Quickly detect the principle of self-damage of IGBT components

0805RGB flashing

Kaixin micro test

Test - lowercase jpg

Test probe PH-5H

Overview:
1 Introduction The main drive of a five-stand mill in a factory adopts Hitachi Hivectol-con/vsi-mh series AC-DC-AC 3-level pwm large-capacity variable frequency transmission. The maximum capacity is 8800kva, which is composed of 4 sets of igbt units in parallel. The power components in each unit are Mitsubishi 3.3kv/1.2ka large capacity igbt. Since the start of the operation, there have been many major failures of igbt and fast-melting large-area breakdown. For example, in a failure in December 2005, all 12 igbt units and 12 fast-melting units in a rectifier were burnt down. The fault processing time exceeded 33 hours, and the direct economic loss reached 1.17 million yuan, which seriously affected the normal production of the factory. After inspection and analysis, it is found that the cause of the fault is caused by the deterioration of one igbt component of the rectifier r phase, and the fault is greatly enlarged due to the long time of the original system igbt pulse blocking process.
After improvement, a new self-loss fast detecting device for igbt components has been put into operation, so that the process time of detection and judgment, pulse blocking and the like is only tens of microseconds, which greatly reduces the possibility of fault enlargement. Repeated testing by the factory also proved the effectiveness of the method.

2 igbt component protection several methods and advantages and disadvantages Due to the limitation of current and voltage capability of igbt components, in the large-capacity transmission system actually used, multiple parallel forms are often used. If a short-circuit fault such as igbt component breakdown occurs, if the igbt pulse is not blocked quickly, it may cause a large amount of igbt damage in the parallel loop and expand the fault range.
In the transmission system, the main circuit of the incoming and outgoing lines is generally provided with a Hall ct for detecting the main circuit current, and the overcurrent condition is processed and judged by the hardware and software of the main control board. If a heavy fault trip signal is issued, the pulse is quickly blocked to protect the igbt component. The advantage is that the overcurrent protection effect caused by the load is obvious, but the disadvantage is that the process time of the overcurrent detection to the pulse blockade is too long, it takes several milliseconds, and the short circuit of the DC loop can not be protected (there is no DC loop in the actual system). Current detection ct). Therefore, it is obvious that this protection method cannot meet the actual requirements of igbt protection.
To this end, other rapid detection methods must also be taken. The currently used methods are as follows:

(1) igbt vce voltage monitoring method This is a more common method. The phenomenon that vge or vce also rises when the collector current ic rises, when the vge or vce exceeds the set allowable value, the output signal blocks the igbt pulse. Because vge changes little when it fails, it is difficult to master and is generally used less. The vce changes greatly, so in practice, the vce monitoring method is often used to protect igbt. The advantage of this method is that the detection is sensitive and the action is rapid, and the large-area damage of the parallel circuit igbt is effectively avoided. However, the disadvantages of this method are also obvious: wiring is required, and the voltage signal between the collector and the emitter of each igbt is introduced into the pulse driving board. In addition, since the voltage between the collector and the emitter is relatively high when the igbt is turned off, it is necessary to increase the corresponding insulation and potential isolation measures of the pulse amplifying board. FIG. 1 is an example of protecting igbt by detecting a voltage vce between a collector and an emitter.