The best secondary winding solution for TOP268KG

The operation conditions of both output diodes and the secondary windings are almost identical, and thus the load sharing between them should be better. For the second option, it is hard to guarantee the load sharing, since thermal of output diode, and flux coupling of the two windings could be different. As you know, any difference may cause huge difference for lcurrent sharing.

regards

I would like to know if you aver finished this design and what your efficiency was for the design. We are making a similar design but our efficiency is only 50% at 36V 1A... can you recommend a transformer design or maybe share your PI designer files... I can give you our PI designer files if requested.

Any help will be much appreciated

Hi mvdeventer,

Can you kindly upload your PI designer files?

Thank you!

Regards

Hi PI

After a lot of investigation we realized that we were measuring the efficiency wrong. We were calculating efficiency with ( V(AC)in x I(AC)in )/ ( V(DC)out x I(DC)out ) instead of ( V(DC)in x I(DC)in )/ ( V(DC)out x I(DC)out ). Our
efficiency is actually 82% to 87%. We realized this after reading the attached PDF document. Attached is also our original PI design.

What we did discover though that double copper on primary and secondary helps a lot for less heat dissipation in the transformer and the E25 CORE when operating constantly at MAX Voltage and current of 36V 1.3A. We had about a 20 degree celsius in drop for the double copper transformer. Also choosing a bigger output diode with a lower forward voltage helped with lower output diode temperature. We also had Output Capacitor sweating a lot and realized that we should rather have used -40 to +105 VERY LOW ESR (<100mOhm) degree celsius CAPS and not -40 to +85 ones.

Operating in an open frame enclosure still causes the average temperature to rise with about 30 degree celsius above ambient and very quickly you start operating at 85 degree celsius, which is too close to the Output Capacitors max operating temperature.

We also had to choose the drain protection DIODE to have a very fast recovery time with a Trr of 20ns (ES3D)as the original diode we used had a Trr of 200ns and this caused the diode to overheat and desolder or burnout or failed open or closed. Special care must be taken with this diode as the juction temperature of this diode of not design correctly can quickly reach 165 degree celsius and will destroy PCB all in its path...

I know the power supply is not designed to operate at MAX Voltage and Current the whole time but we had a condition in the field that caused the system to reset constantly and that in turn caused the power supply to run at Max power output for about 90% of the time.

Hope these suggestion will help any future PI designers.

Thank you mvdeventer!