1. The high-frequency oscillation stops for some reason, and then starts again, so there will be hiccups;

2. In case of output short circuit and other faults of switching power supply, the switching power supply chip works for hundreds of milliseconds after startup, then stops working for a few seconds, then starts working for hundreds of milliseconds, and keeps cycling.
      
 

　3. The single ended flyback switching power supply made of uc3842-3844 sometimes has the phenomenon of "burping". The reasons are as follows:
One is that the power load is too light, even if there is a small duty cycle, the voltage at the output sampling terminal will increase abnormally, making the voltage at the two pins of uc384x greater than 2.5V. So this kind of switching power supply is best to add the appropriate load during maintenance.
Second, the load is too heavy, and even the output terminal is short circuited. The winding supplying power to uc384x gets too little energy to provide enough current for uc384x operation, so uc384x operation current is mainly provided by starting resistance, and this current is relatively small, so intermittent operation phenomenon - "burping" appears.
The speed of "burping" is mainly related to the starting resistance and the size of 7-pin external filter capacitance.

4. Top chip power supply hiccups at low temperature - the equipment uses 12V power supply made by top 245 for low temperature test
Considering the situation of power supply burping, checking the data burping should be that the PWM controller produces protection conditions, mainly over-current or over-voltage or overload! This may be caused by short circuit, overload, breakdown of Zener diode and damage of optocoupler. So I checked these conditions in turn, but the power supply was still on for belch protection. In addition, because of the PI series tube scheme, it also checks whether the TVS tube at DS end of the top tube is damaged. Previously, it encountered the situation that TVS was too low in clamping level selection & lt; = 400, and broke down at high voltage, and then over-current burped all the time, but this situation did not appear on this board.
Continue to check the winding method of the transformer, and doubt whether he has reversed the homonymous end of the second secondary, the power supply is in positive use, the saturation protection, and carefully dismantled a transformer without reversing.

Check the feedback to see if the feedback is broken, because when the feedback is removed, the same hiccup occurs, and all the components of the feedback circuit are intact.