The solution of single chip inverter for electrostatic electromagnetic interference

Today I encountered the electromagnetic interference problem for the first time in my life. There are three fatigue machines of two types. The equipment uses solenoid valves to push the cylinders, and moves up and down through the Hall sensor limit switch. The single-chip microcomputer controls the equipment. The problem is that when the equipment has not completed all actions, it will stop automatically without pressing the stop button. The frequency of the problem is very high, but sometimes it is also good. So the single-chip microcomputer program is preliminarily ruled out. The wiring of the board is 90 degrees right angle, horizontal and vertical, without paving, and the control circuit and relay are on the same board. So it is judged to be interference between the relay and the wiring. When the machine is stopped, the stop light will flash. So the problem is judged to be interference that causes the single-chip microcomputer to stop io incorrectly. After careful observation, one solenoid valve is 220v and the other is 24v. There are emergency stop buttons and panel stop buttons on the equipment connected to the single-chip microcomputer stop io. After careful investigation, it was found that the wires connecting the solenoid valve and the emergency stop light were in a large wiring harness, and the two wires were connected to the aviation plug in a spiral winding state. When the upper and lower limit phases were reached, the relay would be attracted, which would cause very large interference. So the solution was to tear open the large wiring harness, disassemble the wires, plug in and test, and the problem disappeared. For the 24v solenoid valve equipment, since the aviation plugs were all together and there was no way to disassemble the wires, the solution was to connect a 474 capacitor to the io, plug in and test, and the problem disappeared.

Summary: I used to hear about electromagnetic compatibility from others, and I felt it was mysterious and scary. Today, I encountered it, but fortunately, with the help of my colleagues, the problem was solved. However, this problem tells us that when routing low-voltage lines and high-voltage lines, they must be separated and kept far away, otherwise, when they are entangled, the microcontroller will definitely be entangled. If the microcontroller is entangled, the boss will be entangled, and if the boss is entangled, he will definitely make you entangled. So don't underestimate EMI. Write it down to give everyone a warning.
 

There was another time:
 Today I adjusted the equipment, a control cabinet, which is composed of a frequency converter and a single-chip microcomputer, and a motor with rollers and belts on the outside, just like a treadmill. I have tested that thing before, and it runs normally. Sometimes it stops due to interference, but not frequently. There is a motor and a Hall switch on the machine. The motor is regulated by the frequency converter. The single-chip microcomputer is responsible for counting the frequency and number of times. The single-chip microcomputer sets the number of times, then counts and counts the frequency. When the number of times is reached, it stops automatically. The belt on the machine seems to be leather. In order to prevent the heavy object from falling on the belt, the belt is stretched. So under the belt, a nylon board is placed as a pad. As a result, when testing today, a heavy object was placed on it. As a result, the machine turned a circle and then crashed. It couldn't run normally at all. The interesting thing is that if the heavy object is not placed, the machine can run normally. When we lift the heavy object, the machine can run normally again. There is also a situation of stopping. When it stops, the machine will make a snapping sound at a high frequency. Then when the heavy object is taken away, this sound will be much lighter.

  Following the sound, we checked and found that the sound was near the belt. We used a screwdriver and a ground wire to find sparks near the belt, and found that the shutdown would disappear if the sensor was removed. Then I initially judged that it might be static electricity. Later, I asked and found that the original equipment had an iron plate under the belt, but the iron plate might scratch the belt. So this time, a nylon plate was placed. Nylon is not conductive and is plastic. The belt seems to be plastic as well. As a result, the friction between the plastics generated static electricity. Since there is no place for static electricity to be released, it will spark near the sensor. Because the sensor is a Jinshun surface and there is an iron sheet connected to the box, this will cause great interference. The interference is transmitted to the ground of the microcontroller through the metal head and ground wire of the Hall switch, causing the ground level to be unstable, causing discharge after one turn, and the discharge causes the machine to freeze. The final solution is to put a layer of insulating heat shrink tube on the metal head of the sensor, and use a plastic bracket to fix it on the limit to prevent static electricity from discharging with the sensor. As a result, the problem disappeared.

Although the shutdown process may not be what I expected, the reason for the shutdown must be caused by static electricity. So if the problem is solved by removing this interference source, it means that this judgment is correct.