As shown in the figure below, the PWM duty cycle is 50% and the magnification is 2 times. The ripple at point A is about 20mV, but the output ripple reaches 600mV. What is this?
Gena ポストする July 16, 2020
Point A is connected to a resistance to ground.
Hina ポストする July 16, 2020
It seems that the circuit is wrong. The feedback of R33 is moved to the front of R32, amplified first, and then RC low-pass, otherwise C is connected to the output and it is easy to oscillate.
Any design, try to use standard conventional circuit, do not take it for granted.
Buster ポストする July 16, 2020
Disconnect the u9d12 pin and measure whether the signal ripple at point A is still around 20mV. If the data is still at this time, it proves to be caused by the signal of the previous stage. If the signal ripple at point A is not below 20mV after disconnection, It turns out that it is caused by the interference of the subsequent circuit, and it can be further analyzed.
Geena ポストする July 16, 2020
Ripple test: 1. First determine the accuracy of the ripple test method?
Oscilloscope probe settings: 20M bandwidth, 1%, ground clip, etc.
2. After confirming that the test method is correct, you need to determine whether the ripple is from the input stage, output stage, op amp power supply, or the op amp itself.
