During the debugging process of PLC projects, various unusual problems may arise. Signal interference is one of them, and identifying the cause of such issues can be difficult. Today, I will share two cases: one involving interference from a frequency inverter to a PLC analog signal, and the solution to overcome such interference using a signal isolation module.
Phenomenon Description
A Siemens PLC outputs a 4-20mA current control signal from an AO (analog output) point to a Siemens frequency inverter. However, the inverter fails to start.
Fault Diagnosis
Suspected issue with the analog output board. Using a multimeter, I measured the 4-20mA output signal, which was found to be normal.
I then suspected an issue with the input side of the inverter’s control signal. Replaced the inverter with the same model, but the issue persisted.
I used a handheld signal transmitter to simulate a 4-20mA output and sent the standard current signal to the inverter, which then started. This ruled out faults in both the analog output board and the inverter.
It was deduced that interference from the inverter was being transmitted into the analog channel.
To verify, I installed a signal isolation module (TA3012) in the PLC’s 4-20mA output channel. The input terminals 5 and 6 of the module were connected to the analog output module, and the output terminals 1 and 2 were connected to the inverter, with terminals 3 and 4 receiving external 24VDC power. After this, the inverter started normally.
It was concluded that the root cause of the problem was interference from the inverter into the analog channel.
Precautions
In control systems where PLCs and frequency inverters are used together, the following points should be observed:
The PLC power supply and the power supply of the motor system (inverter) should be separately configured. The PLC power supply should use an isolated transformer.
Power cables should be kept separate from signal cables, and signal cables should be shielded.
Signal isolation modules should be used for all analog signal input and output channels.
Implement software filtering design in the PLC program.
The signal ground and power ground should be separately designed.
I recently saw an analysis and solution for an analog signal interference problem. In our actual applications, we frequently encounter similar issues. Here is a shared experience:
A workshop has 10 250kW motors with high-pressure pumps. The frequency inverters (Schneider ATV71) are connected to a Siemens PLC (S300) via Profibus DP. The pressure transmitters are Siemens models, with the 4-20mA signal being transmitted via shielded cables to the PLC.
After successful commissioning, everything ran normally for a week. However, after the vendor left, the 8th pump showed an error. When the set pressure was 40 bar, the actual reading was 70 bar. When the set pressure was 80 bar, the actual reading was 110 bar. Initially, the sensor was suspected to be faulty, but after swapping it to another pump, it worked fine.
The issue then spread to other pumps (3, 4, 5, 6, 7, 9, and 10), all showing similar problems. It was suspected that the frequency inverter’s interference was affecting the pressure transmitters. The manufacturer suggested adding a metal conduit shield. However, considering the installation difficulty (the control room is more than 30 meters away from the motors, and the entire wiring runs underground), I believed the frequency inverter’s harmonic interference should cause pressure fluctuations rather than a linear increase in the readings.
At first, I suspected a programming issue on the inverter since the display always showed 40 bar, but the inverter was outputting a frequency corresponding to 70 bar. The manufacturer disagreed, stating that they were using a standard Siemens PID block.
After much thought, I discovered that the negative terminal of the sensor was connected to the shield, which was also connected to the PLC’s analog input terminal (M). After disconnecting the shield from the negative terminal and grounding it, the fault was resolved.
Conclusion
The 2-wire pressure sensor uses a positive terminal powered by the PLC’s 24V supply, and the negative terminal outputs the 4-20mA current. When the shield and negative terminal were connected, the induced electromotive force on the shield created a current that entered the PLC input, resulting in a combined current that caused the pressure reading to increase linearly.
However, the same problem occurred again after a few days. In fact, one pressure sensor still showed 40 bar even after being removed. A thorough check revealed that the negative wire on the PLC input side was stripped too long, causing a short circuit and leading to signal interference between channels.
It reminded me that during the initial debugging, the vendor asked if the equipment ground and the cabinet ground were at the same potential. The pressure sensor shield was grounded at both ends, which led to severe interference and made the readings impossible to display. At the time, I suggested single-point grounding, and the vendor agreed. Now, I realize that the short circuit was caused by the current output side of each sensor being connected through the shield, and removing the ground from the sensor side resolved the issue, as the signals returned to normal once the shields were not connected together.