In addition to the servo motor and variable frequency drive, how to break the electromagnetic interference problem? After reading it,

There are many problems associated with variable frequency drives and servo motors. We only focus on some of the issues. Readers can easily find variable frequency drives, bearings, overvoltages and electromagnetic interference on search engines to understand the full range of problems and find solutions.

In order to reduce equipment costs, the drive pulse inverter uses a fast switch to drive the pulse to have a rise and fall time of a few nanoseconds (Figure 2), thereby spreading the spectrum of these signals to a few megahertz. Simply turning on and off the current flowing into the motor is simpler, cheaper, and more efficient than producing a step-up and step-down output voltage. These short edge drive pulses are the main source of various problems. If the connection between the controller and the motor is an appropriate RF connection, for example, matching the input and output impedances, selecting the appropriate RF grade cable and similar design, many problems will not exist. However, the purpose of designing the motor is not to transmit the signal correctly, but to perform mechanical work, so its high frequency signal characteristics are rarely considered. The following are some of the problems caused by the spike drive signal or correlation (sequence is independent of importance)

Motor bearing damage

Overvoltage and associated insulation damage

High-level conducted electromagnetic interference in power lines and ground lines

Electrical overload (EOS) problem caused by high level electromagnetic interference current in ground plane

High level radiated electromagnetic interference from cables

Mechanical noise

Motor overheating

In the following sections we will consider the above and discuss ways to solve them.

Let us consider the motor as an electronic circuit. A steep drive pulse is applied to the stator (ie, the sensor of the motor base). The stator has a strong capacitive coupling to the rotor (the spacing between the stator and the rotor's large metal surfaces is very close).

Although the self-frequency of the drive pulse is very low - usually not higher than 20KHz, and is not the focus of consideration - the high-frequency component of the sharp edge of the drive pulse has a small impedance from the capacitive coupling, and now the high-frequency voltage of the motor and the edge of the drive pulse Synchronize. This voltage, in turn, causes current to flow through the only path from the rotor to the motor bearing to the ground plane. Figure 4 shows how the ground current is fully synchronized with the drive pulse edge.

The connection between the bearing ball and the bearing bottom ring is often intermittent, and the presence of insulating lubricants exacerbates this problem. This causes arcing of the current in the bearing, resulting in an EDM phenomenon - spark discharge. Basically, an electric spark will devour a small piece of metal at a time. This phenomenon occurs extensively on other metal parts that are difficult to machine. These metal parts have spherical bearings, and the mechanism of metal machining is basically the same. However, the purpose and output are quite different. EDM caused by electric sparks in bearings originates from small craters, or discontinuities caused by pitting, which can cause further discharges and hazards. A voltage as low as 200 mV on the shaft also produces sparks, although the induced voltage on the motor is very high - up to tens of thousands of volts. Once pitting occurs, it becomes a concentration point for further discharge. Since the drive pulse and the resulting discharge occur thousands of times per second per motor operation (see Figure 5), it will not last long and will not permanently damage the bearing. One of the most common problems is the bearing groove, or the racetrack-shaped arrester - see Figure X (ABB). This problem can spread and eventually cause permanent damage to the motor due to bearing failure. This problem is very common and does not show any signs of elimination.

If the output impedance of the motor controller, the input impedance of the motor and the impedance of the connecting cable are perfectly matched, then the drive pulse is completely a square wave pulse. However, the motor is not a radio frequency device, no one tries to perform impedance matching, and if people try to match it it will not work. Mismatch can cause ringing and overload.

Figure 6 shows very light ringing and overvoltage conditions, although in many cases the overvoltage easily exceeds 60% of the normal pulse amplitude. The red line is the direct drive signal at the controller output; the blue line is the same signal on the motor that passes the 3 foot (~1 meter) cable. It is obvious that this overvoltage and ringing will only increase the ground current through the bearings. According to the US Department of Energy [7] "The fastest rising pulse from a pulse width modulated inverter may cause harmful currents to flow in the bearing even if the overvoltage is not large." Overvoltage will not only cause damage to the ribs, but also oppress the isolation layer of the cable and the internal coil of the motor. In addition, it will cause voltage control drive current damage, motor overheating and noise problems, which does not include other non-serious effects. . This issue has received widespread attention. IEC/TS60034-25 (in conjunction with IEC/TS60034-17) states that the pulse voltage of a motor terminal using standard insulation shall not exceed 1350V. If the pulse voltage rise time of the motor terminal is less than 0.8 μs, the allowable pulse voltage drops to 900 V for a pulse having a 50 ns rise/fall time, as shown in FIG. NEMAMG1-2014 also emphasizes that electromagnetic interference can cause bearing damage and other problems.

It is not enough to focus on the motor damage or the overvoltage of the high frequency drive signal on the inverter/servo motor. The motors do not work alone - they are mounted on the device and these devices may be sensitive to electromagnetic interference generated by the motor. High frequency interference from the drive can cause:

Not compatible with EMC specifications

Interfering with the operation of electronic equipment

Cause measurement error and cause sensor output error

Electrical overload (EOS) for sensitive equipment

The generation of drive pulses causes a large change in the supply current consumption, which in turn produces high frequency conducted emissions, which in turn are returned to the power supply. The use of power line filters on power supplies is the primary method of compatibility design and is the recommended solution for most servo/frequency converter manufacturers. This helps to achieve electromagnetic compatibility. However, there is currently no EMC management department that controls electromagnetic interference within the device, as most management departments are more concerned with how a particular device may affect the operation of other devices. It is not possible to manage electromagnetic interference within the device, which can cause it to interfere with itself, especially when the user or integrator integrates the drive with other electronic devices. At this point, the interoperability of different devices is not strictly checked as if it were produced by a single company.

Most of the internal electromagnetic interference problems from the driver are caused by the driving pulse, and the radiation emission from the sharp edge of the driving second impulse appears in the form of noise on the device ground and the equipment frame. The induced conduction emission on the device trace is difficult to filter. except. As mentioned above, ground current through the motor bearing can damage the entire ground system of the device, reduce the signal-to-noise ratio in the data line, change the signal in the sensor, cause process changes, and sometimes even life-threatening conditions, such as NMR. Wrong reading. Some studies have shown that voltages as low as 1V between the neutral point and ground can cause equipment failure. Some capacitive coupling between the drive cable and the ground of the device can cause ground noise.

Electromagnetic interference from servo motors and frequency converters can also cause electrical overload. The large voltage difference between the semiconductor and printed circuit board mounting equipment ground exposes sensitive equipment to electrical overload, which often causes immediate or latent V damage that can cause the equipment to pass production testing but in practice It will soon be broken. The specific problem is that the electromagnetic interference voltage on the ground plane has a very low output impedance, which leads to high current power, and even a very low voltage difference can damage the device. The IPC-A-610 is the most basic standard for printed circuit board installation - limiting the voltage applied to sensitive equipment, especially the motor's electromagnetic interference transient voltage characteristics as low as 300mV. For electronic products, EOS is becoming more and more important in terms of production and reliability.

You can't control what you can't measure. This sentence is very profound. AC power conduction emission testing for electromagnetic compatibility purposes is well known and is well documented elsewhere and will not be repeated here. Moreover, the tests referred to in this section are not tests conducted by the conventional EMC management department, but are important for device reliability and operability.

Needless to say, directly testing the current through a rotating bearing is at least not a realistic attempt. However, testing the current in the return path of the drive signal is a very relevant test, for example, testing the current in the controller and motor ground, as shown in Figure 7.

The basic premise is that the current through the bearing must flow back to the starting point - the motor controller. The loop path is through the ground (sometimes designed as a PE-power ground). Although some are merely capacitive high-frequency current paths between the stator winding and the ground motor casing, which are insignificant compared to the current flowing through the bearing, the high-frequency current in the test ground is sufficient to estimate the current flowing through the bearing. Figure 7 shows the basic settings that can be used for servo motors and frequency converters. These motors typically have three drive lines, but do not preclude the design of U, V and W, and ground line G, or sometimes as a power ground. The broadband current probe on the ground returns current through the bearing.

Figure 8 shows a typical current through the ground, which is measured by the Tektronix current probe CTI. This probe has a sensitivity of 5mV/mA. The current tests in this paper are performed in this way. As shown, the peak current is 1.72A - the value is relatively large. This peak current is used for 10,000 times/s of bearings. Note that even without motion, the servo motor is still in operation - just keep its position. Not surprisingly, it can cause damage to the bearings, as well as other unwanted effects - disturbing equipment operation and causing electrical overload.

Here we enter the safety aspects that readers must pay attention to because the drive signal can have high voltage (up to 480V) high current power. Exposure to such voltages can cause injury or death. If you are unfamiliar with working in a high-voltage line environment, you should defer testing its connections, or at least with the appropriate training of experts.

You need a battery-powered oscilloscope with a bandwidth of at least 200MHz (there is no benefit from higher bandwidth) and a 100:1 high-voltage oscilloscope probe. Note that spectrum analyzers cannot be used, and AC-powered oscilloscopes generate ground loops due to their connection to a common ground. Moreover, it is noted that the conventional 10:1 probe does not adequately attenuate the drive signal to prevent clamping of the signal. If you set the value of the oscilloscope, you need to set its input to 1 megaohm instead of 50 ohms, because it is important that the high voltage and low impedance are not well matched. This may also introduce a small ringing to the signal, but this is better than damaging the oscilloscope.

In most motor controllers, you may find terminals labeled U, V, and W - they are output to the motor. Connecting the probe ground to the controller ground, the probe tip and the U, V and W terminals (Figure 9) may also need to be extended appropriately, depending on the termination structure. The ground of the oscilloscope probe should be connected to the terminal of the controller ground, and the terminal of the controller is next to the U, V and W terminals. Setting the measurement time based on the oscilloscope helps capture the drive pulse, the rising edge of the pulse and the falling edge of the pulse.

You may expect a square wave drive signal to appear on the high quality motor controller terminals. Testing on the motor is even more difficult, and it seems almost impossible to do this because of the terminal and some difficult problems, but it may be possible with the help of a device expert. Make sure that the ground clamp of the oscilloscope probe is connected to the ground of the motor, not elsewhere, otherwise the test results will be greatly discounted.

Finally, all of the problems described above are caused by the sharp edges of the drive pulses. Therefore, solving such problems requires changing the pulse edges so that both the rise and fall times are slow enough to prevent the capacitive coupling between the stator and the rotor from becoming the main conductive path while still ensuring the performance of the motor. Optimizing the traces, changing the current path through the bearing or blocking the path at the same time is some solution. The problem of pulse width modulation drive motors is not a new problem, but a wide range of problems - there are many solutions, which we will cover in later chapters. Given the breadth of this problem and its economic impact, there are many solutions available for motor users, and various solutions have different effectiveness. The precautions for inverter and servo motor problems are good - users are advised to pay attention to actual technical analysis and mass sales advertising when choosing a motor problem solution.

Undoubtedly, short motor cables provide less ringing and less emissions than long cables. Cables that are placed separately from other cables produce less voltage and current to the other conductors. A properly connected shield (copper mesh weave) helps to reduce the electromagnetic field from the edge of the pulse.

There are many specially manufactured cables for frequency converters and servo motors. A better cable between the controller and the motor is one that has adequate shielding of sufficient thickness (copper mesh braided shield). Some cables have separate ground lines, which further reduces ringing. It is not recommended to use cables that are not intended for conventional cables used in frequency converters as this may exacerbate the problem.