CH665515A5 - Circuit for controlling a power switching transistors with galvanic isolation. - Google Patents

Description

665 515

2nd

PATENT CLAIM Circuit arrangement for controlling a power switching transistor with electrical isolation, characterized in that a first input terminal (VH1) is connected to the base of a first transistor (Tl), the emitter of which is connected to ground and the collector of which is connected to the common point of a first Zener diode (ZD1 ) and a first connection of the primary winding (Nl) of a first transformer (TRI) is connected, a second connection of this winding (Nl) with a current generator (Igl) and a connection of a ninth diode (ND9) and its second connection with the first Zener diode (ZD1) are connected so that a first and second resistor (RI, R2) are connected in series between the connections of the secondary winding (N2) of the first transformer (TRI), with the common point of these resistors (RI, R2 ) the base and at the common point of connection of the winding (N2) and the second resistor (R2) the emitter of a fourth transistor (T4) are connected that the collectors of the fourth and a fifth transistor (T4, T5) together with a connection of the secondary winding (N8) of a third transformer (TR3), which has an output terminal (IZ2), the emitter of the power switching transistor (MST) , to which the one connection of a twelfth diode (Dmst) is connected, is connected that to the common point of the first resistor (Rl) and the connection of the secondary winding (N2) the one connection of the secondary winding (N3) of the first transformer ( TRI) and one connection of a first diode (Dl) are connected, the other connection of which is connected to a connection of a third diode (D3), the base of the power switching transistor (MST) and a connection of the series connection of a fifth, sixth, seventh and eighth diode ( D5, D6, D7, D8) is connected,

that the other connection of the series circuit is connected to the other connection of the secondary winding (N8) of the third transformer (TR3), that between the connections of the secondary winding (N6) of a second transformer (TR2) a third and fourth resistor (R3, R4) in Are connected in series, at the common point of these resistors (R3, R4) the base and at the common point of connection of the winding (N6) of the second transformer (TR2) and the fourth resistor (R4) the emitter of the fifth transistor (T5 ) that the one connection of the winding (N5) and the one connection of the third diode (D3) are connected to the common point of the third resistor (R3) and the connection of the secondary winding (N6), that the other connection of the winding (N3) of the first transformer (TRI) via a second diode (D2) and the other connection of the winding (N5) of the second transformer (TR2) via a fourth diode (D4) to the collector de s power switching transistor (MST), to which the other terminal of the twelfth diode (Dmst) is connected, are connected together with the other output terminal (IZ1), that a second input terminal (VH2) is connected to the base of a second transistor (T2) , whose emitter is connected to ground and whose collector is connected to the common point of a second Zener diode (ZD2) and the first connection of the primary winding (N4) of the second transformer (TR2), the second connection of this winding (N4) being connected to a second Current generator (Ig2) and a terminal of a tenth diode (DIO) and the second terminal of which are connected to the second Zener diode (ZD2), that a third input terminal (VH3) is connected to the base of a third transistor (T3) whose emitter are connected to ground and its collector at the common point of a third Zener diode (ZD3) and the first connection of the primary winding (N7) of the third transformer (TR3), where the second connection of this winding (N7) is connected to a terminal (+ V) and a connection of an eleventh diode (Dil) and the second connection of which is connected to the third Zener diode (ZD3).

DESCRIPTION

The present invention relates to a circuit arrangement for controlling a power switching transistor with electrical isolation.

The technical problem to be solved is the control of a power switching transistor with a feedback diode, which can be integrated discretely or already on the same monolithic circuit as the power transistor, with galvanic isolation also having to be achieved.

The operating states of the transistor are as follows:

- uninterrupted control status;

- uninterrupted locked state;

- all intermediate states with the frequency of 100 kHz;

- continuous operation as a diode;

- The use of a low voltage bipolar transistor for the control of a power switching transistor.

Known solutions for controlling power switching transistors can be divided into several groups:

1. immediate control of the base current;

2. direct control of the base current of larger transistors located in the Darlington circuit;

3. Control of NPN switching transistors with a PNP transistor;

4. Control of a PNP power switching transistor;

5. Control of a power switching transistor via transformers.

1. Direct control means the type in which the energy for supplying the transistor base is supplied from a supply unit which must be galvanically separated from the control electronics. The control signal is transmitted via an optical connection. These circuits are already specified by the manufacturers of power switching transistors and are used by them as test circuits. In addition, other circuits are specified that are simpler and cheaper equivalents of the basic control. The control of the transistor could also be counted with a single galvanically isolated supply stage. The examples are given in the Motorola application book: Switchmode Application Manual, printed in Switzerland in 1980, in the last part of the book, the pages of which are not numbered.

The Westinghouse company also provides basic control solutions in its publication: Merle Morozovich: Where, Why and How to Use the D60T Transistor, Westinghouse Semiconductor Division, Youngwood, PA 15697, 1979, p. 30, Fig. B and C.

In this way, the transistor is driven as much as is necessary, and it can be brought into the saturated state and the losses during conduction can be reduced. This increases switching losses and extends shutdown. This can also be reduced in such a way that a circuit for preventing oversaturation is installed. Any form of base stream can be formed without difficulty. The main disadvantage is that two galvanically isolated supply units are required for electrical isolation, one each for a positive and negative base voltage. The supply unit for the positive base voltage must have the ability to supply a sufficiently high current in the order of magnitude of several amperes, which is quite unfavorable at low voltage because of its low efficiency. The transistors for opening the power switching transistor must be able to supply the entire base current. The capacity of the supply part compared to the network and the mass of the supply electronics exerts a negative influence on the behavior of the circuit, especially at high frequencies. The problem is partially solved with an optocoupler, but with which the information transmission is disturbed at high frequencies.

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

3rd

665 515

2. In order to reduce the necessary base current, two or more transistors were connected in a Darlington circuit. This type is most widely used in practice and is therefore the easiest to find in industrial products. The galvanic isolation is also realized with optically connected galvanically isolated supply units.

The main advantage in this regard is the lower control current with the same collector current. This significantly reduces the power required by the galvanically isolated supply section for the positive base current. The transistors for opening the Darlington circuit also need to have lower power dissipation compared to the variant under item 1. The disadvantages are as follows:

- higher voltage when the transistor is in the on state, which also causes higher losses;

- Because of the delay of the transistor for opening the power transistor, the switch-on and switch-off times are longer;

- Higher price because of the expensive opening transistor, which must have the voltage Uce (SVS) equal to that of the power switching transistor and the collector current I at least equal to that of the power switching transistor;

- As with the variant under point 1, there must be an electrically isolated supply section and the signal transmission must take place via the optoisolator.

This type of control is used in transistorized devices for setting the speed for DC servomotors from Siemens of the types: GRB 2012, GRB 2025, GRB 2030 and the servo device for setting the speed FANUK (documentation code A205-0089-0320 / 01).

3. Often opening a power transistor with a PNP transistor is used. In the majority of applications, galvanic isolation is not necessary, since it can be controlled at the low voltage level of the next transistor that opens the PNP transistor. This solution has been used by General Electric in the HI-AK transistor servo drive, which is protected on the basis of US Pat. No. 3,883,786, for several years. Fundamentally the same solution for controlling the power switching transistors can also be found in US Pat. Nos. 3,652,913 and 3,560,829.

In many applications, such as bridge circuits, this type of control avoids the need for electrical isolation of the power switching NPN transistor.

In this way, a power transistor can be controlled well, thereby reducing losses during the conducting state.

The unfavorable side of this type is shown at higher voltages, since it has a poor efficiency due to loss of power at the resistor for the base current limitation by the PNP transistor. There are large switching losses and long switch-off times.

4. There are several solutions for controlling a power switching PNP transistor and a characteristic example can be found in the book of Motorola Semiconductor Product Inc., Semiconductor Power Circuits Handbook, Nov. 1968, pp. 6-29.

In this way, the power NPN transistor can be opened with a PNP transistor. This Darlington transistor can be brought into a higher saturation state and thereby losses due to the voltage drop across the transistor are reduced when it conducts. The following disadvantages can be seen:

- There are no commercial PNP transistors for high voltages and currents;

- The opening transistor must have the same voltage Uce

(SVS) like the power switching transistor and the collector current Ic must be in the order of the base current of the power switching transistor, which is why these transistors are expensive;

- The turn-off times are not minimal, since the blocking is not carried out with negative current, and both transistors are uncontrolled in saturation;

- Optoelectronic isolation of the signal and a galvanically isolated supply unit for the base current are also necessary here.

5. Controlling the base of the power switching transistor via the transformers enables maximum utilization of the transistor in all operating states, so that it is used more and more.

a) Examples of such an electrically isolated control are given in principle by the Westinghouse company in the publication: Merle Morozovich: Where and How to Use the D60T Transistor, p. 30, Fig. A. This is detailed in Techn. Tips 2-9, M. Morozovich transistorized 50 kHz Popcorn Popper; Westinghouse Power Semiconductor Division, Youngwood, PA 15 15976. Disadvantages are as follows:

- It is not possible to work at a low frequency or even in the operating state of a continuously conducting power transistor due to saturation of the transformer core, which leads to an interruption in the energy on the secondary side. In order to be able to work at lower frequencies, large transformers must be used for the base current over the entire mass and the design is no longer economical.

- It is not possible to work in an operating state in which the transistor is conductive for most of the period and is only blocked for a short time because of the DC bias of the ferromagnetic core. This means that work can only be carried out in an operating state in which the leading state time is 80% of the total period or less.

b) The Motorola company has similar solutions according to the Switchmode Application Manual, Switzerland, 1980. Several possible types of transistor blocking are used, for which it can be determined that they are technically clearer than those according to the literature cited above. In both cases, good and bad properties are similar.

c) The solution according to German patent application P 27 29 407 has the advantage of the above-mentioned solutions that diodes against oversaturation are not necessary, since the collector current itself also opens the transistor. This ensures a short "storage time". Little energy is needed to open the transistor on the primary side. This variant has similar disadvantages as in Example 5a.

d) A similar solution as in 5c is provided in John Klimek's article: Reverse magnetic spike driver switching transistor effi-ciently, Pretoria, South Africa, Electronics, Dec. 29, 1981, p. 94.

The essence of this solution is that, in addition to being opened with the collector current, the transistor is also blocked with the magnetic energy accumulated in the transformer. With this solution, the problem of low frequencies also arises in a working state in which the time of the leading state is much longer than the blocking time.

e) The solution that Helmut Knöll deals with in his doctoral dissertation enables all operating states of the transistor with minimal static and switching losses and maximum speed.

- Of all the solutions dealt with, this is the best from a technical point of view, since the requirement for optimal control of the transistor is fulfilled.

The main disadvantages are the implementation and the price of the overall solution, namely:

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

665 515

4th

- The current source is designed with a transistor and other necessary components and behaves like a constant source with minor changes in impedance. If there is a major change in the circuit impedance during commutation, the source behaves like a choke. A short circuit must therefore be carried out when switching off and only then will the transistors for the supply of the transformers be blocked. The short circuit must last all the time so that the power generator is ready for power supply; in the interval in which the power switching transistor is blocked, all the energy on the power generator is therefore consumed.

- The power generator must have a control circuit, which is why it is expensive.

- Transistors, which supply the current through the transformers, experience high voltages during commutation, which is not mentioned by the author in his article.

f) A technically even better solution was published by H. Boe-Heiniger and F. Bruger in the article: Transformerless transistor pulse converters with output powers up to 50 KVA, E and M, 1979, pp. 539-545.

The same solution is also given in the article: R. Würslin, Transistor converter operation on 380 V three-phase mains, 2nd Annual European Power Conversion Conference, Sept. 3-5, 1980, Munich.

These solutions have no functional disadvantages,

however, due to the high price of the MOS FET transistors used, which can exceed the price of the controlled power bipolar transistor, they are uneconomical. In this case, the MOS FET transistors work partly in an analog and partly in a switching mode and a fairly large amount of power is consumed on them.

The object of the invention is to provide a circuit arrangement for controlling a power switching transistor with galvanic isolation, which does not have the disadvantages of the prior art mentioned.

According to the invention the object is achieved with the characterizing features of the patent claim.

The invention is explained in more detail with reference to the drawing. Show it:

Fig. 1 is a circuit diagram of an embodiment of the circuit arrangement, and

Fig. 2 is a timing diagram for the circuit arrangement.

The input terminal VH1 is connected to the base of the transistor T1, whereas the emitter is connected to ground and the collector to the common point of the Zener diode ZD1 and the primary winding N1 of the transformer TRI. The second connection of the winding Nl is connected to the current generator Igl. Between the connections of the winding N1 there is a series connection of the diode D9 and the Zener diode ZD1 and between the connections of the winding N2 of the transformer TRI there is a series connection of the resistors R1 and R2, the connection point of which is connected to the base of the transistor T4, the emitter being connected to the common point of resistor R2 and winding N2. The collectors of the transistors T4 and T5 are connected together at a common point to the connection of the winding N8 of the transformer TR3, the emitter of the transistor MST and the diode Dmst. The diode D1 is connected to the common point of the resistor R1, the winding N2 and the winding N3 and the diode D2 is connected to the other connection of the winding N3.

In the same way, a transformer TR2 is connected to a connection of the secondary winding N6 the diode D4. The diodes D2 and D4 are connected at a common point to the collector of the transistor MST and the diode Dmst. Between the connections of the secondary winding N5 of the transformer TR2, the resistors R3 and R4 are connected in series, their common point is with the base of the transistor T5, the emitter with the common point of the winding N5 and the resistor R4 and the collector with the collector of the Transistor T4 connected.

The common point of winding N6, winding N5 and resistor R3 is connected to diode D3. The diode D2 and the diode D3 are connected together to the collector of the transistor MST.

The input terminal VH2 is connected to the base of the transistor T2, the emitter being connected to ground and the collector being connected to the common point of the Zener diode ZD2 and the connection of the primary winding N4 of the transformer TR2. The second connection of the winding N4 is connected to the current generator Ig2. There is a series connection of the diode DIO and the Zener diode ZD2 between the connections of the winding N4.

The input terminal VH3 is connected to the base of the transistor T3, the collector being connected to the common point of connection of the winding N7 of the transformer TR3 and the Zener diode ZD3. A series connection of the diode Dil and the Zener diode ZD3 is connected between the connections of the winding N7. The common point of the diode Dil and the connection of the winding N7 is connected to the terminal + V. The diodes D8, D7, D6 and D5 are connected in series between the connection of the winding N8 of the transformer TR3 and the base of the transistor MST.

The diode Dmst is connected in parallel with the collector and the emitter of the transistor MST; these two points simultaneously form the output terminals IZ1 and IZ2 of the circuit, as shown in FIG. 1.

When the command for conducting the power switching transistor MST is received, the transistor T1 or T2 is opened. If the transistor T1 opens, a current flows through the primary winding N1 of the transformer TRI, which current is determined by the current generator Igl. A secondary current must flow on the secondary side of the transformer TRI, since it works like a current transformer. At first, a current flows through the resistor R1 and the base of the transistor T4 and therefore the transistor T4 opens. Now the current begins to flow through the diode Dl into the base of the transistor MST, which therefore opens. If the transistor MST conducts, the base current Ibmst becomes smaller since the secondary current of the transformer TRI begins to terminate via the diode D2 and the collector of the transistor MST. Then the load current and part of the secondary current of the transformer flow through the collector of the transistor MST. It depends on the load current Ib how much of the secondary current Is2 will flow through the base of the transistor MST and how much of this current will flow through the collector. The greater the load current IB, the greater the saturation voltage Uce and the less current will be terminated via the collector of the transistor and the more current through the base of the transistor MST. As a result, the voltage Übe of the transistor MST will increase and therefore the current through the base of the transistor T4 will also be higher. As the base current Ib4 will increase, the voltage Uce4 will decrease and accordingly the voltage on the secondary winding of the transformer TRI will not change significantly. The number of turns of the extended winding (N3 and N5) on the secondary windings of the transformers TRI and TR2 depends on which saturation voltage (Uce) is to occur at the transistor MST. The higher the number of turns (N3 and N5), the higher the saturation voltage; it moves between N2 / 2 (N4 / 2) and zero. Zero mainly occurs in power Darlington transistors.

Since the core of the transformer TRI would saturate if the transistor T1 were open too long,

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

5

665 515

the transistor TI must be blocked and the transistor T2 opened a little earlier. Commutation is carried out in the time when both transistors (T1 and T2) are conducting. Since, due to the inductance of the lines and the leakage inductance of the transformer, the full value of the current cannot flow and it cannot drop to zero for the same reason, the base current can be achieved with a correctly selected time of overlap (commutation) of the transistor MST will be without current indentations (see FIG. 2). This means that the base current will not drop below the necessary value during commutation. Normally, the commutation is extended by so much that the sum of both currents during commutation is greater than the normal base current that flows to the base of the transistor MST if only one transistor is conducting (T1 or T2). It is particularly important that two current generators are used (Igl, Ig2), as this eliminates the mutual influence between the two branches. If the transistor T2 conducts and the transistor T1 is still blocked, the identical picture is repeated as in the case when only the transistor T1 conducts.

During the time when the transistor MST is in the on state, the on state of the transistors T1 and T2 is changed at a frequency, an overlap being carried out continuously.

The role of the Zener diodes ZD1 and ZD2, which are connected in parallel to the primary windings N1 and N4 of the transformers TRI and TR2, is also important. The diodes D9 and DIO prevent the current through the Zener diodes ZD1 and ZD2 when the transistors T1 and T2 are in the conductive state. When the transistors T1 and T2 are blocked, a voltage of opposite polarity and with a height which is limited by the threshold of the Zener diodes ZD1 and ZD2 is induced on the primary sides of the transformers. The energy that was previously stored in the transformers TRI and TR2 during the conduction of the transistors T1 and T2 is now consumed at the Zener diodes.

The voltage of the Zener diodes ZD1 and ZD2 must be high enough so that during the pause, i.e. all the magnetic energy is consumed during the period in which the transistors T1 and T2 are blocked. This ensured that the core would saturate in the following periods due to DC bias.

The transistor MST is blocked in such a way that both transistors T1 and T2 are blocked and transistor T3 is opened for a short time. This ensures a fast blocking of the transistor MST, since it is blocked with a negative voltage between the base and the emitter of the transistor MST. How long this voltage will last depends on the transformer TR3. There is a problem similar to that of transformers TRI and TR2, the diodes D5, D6, D7 and D8 prevent the current through the transformer TR3 when the transistor MST is conductive.

The role of the transistors T4 and T5 is as follows: The transistor MST with the diode Dmst connected in parallel can also work only as a diode. If the transistor MST is used with an integrated reverse voltage diode Dmst, this has a voltage drop in the forward direction of up to 5 V, which is much more than with discrete diodes and the same current.

Because of the voltage drop across the diode Dmst, the current would also flow through the two secondary windings

Terminate transformers TRI and TR2 as well as diodes D2 and D4. Since this leading state of the diode Dmst is of any length, which is primarily dependent on the type of use of the power switching transistor, the cores of the transformers TRI and TR2 reach the saturation state. The resistance value of the secondary windings of the transformers TRI and TR2 is very small and therefore the voltage drop across the diodes D2 and D4 is higher. Since these diodes are designed to be discrete and cause a small voltage drop, a large current flows through the diodes D2 and D4, which would normally flow through the diode Dmst and the transistor MST and thermally destroy them both.

If the transistors T4 and T5 are used, however, the transistors Tl and T2 and accordingly also the transistors T4 and T5 can be blocked in such a mode of operation (operation of the diode Dmst). Now the impedance in the circuit is high and the entire current is terminated by the diode Dmst.

Even if the circuit for opening the transistor MST works, i.e. that the transistors T1 and T2 commutate, the transistor MST with the diode Dmst can be used as a diode. This time, because of the current through the diode Dmst, the collector current of the transistor MST is zero; therefore a base-emitter voltage of zero is necessary. Since the secondary current is forced, the resistor R1 and the base of the transistor T4 in the first period (see FIG. 1) and the resistor R3 and the base of the transistor T5 in the second period of the switching of the transformers TRI and TR2 only flow such a current that the collector current of transistors T4 and T5 represents all of the remaining secondary current. This part of the voltage, which represents the difference between the voltage drops across diodes D2 and D4 and diode Dmst, now occurs at transistors T4 and T5.

Two transistors are used because a transistor that could functionally replace two does not terminate a larger current during commutation. You would have to dimension one transistor for maximum current, which is why it is better to use two for lower current, since they can be dimensioned very precisely.

The electrical isolation with the power switching transistor and the control electronics can be achieved with an insulation between the primary and the secondary windings of the transformers TRI, TR2 and TR3.

The proposed solution enables the control of a power transistor in very demanding working conditions when using inexpensive components, which also significantly cheaper the entire device. This primarily means the transistors TI, T2 and T3, which are bipolar low-voltage switching transistors. This circuit offers the greatest advantage if a power Darlington transistor with an integrated diode is used for the power switching transistor MST. These transistors have a slightly higher saturation voltage and the same gain as discrete transistors.

By using the proposed circuit, these can be used in all working conditions, but at the same time the efficiency of the devices improves because of lower losses on the power Darlington transistor.

5

10th

15

20th

25th

30th

35

40

45

50

55

60

v

1 sheet of drawings