After the delivery of a previous design project, due to the operation error of the on-site commissioning personnel, the exposed wire end touched the circuit board (sprayed with conformal paint) when the rear panel of the equipment was disassembled, which eventually led to a short circuit. The on-site feedback was Say MCU does not work, and hear the sound of "tick tick" when the circuit board is electrified.

Intense troubleshooting begins...

Power supplySchematic diagram

The power supply is Shanghai Belling's BL8033 synchronous step-down switching power supply. The maximum input is 16V, the maximum output is 3A, and the efficiency is as high as 96%.

BL8033 ElectricalCharacteristics

Based on feedback it was a short circuit causing a "tick-tick" sound on the circuit board. Knowing this problem, it can be basically located that the sound appears near the power supply, and after further analysis, it can be determined that it comes from the inductor. So why does the power supply cause the inductor to make a "tick tick" sound when there is a load short circuit? Let's analyze it together.

Why does the inductor whistle?

High-power switching power supply short-circuit howling. When the switching power supply is working at full load, if the switching power supply is suddenly short-circuited, sometimes you will hear the howling of the power supply; or when the current protection is set, when the current is adjusted to a certain level, there will be howling. The reason is that when the load is close to the output power limit of the power supply, the switching transformer will work in an unsteady state.

In the first cycle, because the duty cycle of the switch tube is too large and the conduction time is too long, too much energy is transmitted to the subsequent stage through the transformer; the energy storage inductance of the DC rectifier circuit cannot The energy stored in the first cycle is fully released within 2 cycles; when the third cycle arrives, the power chip will not turn on the switch tube, or the duty cycle to turn on the switch tube very small.

In this way, the energy stored in the energy storage inductor is released after the second and third cycles, resulting in a decrease in the output voltage. In this way, when the fourth cycle comes, the power supply chip will drive the switch tube to conduct a large duty cycle, and the transformer will vibrate at a low frequency, thereby emitting a sound that can be heard by the human ear.

When the power supply is working in an unsteady state, the output ripple voltage is much larger than when it is working in a normal state. When the number of cycles in which the switch tube is fully cut off reaches a certain proportion in the total number of cycles, the switching frequency of the power supply enters the audio range from the high frequency range, resulting in a sharp howling. At this time, the transformer is already in a serious overload state and may burn out at any time.

When there is no load or the load is very small, the switch tube will also have intermittent full cut-off periods. The switching frequency of the power supply has entered the audio frequency range from the high frequency range, thus emitting a sharp howling.

In addition, in no-load or light-load scenarios, the back EMF generated by the transformer cannot be well absorbed, resulting in many clutter signalcouplingTo the primary and secondary windings of the transformer. Resonance occurs when the low frequency components of these clutter coincide with the natural oscillation frequency of the transformer. In order to prevent the resonant frequency from falling into the audio frequency range and produce howling, a frequency-selective loop can be added to the circuit to filter out low-frequency components.

The transformer is poorly dipped, including not dipped. When the transformer is poorly dipped in paint, although the load capacity is generally not affected, it will produce howling and the output waveform will have sharp spikes. It should be noted that when the design of the transformer is poor, it may also vibrate and produce howling during operation.

The ground wire of the primary regulated power supply chip is poorly routed. When the ground wire is poorly routed, the common manifestation is probabilistic failure (some products can work normally, and some products fail). The failure phenomenon is that the load cannot be carried, and even the vibration cannot be started. At this time, it is often accompanied by howling.

The ground wire of the secondary regulated power supply chip is faulty. The grounding of the reference regulator chip of the transformer secondary and the grounding of the primary power regulator chip have similar requirements: it cannot be directly connected to the cold ground and hot ground of the transformer. If they are connected together, it will cause the load capacity to drop and produce howling. The greater the load, the more obvious the howling.

How to solve the short circuit after locating the cause of howling?

The channel between the power supply output and the load can be disconnected if the conditions of hardware permit, and a "FB3" is used in the power supply designed above Magnetic beads, so it is very convenient to eliminate the short circuit of the power supply;

Check the configuration parameters of the power supply, and then re-power on the power supply to ensure that the output power meets the design requirements;

When the power supply problem is eliminated, the short circuit phenomenon still exists, so at this time, the post-stage load needs to analyze how to remove it to confirm the short circuit caused by which device (finally locate the short circuit of the single-chip microcomputer);

The method of eliminating the short circuit of the load, I usually start from small devices to large devices that are easy to solder, remove a point to measure a point, and measure the removed chip itself.

Pay attention to the solder balls on the board (soldering residue);

Why is there a short circuit after spraying conformal paint?

Three defenses are mainly: anti-mold, anti-humidity, anti-salt spray. So it can't effectively prevent the problem of short circuit. There are many components on the PCB that will generate heat when they are working. If the conformal paint is sprayed at this time, the heat dissipation of the device itself will be greatly reduced. In addition, the connectors on the board cannot be sprayed, which increases the risk of short circuit.

Of course, this PCB short-circuit accident should not have happened. Due to the complexity of the on-site debugging environment and the carelessness of the debugging personnel, this low-level problem finally appeared.