Detailed problems encountered in related projects of switching power supply and related solutions

**Project:** A power supply in a laboratory is broken. Upon inspection, it was found that the full bridge controlled by UC3875 needs to be repaired. **Phenomenon:** During the initial check, one of the power transistors was found to be damaged. Since there were no identical transistors available, we replaced them with similar power-rated ones. After powering on, everything worked fine at low input voltage, but when the input voltage increased, the drive signal became unstable and the frequency started to jitter. **Solution:** Increasing the drive resistance of the power transistor eliminated the issue. The power supply returned to normal operation, and the repair was successful. **Analysis:** The new transistors had different parasitic parameters compared to the original ones. This caused faster switching, which introduced more noise. At higher voltages, this interference affected the control circuit, leading to instability. **Notes:** 1. **Careful soldering is essential.** Avoid cold solder joints, as they are hard to spot and can cause serious problems. Also, ensure that components like diodes are placed correctly. I once accidentally reversed the direction of a bridge rectifier diode, which caused reverse voltage across the filter capacitor—this was extremely dangerous. 2. **Twist signal lines during debugging.** If you need to use flying wires for a round-trip signal, twist the signal and return lines together. Otherwise, they can act as an antenna, picking up unwanted interference. 3. **Use both bulk and high-frequency capacitors.** For bus power supplies, it’s not enough to have just a large filter capacitor. Adding high-frequency capacitors improves filtering performance, especially at the output stage. **Project:** UC3845 Dual Tube Forward Converter **Phenomenon:** After turning off both transistors, the voltage across their DS terminals was significantly different from what was expected. Initially, I thought the MOSFETs had mismatched parameters. I swapped the upper and lower transistors, but the result was the same. It seemed unrelated to the MOSFETs themselves. **Solution:** I adjusted the driver circuit to ensure both transistors turned off simultaneously. This slightly improved the situation, but the voltages still didn’t divide evenly. **Analysis:** There were two possible causes. First, differences in PCB parasitic capacitance between the two positions affected the DS voltage. Second, the driver wasn’t fully synchronized, leading to uneven turn-off times. **Project:** UC3845 Flyback with Auxiliary Winding Feedback **Phenomenon:** The main output voltage experienced a large overshoot during startup, but the auxiliary winding used for feedback did not. This discrepancy was puzzling. **Solution:** To adjust the response time, a resistor was added in series with the auxiliary winding. Reducing the resistor value significantly reduced the overshoot on the main output. **Analysis:** Since the feedback came from the auxiliary winding, adding a resistor created a voltage difference between the feedback point and the actual output. This difference caused the overshoot due to coupling through the transformer. **Project:** NCP1014 with Optical Feedback **Phenomenon:** A board that had already been working was re-soldered, and after that, the output voltage became incorrect. **Solution:** Replacing the 431 with another model of the same type restored normal operation. **Analysis:** The original 431 was from Zetex, which had a very low minimum operating current (in the microamp range). The design didn’t account for this, but when it was replaced with a TI 431, which required 1 mA, the circuit failed. **Project:** ICE1PCS01 Controlled Boost PFC **Phenomenon:** The input current waveform looked good across the full voltage range, with minimal high-frequency ripple. However, at around 220V, the ripple suddenly increased, while it was small both above and below that voltage. **Solution:** When using an AC source, the ripple was larger at all voltages, which made me laugh. **Analysis:** The voltage regulator used had leakage inductance. At 220V, the regulator was effectively bypassed, so the leakage inductance had no effect, causing the ripple to increase. **Project:** UC3845 Dual Tube Flyback **Phenomenon:** The drive signal was unstable, with constant jitter and a noisy transformer. Adjusting the loop didn’t help. Oscilloscope checks showed that the sawtooth wave on the UC3845 oscillator pin had jitter, even though the frequency was supposed to be fixed. **Solution:** I separated the ground and power planes of the control circuit and connected them at a single point. This stabilized the drive signal, fixed the frequency, and eliminated the noise. However, conduction worsened, which was strange. **Analysis:** Layout is critical in power supply design. Grounding should be carefully managed, separating power and signal grounds, and connecting them at a single point to avoid interference. High-frequency currents shouldn't flow through the signal ground plane, or they can disrupt the control circuit. **Project:** UCC3895 Current Mode Phase-Shift Full Bridge with Double Current Rectification **Phenomenon:** The transformer was biased. **Solution:** I redesigned the secondary power trace on the PCB, making it thicker and connecting it to the inductor of the current doubler rectifier. The magnetic bias disappeared. **Analysis:** The current doubler rectifier has a unique issue: the average current through the two inductors may differ. With current-mode control, the primary side is balanced, but if the secondary side isn’t, it can cause magnetization bias. The difference in DC resistance between the inductors and the PCB traces caused this imbalance. **Project:** 431 with Optocoupler Feedback **Phenomenon:** The output voltage regulation was poor, and the voltage dropped significantly under load. The measured sampling point was close to the output, yet the drop was noticeable. **Solution:** Adding a small capacitor between the reference pin and cathode of the 431 improved the regulation. **Analysis:** The reference pin of the 431 was being disturbed, which affected the feedback loop. **Project:** IR1150 Boost PFC **Phenomenon:** The switching frequency was 100 kHz, but the input had a 1 kHz ripple current. The X capacitor was also humming. **Solution:** Adjusting the EMI filter parameters resolved the issue. **Analysis:** The EMI filter was resonating, causing the ripple and noise. **Project:** Flyback Synchronous Rectification **Phenomenon:** The voltage peak on the synchronous rectifier was too high and couldn't be absorbed. **Solution:** Replacing the synchronous rectifier with one that had a fast recovery body diode solved the problem. **Analysis:** The long reverse recovery time of the body diode caused a large reverse current, resulting in sharp voltage spikes. **Project:** IR1150 PFC **Phenomenon:** During temperature testing, the MOSFET reached only 80°C, but previously it had gone up to 110°C. Something was wrong. **Solution:** After checking, it turned out the drive resistor was incorrectly soldered—100R instead of 10R. **Analysis:** A high drive resistance increases MOSFET losses. Even if the case temperature seems acceptable, the junction temperature could be dangerously high. Too much resistance leads to insufficient drive power, causing overheating. If the drive resistance is too high, the MOSFET will overheat and fail. A properly sized resistor ensures sufficient drive power, preventing damage. If the PCB trace inductance is too high, it can resonate with the Cgs capacitance, creating voltage spikes. Adding a resistor helps dampen these oscillations. **Project:** L4981 PFC **Phenomenon:** No-load power-on resulted in a weak drive signal and significant oscillation frequency variation. Higher input voltage led to stronger oscillation. I initially suspected poor grounding or PCB layout, but nothing worked. **Solution:** After careful inspection, I found a power line close to the control circuit, connected to the MOSFET's D terminal. I isolated the copper near the control circuit, turning it into a dead area. The interference disappeared. **Analysis:** The MOSFET's D terminal generated a large dv/dt, causing common-mode interference. Keeping the control circuit away from such points is crucial to prevent noise issues.

Customized Mono Solar Panel

Solar Panel

Mono Cell Solar Panel,Monocrystalline Module,Perc Mono Solar Panel,Monocrystalline Solar Panel Price

Wuxi Sunket New Energy Technology Co.,Ltd , https://www.sunketsolar.com