance.

Um, nothing has come thru yet...

Can you please send your mail ID once again.
I have mail you to anna.mclear.net.nz

You need the @ bit in the correct place, look again...