Project: A power supply in a laboratory is broken. When you take a look at it, the full bridge controlled by UC3875 needs to be repaired.
Phenomenon: Initial inspection, the power tube is broken. Because there is no tube of the same type, all the tubes are replaced with tubes of the same power class. After power-on, when the input voltage is low, everything is normal. When the input voltage is high, the drive is chaotic and the frequency is jittery.
Solution: Increase the drive resistance of the power tube, the phenomenon disappears, everything is normal, and the power supply is repaired.
Analysis: The new pipe parasitic parameters are different from those of the old pipe. Under the same drive circuit, the switching speed will be faster, resulting in a larger interference. At high voltage, the interference is too large to affect the operation of the control circuit.
Simply write a few:
1, the component welding should be careful, can not occur in the virtual welding, the virtual welding is very terrible, and it is not easy to see. The direction cannot be reversed, especially the direction of the diode. I used to solder the direction of the bridge rectifier diode, which directly caused the filter electrolytic capacitor to add back pressure, which is very dangerous.
2. If the flying line is needed during debugging, and it is a round-trip signal line, the line and the return line should be twisted together. Because if the line and the return line form a surrounding area, it is equivalent to an antenna, and it is easy to string in interference.
3, the bus power supply must not only have a large filter capacitor, but also have high-frequency filter capacitors. The same is true for filtering at the time of output.
Project: UC3845 double tube forward
Phenomenon: After the two tubes are turned off, the voltages experienced by the DS are very different, not theoretically half of them. The guess is that the parameters of the MOS are inconsistent. The upper and lower tubes are welded down and the position is exchanged. The result is the same. It seems to have nothing to do with MOS.
Solution: Adjust the two-tube drive so that they can be turned off at the same time. The situation is slightly improved, but the voltage cannot be divided equally.
Analysis: This should be caused by two reasons. One is the difference in PCB parasitic parameters. The actual capacitance of the DS in the two positions is different. The other is that the driver is not very synchronously shut down.
Project: UC3845 Controls the Flyback of Auxiliary Winding Feedback
Phenomenon: The main circuit output voltage has a large overshoot at boot time. However, the voltage of the auxiliary winding participating in the feedback does not overshoot.
Solution: In order to adjust the adjustment rate, a resistor is connected in series on the auxiliary winding. The resistance of this resistor is reduced, and the main path output overshoot is significantly reduced.
Analysis: Since the feedback sample is the auxiliary winding, and the auxiliary winding is connected in series with a resistor, the voltage at the auxiliary winding and the voltage at the feedback have a voltage difference, and the output voltage overshoots through the transformer coupling.
Project: NCP1014, Optical feedback feedback
Phenomenon: The mature board that people have already done, after re-welding a piece, found that the output voltage is not correct.
Solution: Replace the original Bom 431 with the 431 of the same model as the other model, and change it back.
Analysis: Originally used is zetex 431, its minimum operating current is uA level, so the design does not consider the minimum operating current. Later, TI's 431 was replaced, and the minimum operating current was 1 mA, which caused the work to be abnormal.
Project: ICE1PCS01 Control boost PFC
Phenomenon: The full voltage range, when adjusted with the regulator, the input current waveform is very good, the high frequency ripple is very small. Only when the input voltage is around 220V, the high-frequency ripple of the input current suddenly becomes large. More than 220V, and less than 220V are very small.
Solution: Just use AC souce, the high frequency ripple is bigger under any voltage, haha.
Analysis: The auto-coupled voltage regulator is used. The self-tuning voltage regulation has a leakage inductance. The leakage inductance can filter the input high-frequency ripple current, but when it is 220V (network voltage), the self-tuning regulator output The end is actually connected directly to the input, and naturally there is no leakage.
Project: UC3845 double tube flyback
Phenomenon: The drive is unstable, the jitter is constant, and the transformer is noisy. The adjustment loop is useless. Use the oscilloscope to look at the sawtooth waveform of the uc3845 oscillatory foot and find that the sawtooth wave has jitter. The UC3845 is fixed frequency and appears to be interfering.
Solution: Separate the ground and power of the control circuit strictly, and then connect at a single point. The drive signal is stable, the frequency is fixed, and the transformer is not called. But the awful thing is that the conduction is actually worse. Perhaps the legendary frequency jitter is indeed good for conduction.
Analysis: layout is very important in power supply design, especially the layout of the ground, the power ground and signal ground are separated, and grounded at a single point. It is to avoid high-frequency power current flowing through the signal ground plane, otherwise it will interfere with the control circuit.
The ground of the IC, the land of MOS must be strictly separated, and then single point.
The auxiliary winding supplies power to the IC, so the ground of the filter capacitor of the auxiliary winding is formed separately and then grounded at a single point with the signal. Thus, the high frequency current on the auxiliary winding is absorbed by the capacitor without being strung to the signal ground.
Project: UCC3895 current mode control phase shift control full bridge, double current rectification
Phenomenon: Transformer is biased
Solution: Make a PCB power trace of the secondary power circuit bold. The PCB trace is connected to an inductance of the current doubler rectifier circuit. The magnetic field disappears~~~~
Analysis: The current-flow rectifier circuit has a unique problem, that is, the average current on the two inductors will be inconsistent. If current-mode control is used, the control signal will ensure that the positive and negative current peaks of the transformer primary are the same, then if the transformer secondary is positive If the negative current is inconsistent, it will cause a bias magnet to appear.
The average inductor current is inconsistent because of the difference in DC impedance between the two inductors. But in fact, the inductance of the same batch of ground is not so different. On the contrary, the PCB traces connecting these inductors are quite different, resulting in a large difference between the actual DC resistance of the two inductors (plus the resistance of the PCB trace).
Project: 431 plus light feedback back
Phenomenon: The output voltage regulation rate is very poor, and the voltage drops significantly with the increase of the load. The voltage difference between the measured voltage sampling point and the output pin is not large.
Solution: Place a small capacitor between the reference pin of 431 and the cathode. The adjustment rate has become better.
Analysis: The reference foot of 431 is disturbed.
Project: IR1150 boost PFC
Phenomenon: The switching frequency is 100K, but the input actually has 1Khz ripple current. The X capacitor is also screaming.
Solution: Adjust the EMI filter parameters.
Analysis: The EMI filter resonates itself.
Project: Flyback Synchronous Rectification
Phenomenon: The voltage peak of the synchronous rectifier is very high, and it can't be absorbed.
Solution: Replace the synchronous tube with a tube with a fast recovery body diode
Analysis: Since the reverse recovery time of the body diode of the sync tube is too long, a large reverse recovery current is caused. Causing sharp voltage spikes
Project: IR1150 PFC
Phenomenon: When the temperature is tested, the temperature of the MOSFET is only 80 degrees, and the chicken is fried. In the previous few, the temperature of the MOS case reached 110 degrees, and everything was safe.
Solution: Find out the reason, the drive resistance welding is wrong, originally 10R, the result is welded to 100R.
Analysis: The drive resistance is too large, resulting in a large MOS loss, the same junction-to-shell thermal resistance, and large power dissipation can cause large temperature differences. Although the shell temperature is only 80 degrees, the actual junction temperature has exceeded the MOS's tolerance.
If the drive resistance is large, the power of the drive will be seriously insufficient, and the tube will be hot!
If the driving power is large enough, it will not fry the chicken.
If the inductance caused by the PCB trace is large enough, it will resonate with the capacitance Cgs of the GS terminal of the MOS, and the peak will be superimposed on the drive signal. In severe cases, the fried chicken will be caused. The resistance is added to attenuate the oscillation.
Project: L4981 PFC
Phenomenon: No-load power-on, the drive is inconspicuous, and the oscillation frequency changes significantly. The higher the input voltage, the more powerful it is. I began to think that the ground wire was not laid, the PCB was cut and cut, and it could not be solved.
Solution: Look carefully at the PCB and find that there is a power line that is close to the control circuit. The power line is connected to the D pole of the MOSFET. It is useless to block the power line and let the power current go away from the control circuit. The PCB copper wire close to the control circuit is made into an island, making it a dead copper, and the interference disappears.
Analysis: Electric field interference, the D pole of MOS is a large dv / dt, resulting in a large common mode interference. So the control circuit should be as far away as possible from this point.
Bakelite Material,Bakelite Plastic,Bakelite Made From,Bakelite Synthetics
WENZHOU TENGCAI ELECTRIC CO.,LTD , https://www.tengcaielectric.com