In the simplest relay circuit, the output socket is not connected to any load, and there is no contact sparking or interference observed. To begin with, the system uses a 5V total power supply, provided by an AC-DC power module, while the microcontroller runs on 3.3V, supplied through an LDO. The microcontroller's IO ports directly drive a transistor, and are set to push-pull mode. There are two boards in the system: one is the main control board, which contains the microcontroller and an LCD screen, and the other is the power board, housing the AC-DC modules, relays, and output sockets. These two boards are connected via an FPC cable.
Interference issues arise when the 220V AC power is turned on. The relay toggles on and off instantly, and there's a 50% chance that the system crashes. During these crashes, an oscilloscope detects a pulse interference on the 5V power line. Additionally, the power indicator on the power module flickers, indicating a significant voltage drop. Several tests were conducted:
1. A 22uH inductor was added in series with the relay coil, but the system still crashed.
2. A Schottky diode was placed in series with the relay coil, along with a 1N4007 in parallel, but the system still crashed.
3. Multiple attempts were made, yet the crash persisted.
However, when the relay contacts were disconnected from the 220V AC (as shown in the image), the pulse interference on the power supply disappeared, and the system no longer crashed. Replacing the AC-DC power supply module reduced the pulse interference but did not eliminate the crash entirely. This suggests that the problem may lie in the relay coil or the PCB’s EMC design.
The "Big Ant Master" mentioned that the best approach to solving EMI issues isn’t just to eliminate interference but to understand how it affects the system. Otherwise, passing EFT testing would be impossible. Everyday devices like lights and hair dryers generate similar pulses, so the system should be able to handle them.
Uncle T suggested that the interference might come from either the back EMF of the relay coil or the inductive kick from the contacts. He recommended adding small resistors in series with the 1N4007 to help dissipate energy and prevent oscillation.
Master Chunyang believed that poor insulation between high-voltage and low-voltage circuits could be the root cause. Another possibility is that the system’s EMC performance is inadequate, making it vulnerable to electromagnetic interference.
To test this, he advised using a different relay, connecting its windings with short wires, and keeping the contacts isolated from the PCB. Then, after powering up, manually connect the relay contacts to the AC line and observe if the system crashes. If it doesn’t, the issue may be with the PCB’s leakage or insulation.
If the power supply has a Y capacitor, grounding must be properly done. Otherwise, the system may float at a high voltage, leading to instability when interference occurs.
A user named Yanruiqi asked why disconnecting the L terminal eliminated the interference. It was speculated that even without a load, the relay’s contact movement could create a magnetic field change or capacitance effect at the contact point.
It was recommended to try a different relay or isolate one contact from the PCB to prevent leakage between the two poles of the socket.
From the perspective of the phenomenon, it could be that the power supply has a weak transient response, and the relay requires more power than expected during activation. Using a DC24V relay instead of a 5V one might reduce the current draw.
In general, driving a relay directly from a microcontroller is not ideal for industrial use. Many engineers prefer optocouplers or driver ICs like the TLP127, which include built-in flyback diodes and can better handle inductive loads.
User Hu Fengwei suggested that the circuit design itself might be flawed, and that isolation between the relay and the microcontroller is essential to prevent interference from reaching the sensitive components. Without proper isolation, even small transients can cause the system to crash.
In summary, while replacing the power supply may temporarily solve the issue, the real solution lies in improving the circuit design—especially in terms of EMC, isolation, and relay management.
Solar Panel,,High Efficiency Mono Solar Panels,Bifacial Jinko Solar Panels
PLIER(Suzhou) Photovoltaic Technology Co., Ltd. , https://www.pliersolar.com