24V 3A 72W universal input adapter

Posted by: kemred on 12/17/2012

I have designed a 72W SMPS with TOP270 E sip package, It works well in over all range at first, low ripple, noiseless etc.. But after working in below 100V AC for a 2-3 minutes, it stops working for 20-30 seconds and starts working again for 2*3 minutes.. What can be the reason, I use a big heatsink, and it doesn't seem too hot,

I also want to ask why PI wants to stop TO-220 package production, it is very hard to conect heatsink to mosfet with clips, I am using the clips(connectors) send by PI distributor..

Comments

Would you like to increase the capacitane of input capacitor?

After some time, the efficiency of power supply is lower and lower, and it is harder and harder to maintain the regulation and thus causes the board shut down.

Maybe no TO-220 packacge is caused by not too much volume, but I will consult with some other guy to understand the TOP-220 package.

Regards

I also increase the input capacitance (first one is 150uF, design made for this setup) then I increase to 250uF. but top switch is too hot, I measured with thermocoupler and its surface temperature reaches 84 Celcius decreases and stop working until it reaches 64 Celcius... I thing internal temperature reaches the limit values..

I read the AN 14 fot troubleshooting in hot top switch and do the all suggestions written in this note but it doesn't seem any help..

I designed transforner for KP 0.6 so rms current is smaller,

I also changed the rectifier diodes with UF and 30A capacity ,

I use UF4007 for primary snubber,

I use a very big cupper heatsink...

I thing the main reason is the Rds of TOP270EG is around 1.8 ohm (very high) in 100 decreases junction... may be using an external mosfet with low Rds on will be better but I thing PI haven't any MOSFET driver IC...

I also designed a 36W SMPS 6 month ago and it is TO-220 PAckage HX series and it does not have such temperature problems,

Best regards,

Would you like to try TOP271? The condcution loss will be reduced about 20% and the switching loss should be a little bit worse. At steady state, the flux density of the transformer should be same as the board with TOP270. But you need to double check the Ids during start up to make sure the transformer is not saturate.