I wanted to modify the output voltage of a 20V3.25A notebook power supply board. First, I couldn’t understand how the circuit in the voltage feedback works. I just posted a post yesterday asking for help. I finally figured it out and encountered a new problem. It is the process of reducing 20V to 9V by replacing the sampling resistor of the op amp. Because the voltage is too low, the 3843 power supply under-voltage protection has become an intermittent output. Before the change, the 7-pin power supply pin was measured at 12.5V at a 0.5A load. It is estimated that the 7-pin power supply pin cannot reach the required minimum 8.5V because of the change too much.
Let me talk about how this board was sacrificed:
Since the voltage feedback is full of patches, in order to modify the 23.7k sampling resistance to a smaller value, I used a straight plug of 33k color ring resistor with two legs bent directly on its pad, and the output became intermittent. . In order to measure whether the voltage on pin 7 of UC3843 (also a patch) is undervoltage at this time, put the multimeter pen in the right hand on the measuring point of pin 7 and a hundred and eight thousand miles away, and hold the long resistance in the other hand. The feet are ready to connect to the two small pads. This difficult action is to save a little trouble without using the soldering iron that can’t connect to the ground…
I was thinking that my right hand should not be slippery. Before I finished thinking about it, the red test pen slipped out, only to hear a crisp sound, an electric light flashed on the side of the board, and then there was a muddy smell… I saw that there was no output from the load. , I guess it is slipped to the 8-pin 5V reference output pin, 7 and 8 feet are short together. I quickly unplugged the power to see that there was no trauma on the board, and there was no tripping in the house, so I plug it back in and test it again. It turned to ~600V. The PFC is intact. There is no voltage at the rectifier bridge. Make sure that the 2.5A250V cylindrical fuse is used. Ruined.
I thought that 3843 might not be broken, only the insurance was burned, so I took the alligator clip to connect the two pads of the fuse together, and plugged it in again, which brought a louder sound and brighter sparks. With the pungent smell, all the alligator clips flew some distance away! Turning over, a piece of the green NTC resistor sheath next to the cable inlet has been blown off, and the 2SK2843 tube next to the 2SK2843 tube in the switch circuit is covered with a heat shrinkable tube. The PCB underneath is also burnt, and the resistor is 0.27. Ω, burned. Measure that the pin of the gate, drain and source of the tube is all short-circuited. There is a short-circuit between pins 7 and 5 of the UC3843. The rectifier bridge will not be measured. The large capacitance should be fine. The pictures above are not clear.
This is the magnification of the fault caused by carelessness. Seniors, let’s talk about how to deal with it…
