problem in design DC DC converter 12V - 310V

[Tankesh Rajput]

Hello,
I am trying to make DC - DC Converter using SG3525 ic. which circuit diagram i mentioned. its work perfectly when the load is under 60W, when i increase load. transformer create noisy sound and waveform of mosfet gate is change ( frequency increase and waveform is unstable). i use a ferrite core transformer which i take from computer power supply. its a EE type core which core area is 0.785 cm2. transformer calculation is on attachment. primary wire is 0.6mm *10 wire and secoundry side is 0.45mm wire *1. please tell me what changes i need to increase the output power to 350W. i use irf3205 Mosfet instead of irfz44.

[schmitt trigger]

Do you know for sure, that the unknown magnetic core will sustain 1700 Gauss without saturation?

[Tankesh Rajput]

I dont have datasheet of core. most of website i see that the guass value is in between of 1300G - 2000G. so i assume that its not saturate in 1700 Gauss.

[Tankesh Rajput]

I also try to 3 other ferrite core with same design but problem remain same.

[Tankesh Rajput]

I think that i use the wire at primary side to handle to capable to more than 60w so i try the 12mm (width) copper insulated foil instead of copper round wire. but wonderfully its increase my problem. its consume 26W power when no load is connected at output. 

[Conrad Hoffman]

Just some general things I learned when doing a large step up long ago. First, app notes and online info tends towards low ratio step ups. They can mislead you for large step ups. Odd things can happen with larger ratios and transformer construction may require "tricks" like multiple layered windings and Litz wire. Make sure insulation is sufficient- single build may not be enough. Make sure to measure coupling and such at the frequency of interest. I always had to include a small gap in the transformer to prevent saturation. Remember that saturation isn't a sudden thing, but comes on gradually. Layout is important- keep current loops very short. Even with a ground plane you can have ground loop problems, so  single point grounding of some traces may be necessary. Be sure to look at the ESR of input caps along with output caps.

[strawberry]

insert 0.1ohm resistor in between ceter tap of the transformer and B+ and run MOSFET at max Duty cycle and show some waveforms

is there any gap in core?

copper foil gave you bad resoults doe to winding centertap asymmetry, therefore saturation and leakage inductivity is high

[boB]

Push-Pull topology can be kind of picky from duty-cycle volt-seconds differences on opposite sides of the core.  Make sure that left and right sides are very close.

One thing you might want to try, if you have not already is to wind both primaries in hand so that the length of each primary half is equal.

This may be what Strawberry was eluding to.

[nctnico]

This looks like a forward converter to me. The transformer shouldn't have an air-gap. A simple approach to a forward converter is to use the same equations as a simple step-up converter. The transformer merely transforms the duty-cycle. The energy is stored in the secondairy inductor which should be able to handle the current swings. This should give you an idea of which currents you are dealing with.

When it comes to designing the transformer: there are many ways to design it. Based on core loss, based on copper loss, based on transformer bobbin size... In general more windings means more copper loss and less core loss (the flux swing gets lower). AFAIK a good approach is to take the peak current (primary or secundairy; doesn't matter) and calculate the number of turns to get near the core saturation point. Then calculate the core and copper losses. If the core and copper losses are acceptable then wind a prototype. You can probably put the formulas into an Excel sheet and play with it a little bit. It isn't an exact science.

BTW I'd use a push-pull driver for MOSFETs.