Arkori Qingqing
Jiangsu Ankerui Electric Appliance Manufacturing Co., Ltd., Jiangyin, Jiangsu 214405, China
Zhao Xuelian 1 Shen Biao 2
(1. Qinghai Sanjia Engineering Design Consulting Co., Ltd., Qinghai 810000)
(2. Ankerui Electric Co., Ltd., Jiading 201801, Shanghai, China)
Abstract: An Insulation Monitoring Device (IMD) for industrial ungrounded systems is introduced. According to the deficiencies of the existing technologies, a new hardware platform is provided to monitor 400V AC/DC non-grounded systems and details Insulation monitor hardware and software design principles. At present, the insulation monitor has passed the test verification and is sold in large quantities in the market, providing reliable insulation monitoring for industrial ungrounded distribution systems.
Keywords: AC/DC ungrounded system Insulation monitoring device Adaptive IMD
Â
0. Preface
In places where the continuity of power supply is relatively high (such as: mines, chemical plants, glass factories, metallurgical plants, safety lighting for certain assembly sites, and test equipment for certain electric furnaces, etc.), equipment failure will bring huge power outages. The loss, therefore, the use of ungrounded systems can effectively reduce the frequency of power outages. This is due to the fact that when an unearthed system first experiences a ground fault, the system can continue to be used without power failure. Ground fault is caused by human beings, but basically no harm to the human body, but at this time the system has hidden safety problems. If the fault is not eliminated in time, when the out-of-phase ground fault occurs again, the system may be powered off, resulting in Serious consequences. The installation of the insulation monitoring device can display the insulation resistance of the system to ground in real time. When the system first experiences an insulation fault, an alarm signal is issued to prompt the maintenance personnel to troubleshoot the system. The system does not need to trip in a short time, thus ensuring the IT system power supply. The reliability and continuity [1]. According to Article 7.2.3 of JGJ 16-2008 "Design Code for Civil Building Electrical Appliances", IT distribution system must be equipped with an insulation monitor [2]. Foreign countries also attach great importance to this. In the 1960s, various developed countries have begun to study the power system, but its rapid development was in the 1970s and 1980s. During this decade, the integration of digital circuits, the rapid development of computers, and the appearance of various types of sensors have promoted the development of electronic measurement. At present, some domestic manufacturers are paying more and more attention to the research of insulation monitoring products. The mainstream measurement methods include DC signal injection method, AC signal injection method, balance bridge measurement method, and so on. The above measurement methods have their own advantages, but due to the differences in the environment of the application site (the presence of leaked capacitors, DC signals, etc.), there may be a narrow measurement range, low measurement accuracy, and low allowable leakage capacitance in the system. The disadvantage of long measuring cycles, which can only be used in AC systems. This paper proposes a design principle of a new type of insulation monitoring device. The device adopts the method of adaptive system frequency. It has a wide range of insulation resistance measurement, allowing the system to have large leakage capacitance, fast response, and short measurement period.
1. The principle of insulation monitoring device
Figure 1 shows a schematic diagram of the measurement circuit:
.png)
Figure 1: The principle diagram of the insulation monitoring device
In Figure 1, R1 and R3 are equal resistance coupling resistors, R2 and R4 are equal resistance sampling resistors, Rf is system ground resistance, Ce is system leakage capacitance, and G is signal generator. The live conductor of the power supply is not grounded and is only used as the protective ground of the device housing. The insulation monitor injects +20V and -20V pulse signals into the system through G and returns to the insulation monitor through R1, R2, R3 and R4 to form a closed loop. Signal processing and acquisition of R2 and R4 voltages can be used to calculate the system. Ground resistance and system leakage capacitance.
2. Hardware design
The hardware circuit of the device mainly includes a central processor module, a disconnection monitoring module, and a signal injection module. The central processor chooses the ARM cortex-M3 core MCU. The chip has high frequency and abundant peripherals, which greatly simplifies the design of peripheral circuits.
The following discussion of the hardware circuit:
2.1 signal control circuit
The CPU determines the output of the signal by controlling the analog switch. Among them, the +2.5v signal is from the reference chip, -2.5v is obtained after inversion by +2.5v, and then enters the signal generation circuit.
2.2 signal generation circuit
The +2.5v or -2.5v signal in the signal control circuit is amplified by the high voltage op amp to generate a +20v or -20v pulse signal, which is the signal injected into the ungrounded system.
2.3 Signal Detection Circuit
The ±20V signal in the signal generation circuit forms the loop through the coupling resistance in Figure 1 and the insulation resistance of the system to ground. The insulation resistance of the system is calculated by detecting the signals of two sampling resistors; the PE/KE is determined by detecting the signal voltage on the PE. Whether the line is broken or not; during the operation of the device, the system type is detected in real time, and an appropriate measurement method is selected according to whether there is a DC component in the system.
2.3.1 AC System or Offline Status
The signal flows from the sampling resistor through a low-pass filter circuit with a cutoff frequency of less than 10Hz. When the system is an AC system or is in an off-line state, the interference signal is mainly from the 50Hz signal of the ungrounded system, and the frequency is much larger than the cutoff frequency (less than 10Hz) of the filter, the interference signal will be attenuated to Ignore the amplitude, and then add, amplify, and raise the two signals through the signal processing circuit, and finally be sampled by the single-chip ADC.
Filter effect can refer to simulation results. The circuit is simulated in PSPICE. A 300V (frequency 50Hz) voltage is applied between L1 and L2 (analog ungrounded system). The comparison of the signal before and after the fourth-order low-pass filter circuit is shown in Figure 2. The waveform in Fig. 2 is the result of superimposition of the injected ±20V and 300V system voltages. It can be seen that the 300V voltage has a great influence on the signal voltage on the sampling resistor. Referring to the lower graph of Fig. 2, it can be seen that after passing through the low-pass filter circuit, the 300v (frequency 50Hz) signal is attenuated to a negligible amplitude.
.png)
Â
.png)
Figure 2. Signal comparison before and after filtering
The two signals in Fig. 2 are the results of cross conversion of +20V and -20V, respectively. Since the system has leakage capacitance, the waveform shows a slow charge-discharge curve. This process is also a process in which the sampling resistor voltage tends to be stable. The final voltage on the voltage dividing resistor is only related to the system voltage and the proportion of the system voltage. It is not related to the capacitance. Therefore, the resistance measurement is related to the positive and negative half-cycle stabilized voltage. The following is a brief description of the calculation process:
.png)
Figure 3. Two-way signal synthesis
Suppose the voltage of “ADC_R†(sampling voltage) in Figure 3 is stable after the voltage is V1. At this time, “VOUTF†is at the voltage V2, and “VOUT1†and “VOUT2†voltage is at V3. At +20V, there are:
.png)
1
V1 and V2 (lift voltage) are known and V3 can be found. Set the sampling resistor voltage to V4. Since only the low-pass filter circuit and a signal lifting voltage V6 from V4 to V3, the low-pass filter circuit has little effect on the signal amplitude, then:
.png)
2
V4 is also the partial pressure of R2 and R4 in Fig. 1. With the power supply voltage V5, then:
.png)
3
By combining equations 1, 2, and 3, the insulation resistance Rf can be found.
The calculation of the capacitance depends on the size of the resistance and the curve of the waveform. Assume that the voltage has two points M1 and M2 on the waveform with respect to time t, and the corresponding coordinates are (V1, t1), (V2, t2) according to the capacitor charging formula:
.png)
Corresponding to M1 and M2:
.png)
.png)
After processing:
.png)
In the actual calculation process, point calculations can be performed multiple times, averaged, and measurement accuracy improved.
At -20V, the insulation resistance Rf and leakage capacitance are calculated in a similar manner.
2.3.2 The system has a DC component
When the system has a DC component, it still needs a fourth-order filter circuit to filter out the system AC signal (the DC signal still exists), and then passes through a signal holding circuit as shown in FIG. 4 :
.png)
Figure 4. Signal holding circuit
The input signal is divided into positive and negative half-cycle signals, but both contain the DC component in the system. By opening and closing the switch, the positive and negative half cycle signals can be subtracted. Since the DC voltage amplitude of the system changes very little, the subtraction occurs. After the signal no longer contains a DC component, only ±20 V is applied to the result of the sampling resistor at this point in the sampled signal, and the final signal is amplified to the microcontroller sampling module. Into the ADC sampling waveform can refer to the PSPICE simulation results shown in Figure 5:
.png)
Figure 5. Two Independent Signal Waveforms
The system can independently monitor the insulation condition at +20V or -20V. In this way, the measurement period is at least one-half that of the fixed-cycle product measurement period. The calculation of the resistance in the DC system is the same as that described in the AC system. The magnitude of the resistance depends on the voltage value after the waveform is stable. The calculation of the capacitance is still dependent on the resistance. The calculation method is similar to that of the ADC sampling signal that can be inverted at +20V and -20V. When the partial pressures of R2 and R4 in FIG. 1 are obtained, the insulation resistance value and the leakage capacitance value can be obtained.
2.4 Other Circuits of the Instrument
In addition to the above circuits, there are disconnection detection circuits (PE/KE disconnection, L1/L2 disconnection detection function), 485 communication circuits, and other communication circuits.
3. Software design
3.1 Software Process
The insulation monitoring device uses a structured programming concept and is written in C language. The main function determines the execution of the corresponding module by querying the state of the flag bit, and the flag of each module changes within the timer. This method improves the real-time nature of the software, and later software maintenance is relatively easy.
3.2 Adaptive Frequency
Currently, most of the products in the market adopt the method of injecting fixed-cycle signals into the system. This method must consider the maximum resistance and capacitance of the system. The measurement period must meet the requirements of the maximum resistance and the maximum capacitance. Therefore, the period is also the longest, and unable to be changed. Adaptive frequency is a new type of cycle adjustment method, which adjusts the cycle size by monitoring the signal waveform of the system. Take two points of the voltage signal on the signal waveform, when the signal voltage change is very small, regard as the stability, then reverse the pulse signal, and preserve the cycle running time as the cycle of next pulse. Since the waveforms are monitored and calculated both in the positive and negative half cycles, the adjustment of the signal waveform will be timely, and the calculation result of the resistance will be updated relatively quickly. In addition, once the resistance and capacitance measurement results are stable, the system will calculate the theoretical cycle and compare it with the actual measurement cycle, and then assign the theoretical measurement cycle to the next pulse cycle. This method ensures that the measurement cycle can be minimized on the premise of accurate measurement results.
3.3 Response time
Table 4.6 of IEC 61557-8 Part 8 “Insulation Monitoring Devices in IT Systems†states that in a pure AC system, when the leakage capacitance 1uF and insulation resistance are 0.5 times the alarm value, the response time should be less than 10 seconds. Under the premise of measuring accuracy, the response speed of the device can be less than 6s. The following is a brief analysis of the influence of the resistance mutation on the waveform. Xiang is shown in Figure 6:
.png)
Figure 6. Fault simulation waveform
Solid line: Waveform 1 Dotted line: Waveform 2
Before t1, the system cycle has stabilized. Assuming that the resistance at t1 (voltage V1) suddenly decreases below the alarm value, the waveform changes. When the sampling time t2 is reached, the voltage V2 is measured at this time. The CPU determines that the difference between the two is greater than The fixed value is doubled in the second half of the week and becomes 2T (previously T). Since the capacitance is very small, the system will stabilize early before the 2T time runs out. Although the system will end prematurely before the cycle is completed, the response time will increase. If a complete positive and negative cycle signal is used as the basis for the alarm response, the response time is greatly increased. To solve this problem, the system calculates the resistance value (independent signal) after the end of the half-week. If the resistance value is less than the set alarm value, an alarm signal is issued. The response value is t2~t1 in Figure 6. After the actual test, The response time is basically kept within 5s and the longest is not more than 6 seconds.
3.4 Other Software Description
The software calibration adopts the linear segmented calibration method with a total of 8 calibration points to ensure the accuracy of the instrument; in order to filter the noise interference in the signal, the digital filtering uses the bubble method (sequencing the data) and the median value filtering method in turn. The average value filtering method processes the data to ensure the reliability and stability of the signal.
4. Test results
The product has passed the type test of the Xuchang Cape Test Center, and its functions and performance meet the requirements of international standards. The test verified that under the condition of resistance 1K-5M and capacitance 0-150uF, the ratio of the displayed value to the actual value is kept within 10%, the measurement accuracy is up to standard, and it can meet the insulation monitoring of ungrounded systems in various environments. demand.
5 Conclusion
This article describes a new type of insulation monitoring device, compared with the market insulation monitoring instrument, its advantages are the ability to monitor the DC ungrounded system, allowing the system to leak large capacitance, short measurement period, short response time and so on. After tests, the insulation monitoring device described in this article can work reliably in AC and DC ungrounded systems, and can provide a reliable monitoring for ungrounded systems.
Article Source: "Intelligent Building Electrical Technology" 2016 No. 3.
references
[1] Wang Houyu. On the application of it system. China Hang Kong Industrial Planning and Design Institute (Beijing).
[2] JGJ 16-2008 Civil Building Electrical Design Specification [S].
[3] Liu Guoping. Marine Electrical and Communication. First Edition. Beijing: Ocean Publishing House, 2004.
[4] Wang Yi, Wang Jinquan, Yang Tao, et al. Discussion on several key issues of low-voltage IT system[J]. Building Electrical, 2011(11):47-50.
[5] MA Tao, WANG Jin-quan, JIN Wei-yi, et al. Research on three-phase four-wire IT system insulation monitoring technology plan [J]. Marine Electric Technology. 2008 (5): 277-280.
[6] IEC 61557-8Electrical safety in low voltage distribution systems up to 1000Va.c.and1500V dc - Equipment for testing, measuring or monitoring of protective measures – Part8:Insulation monitoring devices for IT systems
Yokohama Fender,Net Type Pneumatic Fenders,Yokohama Rubber,Yokohama Floating Fenders
Shandong Nanhai Airbag Engineering Co., Ltd. , https://www.nanhaimarine.com