I need help on how to calculate turns in flyback transformer for the above mentioned power supply. Now I know about designing Half bridge, full bridge and forward converters but I am unable to figure out about flyback converters.
I indent to build one out of scratch even though I know I can buy one. I think I'll use it as aux supply for the variable half bridge SMPS bench supply I'm currently working on.
Here's a sample circuit although it seems to be crude and does not have an RCD snubber.
Figure around 50% duty cycle in BCM, i.e., primary peak volts are double the DC supply (consider the output voltage if the primary winding were used as a boost converter).
Flyback is more flexible than forward, so you have some freedom to choose this peak voltage. It can be higher or lower than the supply. Higher means more stress on the transistor (need higher Vds), higher primary duty and lower secondary duty. (Low duty is bad because it means all the input or output power is drawn during a relatively short time, making the RMS current in the filter capacitors that much higher.) Lower peak voltage means less stress (lower Vds rating), lower primary duty and higher secondary duty.
Because of duty cycle, you probably don't want peak voltage below 1.5 Vin, or above 3 Vin.
Because of Vds rating, a somewhat lower peak voltage is preferred, so we might aim for 1.5 to 2 Vin.
If the input is 360V and the flyback is 240V above that, we get a peak of 600V, and would choose Vds >= 800V (a commonly available rating). For say 12V output, we need a ratio of 240/12 = 20:1.
If you have the winding area, the primary should preferably be wound in a single layer over the secondary, or interleaved around the secondary (or the secondary around the primary, it doesn't matter). This keeps leakage inductance low. This is a priority for higher power supplies (>50W?), but not critical for smaller ones (5-20W?), and actually it swings the other way for low power ones (1W?) where capacitance instead dominates so a looser (multilayer?) winding is preferred. (The actual property being controlled for here is characteristic impedance, not just leakage inductance or capacitance but in fact the square-root of their ratio.)
Tim