COMMON PROBLEMS AND SOLUTIONS IN IGBT APPLICATIONS

The igbt, which combines the advantages of gtr and mosfet, is a new type of insulated gate bipolar transistor that was introduced in the 1980s. It has convenient control, high operating frequency, fast switching speed and large safe working area. With the continuous improvement of voltage and current levels, igbt has become an ideal power switching device for high-power switching power supplies, variable frequency speed regulation and active filters, and has been widely used in power electronic devices.
 
With the development of high-frequency and high-power of modern power electronics technology, the potential problems of switching devices are becoming more and more prominent in the application. The voltage and current overshoot caused by the switching process affect the working efficiency and reliable operation of the inverter. Sex. In order to solve the above problems, measures such as overcurrent protection, heat dissipation, and reduction of line inductance have been actively adopted. The snubber circuit and soft switching technology have also been extensively studied and rapid progress has been made. This article reviews this aspect.
 
2 igbt application areas
 
2.1 Application in frequency converter
 
The block diagram of the spwm variable frequency speed control system is shown in Figure 1. The main circuit is a standard topology circuit of a voltage source type spwm inverter with igbt as a switching element. The capacitor is charged by a rectifier circuit, and the spwm signal generated by the control circuit controls the output waveform of the inverter through the driving circuit; The three-phase AC voltage of the corresponding frequency, amplitude and phase sequence is output to the asynchronous motor to operate at a certain speed and direction of rotation.
 
 
2.2 Application in switching power supply
 
Figure 2 shows a block diagram of a typical ups system. Its basic structure is a set of rectifiers and chargers that convert alternating current into direct current and inverters that convert direct current into alternating current. The battery stores energy when the AC power is normally supplied and maintains a normal charging voltage, and is in a "floating state". Once the power supply is out of the normal range or interrupted, the battery immediately supplies power to the inverter to ensure that the ups power supply outputs an AC voltage.
 
 
The main control object in the ups inverter power supply is the inverter, and the most widely used control method is the sinusoidal pulse width modulation (spwm) method.
 
2.3 Application in active filters
 
 
2.2 Application in switching power supply
 
Figure 2 shows a block diagram of a typical ups system. Its basic structure is a set of rectifiers and chargers that convert alternating current into direct current and inverters that convert direct current into alternating current. The battery stores energy when the AC power is normally supplied and maintains a normal charging voltage, and is in a "floating state". Once the power supply is out of the normal range or interrupted, the battery immediately supplies power to the inverter to ensure that the ups power supply outputs an AC voltage.
 
The main control object in the ups inverter power supply is the inverter, and the most widely used control method is the sinusoidal pulse width modulation (spwm) method.
 
2.3 Application in active filters
 
The schematic diagram of the parallel active filter system is shown in Figure 3. The main circuit is an inverter with igbt as the switching element, which injects the reverse harmonic value into the system, and theoretically can completely filter out the harmonics existing in the system. Different from the frequency converter, the modulated wave of the active filter pwm control signal is a composite waveform of each harmonic that needs to be compensated. In order to accurately reflect the harmonic components of the modulated wave, the carrier must be greatly improved. Frequency of. This also places higher demands on the switching frequency of the switching device.
 
Analysis of common problems in 3 igbt applications
 
Obviously, igbt is applied to various systems as a switching element of an inverter, and the commonly used control method is the pwm method. Both theoretically and in fact, it has been proved that if the switching frequency of the pwm inverter is increased above 20khz, the noise of the inverter will be smaller, the volume will be smaller, the weight will be lighter, and the output voltage waveform will be more sinusoidal. It can be seen that high frequency is the development direction of inverter technology. However, in a typical pwm inverter, the switching device is turned on at a high voltage, turned off at a large current, and is in a forced switching process. When operating at a high switching frequency, it is limited by a series of factors:
 
(1) Generate a holding effect or dynamic holding effect
 
Igbt has a four-layer structure, so that there is a parasitic thyristor in the body, and the equivalent circuit is shown in Figure 4. There is a body short circuit rs between the base and the emitter of the npn tube. The lateral hole flow of the p-type body region will produce a certain voltage drop, which is equivalent to a positive bias voltage for j3. Within the specified range, this positive bias voltage is not large and the npn tube will not conduct. When ic is greater than a certain degree, the positive bias voltage is sufficient to turn on the npn transistor, so that the npn and pnp transistors are in a saturated state, so that the parasitic thyristor is turned on, and the gate loses control, that is, the holding effect, which increases ic. Causes excessive power consumption and even damage to the device. An increase in temperature will cause a serious drop in the icm that the igbt will hold.
 
In the dynamic process of igbt turn-off, if the dvce/dt is higher, the displacement current cj2dvce/dt caused in the j2 junction is larger. When the current flows through the body short-circuit resistance rs, it can generate enough for the npn transistor to be turned on. The forward bias voltage satisfies the conditions of the parasitic thyristor turn-on and forms a dynamic holding effect. An increase in temperature will increase the risk of igbt's dynamic holding effect.
 
(2) Excessive di/dt causes voltage overshoot at the time of switching through the line inductance between igbt and the snubber circuit
 
The circuit is analyzed with the line inductance lб≠0. As shown in Figure 5, during the turn-off process, the inductive load current iб remains unchanged, ie iб=it+id remains unchanged, and it increases from zero to iб. Since diode d is turned on, vee=0, since it linearly decreases with time, the induced voltage vl=vbc=lбdit/dt at both ends of the inductor lб should be a negative value, and vb is a positive value, that is, the potential at point c is higher than the potential at point b. .
 
Since it=i0(1-t/tfi)
 
Therefore vl=vbc=lбdit/dt=-lбi0/tfi“0
 
Vcb= -vbc= lбi0/tfi
 
During the tfi of its falling, the voltage across the switch
 
Vt=vcem=vd-vl=vd+lбi0/tfi
 
Therefore, at the beginning of the shutdown process, vt immediately rises from zero to vcem, and during the period from i0 to zero, vt=vcem does not change. Until it=0, id=i0, vt drops to the power supply voltage vd, as shown in Figure 5(b). The value of vcem over vd depends on lб, tfi and load current i0. Obviously, too fast current drop rate di/dt (ie tfi is small), excessive stray inductance lб or excessive load current will cause serious components when shutting down. Overvoltage, and with a lot of power consumption.
 
It can be seen that although the fast turn-on and turn-off of igbt is beneficial to shorten the switching time and reduce the switching loss, too fast turn-on and turn-off are harmful under large inductive load. When turned on, there is a freewheeling diode reversal. The recovery current and the discharge current of the snubber capacitor, the faster the turn-on, the larger the peak current that igbt is subjected to, and even rises sharply, causing damage to the igbt or freewheeling diode. When turned off, the large inductive load turns on and off with the overspeed of igbt, which will generate a high-frequency, high-amplitude, narrow-width spike voltage ldi/dt in the circuit. The conventional over-voltage snubber circuit is exposed to the diode. The limit is difficult to absorb the spike voltage, so vce suddenly rises to produce overshoot, and igbt will withstand higher dvce/dt impact, which may cause damage to other components in the circuit or itself due to overvoltage breakdown.
 
(3) The state of operation of the switching device at the moment of turn-on and turn-off is beyond the reverse safe working area (rbsoa);
 
The reverse safe working area (rbsoa) is surrounded by three limit boundary lines of maximum collector current icm, maximum collector pole voltage vce and voltage rise rate dvce/dt, and is added with dvce/dt when igbt is turned off. Change, the higher the dvce/dt, the narrower the rbsoa, so the high dvce/dt generated at the turn-on and turn-off moments will make the state of the switching device more easily beyond the rbsoa, affecting the switch reliability.
 
(4) The surge voltage when dv/dt and igbt are turned off during diode reverse recovery will cause overcurrent during switching.
 
It is well known that igbt has a Maitreya capacitor ccg and an input capacitor cge. The voltage overshoot across igbt will pass through the ccg coupling gate, causing the gate voltage to rise instantaneously because of the gate negative bias and the input capacitor cge. The height reached by the voltage is much lower than the overshoot of the collector, but it may exceed the threshold and cause the tube that should be turned off to conduct, so the upper and lower arms are straight through and overcurrent.
 
If the resulting gate voltage is sufficient to saturate the tube, it is not through but shorted. This overcurrent or short circuit also appears repeatedly in the oscillation decay process after the collector voltage overshoot. Experiments prove that this phenomenon does exist.
 
4 commonly used solutions
 
For the above problems, the practical measures generally adopted are: selecting an effective overcurrent protection circuit, using a non-inductive line, actively dissipating heat, using an absorption circuit and a soft switching technique.
 
4.1 Select an effective overcurrent protection drive circuit
 
In igbt applications, the key is overcurrent protection. The overcurrent time that igbt can withstand is only a few microseconds, which is much smaller than that of devices such as scr and gtr (tens of microseconds), so the requirement for overcurrent protection is even higher. Igbt's overcurrent protection can be divided into two types, one is low multiple (1.2~1.5 times) overload current protection;
 
The other is high-multiplier (8~10 times) short-circuit current protection. For overload protection, protection can be achieved by instantaneously blocking the gate pulse. For the short-circuit current protection, adding the instantaneous blocking gate pulse will be too large due to the di/dt of the short-circuit current drop, and it is easy to induce a high collector voltage to pass through the igbt on the loop stray inductance, so that the protection fails.
 
Therefore, for igbt, reliable short-circuit current protection should have the following characteristics:
 
(1) First, the gate voltage should be softly lowered to limit the peak value of the short-circuit current, prolong the allowable short-circuit time, and win the time for the protection action;
 
(2) Protection to cut off the short-circuit current should be implemented with soft shutdown
 
The igbt drivers exb841, m57962 and hl402b meet the above requirements. However, these drivers cannot completely block the pulse. If no measures are taken, the soft-off protection of each cycle will be caused once the fault does not disappear. The heat accumulation thus generated will still cause damage to the igbt. To this end, the fault detection output of the driver can be used to completely block the gate pulse through the optocoupler, or reduce the operating frequency to below 1hz, and automatically return to the normal operating frequency when the fault disappears.
 
As shown in Figure 6, the igbt driver module m57962l has its own protection function. When the detection circuit detects that the detection input pin 1 is 15v high level, it determines that it is a current fault, immediately starts the gate shutdown circuit, and sets the output terminal 5 pin. Low level, make igbt cut off, and output error signal to make the fault output terminal 8 low level, to drive the external protection circuit to work, delay 8~10μs to block the drive signal, which can achieve overcurrent protection well. After a delay of 1~2ms, if the input is detected as high level, m57962l is reset to the initial state.

Pre:IGBT KNOWLEDGE 3

Next:QM75DY-24 DATASHEET

Shunlongwei Co. Ltd.

Contact: Alice Peng

Tel:+86-755-82732562

E-mail: sales@shunlongwei.com

ADD:Rom512,Bldg#505, Shangbu Industrial Area, Huaqiang North Rd., SZ,518000,China.