1. Introduction
In the modern power electronics industry, rectifier and inverter power systems have made great progress. In this power system, the role of the DC link capacitor is to prevent the DC bus and the parasitic inductance of the capacitor from generating induced electromotive force due to sudden changes in load, which leads to a large sudden change in the DC bus voltage. This article mainly introduces the reduction scheme of the stray inductance of the metallized film DC link capacitor itself and the discussion of the current distribution scheme inside the product. The purpose is to ultimately make the DC-Link capacitor product have lower heat loss and more uniform temperature distribution.
2. Brief description of the principle
The typical application circuit of a DC support capacitor (i.e., a DC-Link capacitor) in a power system is shown in FIG1 , where Ls is the parasitic inductance of the system connection.
In this application, the IGBT inverter can be regarded as the load of the rectifier circuit. This switching device load has a sudden current. According to formula (1),
V=L×di/dt-----------------------(1)
Under the action of inductance, the sudden current will generate an induced electromotive force. When the circuit and the stray inductance of the capacitor product itself are affected, this induced electromotive force may reach tens or even thousands of volts. Such a high sudden voltage will cause serious interference to the system and even damage the system. The role of the DC support capacitor is to use the characteristics that the capacitor voltage cannot change suddenly and the capacitor reactance decreases with the increase of frequency to provide a low-impedance channel for the system in a very wide frequency band, thereby reducing the AC impedance of the DC bus. According to formula (2)
Where ω=2πf, ESL is the stray inductance, and ESR is the equivalent series resistance
From the above formula, we can know that the impedance of a capacitor is not only related to the capacitance, but also to the equivalent series resistance, stray inductance, and system frequency. The relationship curve between its impedance |Z| and frequency is shown in Figure 2:
As can be seen from the figure, as the frequency increases, the impedance gradually decreases. When f=f0, it has the lowest impedance, which is the equivalent series resistance ESR. When f>f0, the capacitor no longer has the property of capacitive reactance, but presents inductive reactance, and the capacitor has lost its function. Therefore, the operating frequency of the capacitor should be much smaller than the resonant frequency. The resonant frequency f0 is determined by the following formula:
For a capacitor of a selected capacity, if we want the capacitor to exhibit capacitance over a wider frequency band, that is, if we want it to have a relatively high resonant frequency, then we must have a smaller stray inductance.
On the other hand, if the equivalent series resistance (ESR) of the capacitor is relatively large, the capacitive reactance of the capacitor will be lower than the ESR at relatively low frequencies. At this time, the AC impedance of the capacitor mainly depends on the ESR, and the requirement of low AC impedance cannot be achieved well. This article does not discuss the impact of this factor in detail.
Next, we will look at the main causes of stray inductance ESL in DC-Link capacitors.
The internal equivalent circuit diagram of the DC-Link capacitor is shown in Figure 3:
Since our DC-Link capacitor products use metallized polypropylene film capacitors, due to their small specific capacitance, in order to obtain a relatively large capacity, their volume is relatively large. The product is composed of multiple core units connected in series and parallel. If the DC-Link capacitor has a relatively large stray inductance and the internal connection is unreasonable, it will cause uneven current distribution between the core units inside the product, and the external manifestation is that the local temperature rise of the product is too high. The inductive reactance increases with the increase of frequency, so this phenomenon will be particularly obvious in high-frequency conditions, and in severe cases it will cause thermal breakdown of the capacitor and cause accidents.
The main sources of stray inductance ESL inside the film DC-Link capacitor are as follows:
(1) The core formed by winding the metallized film itself;
(2) Caused by core unit string, parallel lead or copper busbar;
(3) Metal casing inductance: This is caused by the connection between one electrode of the product and the metal casing. This factor does not exist in other cases.
We will discuss the solutions to the above three reasons in the following case analysis.
3. Case analysis
The following is a specific analysis of the DC-Link product provided by our company for a certain company:
The product model is MKP-LG6000μF/1200V.DC with a nominal effective current of 300A. The shell is made of non-magnetic steel. First, we give a set of our temperature rise test data, see Table 1
Note: The data collection points in Table 1 are all marked in Figure 4; the test current is 310A; the test frequency is 13.75kHz.
From the data, we analyzed that the temperature difference between points 5 and 7, and between points 6 and 8 is large, reaching 8~10℃, and the temperature distribution of each point on the upper surface of the product (this is the epoxy surface) is also uneven, with a large temperature difference, which affects product reliability.
Figure 4
The product used in the above test is a structure designed by our company in the early stage, which did not consider the influence of stray inductance and the optimization of the internal current distribution of the product. In addition, since the capacitor itself has a concentration effect during use, the current will be concentrated on the upper part of the capacitor. In the above scheme, the arrangement structure of the core inside the product can be simply represented as shown in Figure 5.
Figure 5
As can be seen from the figure above, the rectangle is the copper plate for wiring. Since the copper plate has a certain inductance, the impedance is relatively large for high-frequency current. According to the formula I=U/Z=U/(XL+XR+XC)
XL=2πfL--------------------------(4)
Xc=1/(2πfc)------------------(5)
(Assume 2πf=ω) It can be seen that when the frequency is fixed, the larger the inductance, the larger the inductive reactance. When the frequency is low, for example, at an industrial frequency of 50Hz, the inductive reactance Esl generated by the stray inductance in the circuit is low, much smaller than Esc, so Esl can be ignored, and Esr and Esc are the main influences. But when the current frequency is as high as 600kHz, the capacitive reactance is low, about 0.005Ω/mm, and the inductive reactance is very large, about 0.3Ω/mm, much larger than Esc, and it is the main influence in the equivalent circuit. If the average stray inductance of each 1mm copper plate is about 1nH, and the distance between each terminal is 60mm, then the total stray inductance in the circuit is 60nH, and the stray capacitance is 40nH, then the inductive reactance of the first capacitor is XL=40ω, the inductive reactance of the second capacitor is XL=2Xl+Xc=160ω, and the inductive reactance of the third capacitor is: XL=4Xl+Xc=260ω. Because I=U/XL=200A, the ratio of the current passing through these three is I1:I2=4:1, I1:I3=6.5:1, from which it can be concluded that 1.4I1=200A, and the current flowing through C1 is the largest, about 143A, the current flowing through C2 is about 36A, and the current flowing through C3 is about 21A. Therefore, the current of C1 is seriously heated, while the heating of C2 is normal, and the heating of C3 is less, which can easily burn C1, so this connection method cannot be adopted. At the same time, in the case of relatively low frequency, such as the power frequency of 50Hz, the shell material has little effect on the product. However, when the frequency reaches 10kHz or above, during the use of the product, if the shell material is magnetic, it will also heat up due to induction heating, which will have an adverse effect on the overall heating of the product. We started to improve the solution from four aspects.
First, we calculate the inductance based on a single rectangular cross-section wire:
Where A is the thickness of the rectangular wire, H is the width of the rectangular wire, and l is the length of the rectangular wire.
In view of the influence of internal leads on stray inductance, we use formula (6) to estimate the wire inductance. By adjusting the wire cross-section and length, we can achieve the goal of minimizing the stray inductance of the product itself while satisfying comprehensive factors such as good product overcurrent and cost, so as to achieve the purpose of reducing the heat loss of the inductor part.
Secondly, we adjusted the core connection structure of the product to achieve a more even current distribution between each core unit of the product. Figure 6 shows the connection structure after our adjustment:
Figure 6
As shown in Figure 6, the same device, but with different connections, has a different equivalent circuit diagram. The conditions are the same as the original solution, but the equivalent inductive reactance is different. From the equivalent circuit diagram, we can see that XL1=XL2=XL3=2πfL, and Esr and Esc are equal, so the current flowing through each capacitor is I=U/R=40A. This can evenly distribute the current to each capacitor. In this way, the problem of uneven current distribution causing serious heating of some capacitors is solved.
Thirdly, we have improved the connection method of the end face of the core inside the product: we changed the previous method of direct connection with a whole copper busbar. By adjusting the size of the copper busbar and cutting it appropriately, we can make the stray inductance distribution of the copper busbar itself more reasonable, and at the same time weaken the impact of eddy current on the end face connection and reduce heat.
Finally, since non-magnetic stainless steel still has a certain degree of magnetism, it is easy to produce additional heating under medium and high frequency conditions. Therefore, we changed the shell material to aluminum, which greatly eliminated the impact of the heating of the shell itself on the overall temperature rise of the product.
After the improvement, the temperature rise effect of the product is shown in Table 2 (the test points are the same as Table 1):
Table 2 MKP-LG6000μF product - after improvement, overcurrent test data excerpt
Note: Temperature point 7 was not measured, the test current and frequency are the same as in Table 1
From the data analysis of Table 2, it can be seen from points 5, 6, and 8 that the temperature distribution of each point on the product shell surface is relatively uniform, and the temperature difference does not exceed 3°C. And from the analysis of the data of the corresponding points in Tables 1 and 2, it can be seen that the temperature rise of the improved product is lower than that before the improvement, especially the upper epoxy surface, which is 18.1°C lower at the highest point. The improvement effect is very obvious.
4. Conclusion
This solution not only solves the problem of uneven current distribution on the capacitor core group, but also reduces the power loss of the equipment, thereby increasing the service life of the machine.
With the needs of industrial development and the theme of environmental protection and energy saving, inverter power supplies are increasingly used, so the technical requirements for them are more stringent. The quality requirements for its core part, the DC-Link capacitor, are also increasing. We fully consider the arrangement of the capacitor's internal core, the distributed inductance of the lead wires, and the influence of the heating of the magnetic material, and select a more optimized wiring method and design scheme, so that the DC-Link capacitor can meet the needs of the continuous development of technology, which in turn promotes the advancement of technology.
Previous article:Design of 5kW off-grid photovoltaic inverter using DSP power inductor
Next article:Master shares with you how to handle leakage inductance spikes in push-pull inverters
Recommended ReadingLatest update time:2024-11-22 11:31
- Bourns Launches Two Thick Film Resistor Series with High Power Dissipation Capabilities in Compact TO-220 and DPAK Package Designs
- Bourns Launches New High-Pulse Brake Resistor Series with Superior Energy Dissipation Capabilities
- Nexperia launches new 120 V/4 A half-bridge gate driver to further improve robustness and efficiency in industrial and automotive applications
- Vishay's New 150 V MOSFETs Offer Industry-Leading Power Loss Performance
- The first generation of SGT MOSFET series from Qiangmao: innovative trench technology, automotive-grade 60 VN channel, breakthrough in high-efficiency performance of automotive electronics
- DC/DC Power Supplies for Automotive Applications
- MathWorks and NXP Collaborate to Launch Model-Based Design Toolbox for Battery Management Systems
- STMicroelectronics' advanced galvanically isolated gate driver STGAP3S provides flexible protection for IGBTs and SiC MOSFETs
- New diaphragm-free solid-state lithium battery technology is launched: the distance between the positive and negative electrodes is less than 0.000001 meters
- Intel promotes AI with multi-dimensional efforts in technology, application, and ecology
- ChinaJoy Qualcomm Snapdragon Theme Pavilion takes you to experience the new changes in digital entertainment in the 5G era
- Infineon's latest generation IGBT technology platform enables precise control of speed and position
- Two test methods for LED lighting life
- Don't Let Lightning Induced Surges Scare You
- Application of brushless motor controller ML4425/4426
- Easy identification of LED power supply quality
- World's first integrated photovoltaic solar system completed in Israel
- Sliding window mean filter for avr microcontroller AD conversion
- What does call mean in the detailed explanation of ABB robot programming instructions?
- Chuangshi Technology's first appearance at electronica 2024: accelerating the overseas expansion of domestic distributors
- "Cross-chip" quantum entanglement helps build more powerful quantum computing capabilities
- Ultrasound patch can continuously and noninvasively monitor blood pressure
- Ultrasound patch can continuously and noninvasively monitor blood pressure
- Europe's three largest chip giants re-examine their supply chains
- Europe's three largest chip giants re-examine their supply chains
- Breaking through the intelligent competition, Changan Automobile opens the "God's perspective"
- The world's first fully digital chassis, looking forward to the debut of the U7 PHEV and EV versions
- Design of automotive LIN communication simulator based on Renesas MCU
- When will solid-state batteries become popular?
- How to solve the problem of unstable GND?
- Some arguments for your own capacitor filtering
- [A goddess who gives a good gift will not run away] Light up my heart
- Two current-type control methods of inverter
- 24 GHz to 44 GHz Wideband Integrated Upconverter and Downconverter Boosts Microwave Radio Performance While Reducing Size
- The relationship between baud rate and time, and the application of transistor/MOS in communication interface level conversion
- Medium and high voltage GaN devices: vertical or horizontal?
- [Help needed] This is a slightly complicated schematic diagram, about the TI processor and USB power supply part. There are some things I don't understand and I would like to ask everyone for help.
- What is GAN semiconductor technology and how does it work?
- Research on a New Type of ZVS Shutdown Circuit