JPS6013387B2 - power control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in a power control device operated from a DC power source, and in particular to an improvement in a switching circuit having bridge-connected solid state switches.

First, we will discuss conventional problems when the solid state switch is a thyristor.

In forced commutation inverters that operate on series power supplies, bidirectional current-carrying DC chopper control devices (pillateral chopper), etc., the state side is responsible for dealing with external issues such as low voltages and fault currents.

In addition, the prior invention shown in FIG. 1a is known as an excellent prior invention in that the short-circuit current is small when commutation fails. However, this conventional device has the drawback that the voltage of the commutating capacitor 6 rises each time the commutation is repeated cyclically, and the voltage applied to the thyristor also becomes excessive. Furthermore, there is a drawback that the energy level L1 between the inductance L of the commutating reactor 4 and the load current 1 is ultimately ``missed''.

This conventional problem will be explained in more detail below. In FIG. 1a, it is assumed that, prior to the nth commutation, the charging voltage of the capacitor 6 was of the illustrated polarity v6(n).

and DC power supply terminal P-first reactor winding 4a-first
It is assumed that a load current 1 is being supplied to an AC load 7 or a DC load 7' via the thyristor 2a-output terminal OG. Now, in the operating waveform of FIG. 1b, when the extinguishing thyristor 3a is ignited at time t, the first thyristor 2a is reverse biased and extinguished.

Load current 1 flows through P-4a-3a-6-○. Furthermore, during this time, 4a-3a-6 - first rectifying element 5a-4
A closed loop consisting of a is formed, the capacitor voltage V6 is also applied to the first reactor winding 4a, and the capacitor current i6 and the first reactor current i4a both increase by a bit and then return to the initial current value 1. At the point when the initial current value returns to 1, the absolute value of the capacitor voltage v6 is the initial charging voltage value v6.
(n) and coincides with the time t2 when the polarity is opposite to the illustrated polarity. Now, it is assumed that the initial charging voltage value v6(n) is equal to or larger than the voltage E between the DC power supply terminals.

If it is assumed that v6(n)ZE, the rectifying element 5b becomes suitable at this point. However, since a current corresponding to the load current 1 (and thus significant energy) still remains in the first reactor winding 4a, the voltage v6 of the capacitor 6 continues to be charged to the opposite polarity as shown. and the first
At time t3 when the current in the reactor winding 4d and the capacitor 6 becomes zero, the capacitor voltage v6 becomes the final value v6 (
n+1)=v6(n) reaches △v6. This overcharge △v6 is △v6=
1 no LC...{11.

And the overcharging period (ra-t2) is Buddha t2):? Before
．．・・・．・It is a leg.

During this overcharge period, the current i5b of the rectifying element 5b increases from 0 to 1. If it is an inverter (AC load 7), then the second thyristor 2b is ignited, and the load current polarity is reversed. In the case of a DC chopper (DC load 7'), the load current polarity usually does not change even if the second thyristor 2b is fired. Next, the arc-extinguishing thyristor 3b is ignited, and the second thyristor 2b is extinguished.

At the (n+1)th commutation time of this strategy, the charging voltage obtained by adding up the previous overcharging amount △v6 {v6 (n) + △v6}
is the initial charging voltage, so the capacitor voltage increases cyclically each time commutation is repeated. As described above, the capacitor voltage increases extremely, and in the end, L12 of the reflector accumulated energy becomes a loss in the rectifying element, reactor, capacitor, thyristor, and other wiring, and the voltage increases until it is balanced with these losses.

Therefore, in the conventional device described above, in order to suppress the rise in capacitor voltage, it is necessary to intentionally provide a loss generating element in the reactor, arc-extinguishing thyristor, capacitor, rectifying element, etc.

In addition, as a method of preventing the capacitor voltage from rising, it is possible to
During the period when the current a remains within the initial current 1, the second
While igniting the thyristor 2b, the first reactor winding 4
a and the second reactor winding 4b are combined, and the above 2
There is a method in which the remaining stored energy is commutated to the third reactor 4b by ignition of b, and the remaining stored energy is short-circuited in a closed loop passing through the rectifying element 5, the second thyristor 2b and the second reactor winding 4b.

However, when such a method is used, the short total current flows through the thyristor 2 and the rectifying element 5, and this extra current increases the current burden on the elements. Moreover, in order to reduce this short-circuit energy, it must be consumed as circuit loss. Furthermore, when the superactor windings 4a, 4b are combined, this secondary embodiment has the disadvantage that the voltage applied to the thyristors 2a, 2b is doubled, as will be explained later. In this way, the conventional device
The capacitor voltage rises and 2a touches the candle for 12 minutes and 3
There is a problem with the voltage applied to the thyristor increasing.

Alternatively, a problem arises in which the lb storage energy is lost by 12 minutes and the voltage applied to the 2b thyristor is doubled. Now, in the above conventional example, the solid state switch is a thyristor.

In FIG. 1a, if the solid state switches 2a and 2b are self-off type switching elements such as transistors and gate turn-off thyristors, the commutating capacitor 6 and the arc-extinguishing thyristors 3a and 3b are unnecessary. In this case, instead of the voltage rise of the commutating capacitor, the problem appears as a surge voltage during self-CFF of the solid state switch 2a. On the other hand, the reactor 4a
, 4b, the switching responsibilities such as the shoe group and younger brother will become stricter.

In addition, the short-circuit current suppressing effect disappears when a small amount of simultaneous conduction of the series solid state switches occurs (due to ON state due to turn-off delay or failure). Therefore, it is desirable to provide a series reactor from the viewpoint of improving the reliability of the switching circuit and reducing switching duties. The present invention aims to improve a power control device using a solid-state switch in which reactors are connected in series, and to reduce the switching duty thereof. First, the circuit operation according to the embodiments shown in FIGS. 2 to 4 will be described in detail for the case where a thyristor is used as the solid-state switch.

Figure 2a is a connection diagram showing an embodiment of the present invention, in which la and lb are DC power supplies, P is the first potential terminal (positive terminal) of the DC power supply, and N is the second potential terminal (positive side terminal) of the DC power supply. negative terminal). 2a, 2b, 3a, 3b are a pair of bidirectional current-carrying terminals A, , A2 and a pair of unidirectional current-carrying terminals D, , D2
and 4a is a unidirectional current-carrying terminal D. of one of the thyristor prisms, and 4a is a thyristor that is fringe-connected between these terminals to form a four-arm thyristor bridge. A first reactor winding 4b is connected between the other unidirectional current-carrying terminal D2 of the thyristor bridge and the second potential terminal N of the DC power supply. 5a is a first rectifying element connected between one bidirectional current-carrying terminal (fluctuation potential output terminal) A of the thyristor bridge and the first potential terminal P; 5b is a A second rectifying element 6 is connected between the bidirectional current-carrying terminal A and the second potential terminal N, and 6 is a commutating capacitor connected between the pair of bidirectional current-carrying terminals A, . 7' is a load, 8a is a first coupling winding connected in series to the second rectifying element 5b and magnetically coupled to the first reactor winding 4a, 8b is connected in series to the first rectifying element 5a, and is connected in series to the first reactor winding 4a. This is a second coupled winding that is magnetically coupled to the second reactor winding 4b.

FIG. 2b is an operational waveform diagram for explaining the operation shown in FIG. 2a.

Next, the operation will be explained.

In FIG. 2, power source la-first reactor winding 4a-
The load current 1 is on the path ■ passing through the first thyristor 2a and the load 7.
Assuming that the current is being applied, the capacitor 6 is charged to the polarity shown in the figure prior to extinguishing the current.

Let this initial charging voltage be v8(n). Next, when the arc-extinguishing thyristor 3a is ignited, the first thyristor 2a is reverse biased and the load current 1 is changed from the first potential input terminal P to the first reactor winding 4a to the first arc-extinguishing thyristor 3.
Move to path ■ of a - commutation capacitor 6 - output terminal ○.

At the same time, a closed loop passing through P-4a-3a-6-first rectifying element 5a-second coupling winding 8b-P is formed, and the voltage of the capacitor 6 is connected to the first reactor winding 4a and the second coupling winding 8b-P.
A divided voltage is applied to the coupling winding 8b, and the current i of the capacitor 6
6 is increasing. When the voltage of the capacitor 6 passes through zero and becomes reverse polarity, the current of the capacitor 6 passes through the peak value. Continuing, capacitor 6 connects load current 1 and winding 4
The winding 8b, which is the release of the accumulated energy of the coils a and 8b, is charged with the reverse gutter characteristic as shown in the figure due to the sum of the current, and the first
The voltage of the thyristor 2a also increases in the negative direction. If circuit loss is ignored, the absolute value of the capacitor voltage is the initial value v
6(n) and the polarity is reversed at time t2, the capacitor current i6 and the first reactor winding current i4a
returns to the initial value 1, and the first rectifying element 5a becomes non-conductive.
Of course, at this time, the current in the coupling winding 8b also returns to zero as the energy is released. Now, if v6(n)≧E (however, E
is the voltage between DC power supply input terminals P and N), then the above time t
At step 2, the first rectifying element 5a is turned off, and the second rectifying element 5b and the first coupling winding 8a are energized, and the load current shifts to approximately (2).

And the capacitor voltage at this time v6(n+1)=i(n
) (ignoring losses). On the other hand, if v6(n) < E or when i4a=i6 returns to 1 due to loss, the capacitor voltage v6(t2) becomes
Then, capacitor 6 will continue to be charged with load current 1 and v
When 6=E is reached, r2 [Actually, it becomes v6(n)±E, so there is no big difference between L and r2. Therefore, the second rectifying element 5b and the first coupling winding 8a are energized at 1, which is not particularly distinguished in FIG. 2, and the load current 1 is transferred to the path (2). Here, in this invention, the first reactor winding 4a and the first coupling winding 8a are coupled, so if leakage inductance is ignored, the current i4a of the reactor winding 4a immediately flows into the coupling winding 8a. can do.

Thereafter, the current of the reactor is inherited by the coupling winding 8a, or the actuator's accumulated energy, in other words, is released to the load and used effectively. Therefore, unlike the prior art embodiment, a subsequent charging period (t'2-t3) does not occur. And additional charging △v
6 does not occur either. Therefore, the commutation phenomenon is completed at the above-mentioned time point t2'. Then, when starting from the startup initial value v6 (0) = E, this final capacitor voltage value is also v6
(n)=v6(n+1)=E. FIG. 2b shows the principle operation waveforms ignoring this leakage. If the first reactor winding 4a and the coupling winding are glued together, the cage inductance is connected to the continuous charging period (ra-t3).
This causes additional overcharging by Δv6=1. In this invention, the additional overcharge voltage Δv6 is actually 4.

This accumulation of overcharge (cyclic increase) can be prevented with a small loss (equivalent to F). That is, there is no need for a resistor or the like to intentionally prevent overcharging. And the loss itself will be small. Moreover, in this invention,
As shown by the dotted arrow in FIG. 2a, the reactor windings 4a and 4b can be combined and wound around the same magnetic core, allowing the reactors to be integrated. In this case, P-4a-3a
-6-5a-8b-P, the inductance L of the commutating capacitor bypass inversion loop is the first reactor winding 4.
Winding inductance L4a of a, winding inductance L3b of the first coupled winding 6b, mutual inductance M4a-
Since the combined value L=La+2M4a−E1La of 8a acts, the number of turns of each winding is reduced for the same required value L of commutation inductance. Therefore, even if a coupled winding is provided, the size can be reduced in combination with the above-mentioned integration. In this proposal, during the commutation period (t2), lb-la-4a-3a-6-
Regarding the loop of thyristors 2b-4b-lb during voltage blocking, capacitor voltage v6 and actuator winding voltages v4a and v
4b≠v8b cancel each other out.

That is, since (v4a+v4b)=v6, the voltage applied to the thyristor 2b during voltage blocking is v2b=E-(v4a + v4
b) +v6=E, regardless of whether reactors 4a and 4b are coupled. Therefore, as described above, 4a and 4b can be combined and made integrally. On the other hand, in the subordinate embodiment, when 4a and 4b are combined, v4a=v4b=-v8 (where v6>0) with respect to the same loop.

Then, the second half of the commutation period (v6ku0.v4a
= v4b 0), the voltage of thyristor 2b during voltage blocking is v2b = E - (v4a + v4b) + v6 = E - v6
! E1lv6l. And the wave height value is E+v6(
t2)=E+v6(n+1). Originally, the capacitor voltage v6 rose to a high voltage and v6(n11)>E, so
An extremely high overvoltage is applied to the thyristors 2b and 2a. If 43 and 4b are combined, the above problem will occur, so
In FIG. 1a of the secondary structure, 4a and 4b cannot be magnetically coupled (same energy, integrated).

FIG. 3 shows another embodiment of the present invention, in which a reactor winding 4 and its coupling winding 8 are used in common to form a three-phase fringe connection circuit.

In the figure, R, S, and T are three variable output potential terminals, respectively. The commutation operation during each commutation period for each phase in the example shown in FIG. 3 is the same as the example shown in FIG. 2 as a circuit phenomenon, so a description thereof will be omitted. The circuit and operation of the present invention have been described above.

Next, a reactor device used in the present invention will be explained with reference to an embodiment shown in FIG. 4.

The reactor device used in this invention reduces the cage inductance between at least one coil {4a, 8a} or {4b, 8b}, or at least improves the predetermined leakage for the evening-on control of the main thyristor 2. It is required to reduce it to the level of inductance.

When using a magnetic core such as an iron core or ferrite, or a core of an oven magnetic path, the supercarrier-cage inductance ratio can be reduced, and it is easy to set the coupling coefficient to the required value. However, in large-capacity devices or high-frequency devices, magnetic cores can cause problems such as being expensive, causing noise, and being difficult to construct.

In such cases, an air-core reactor device is desirable. Of course, a magnetic core may be included in a relatively small capacity device. However, it has been difficult to minimize leakage inductance in coupled reactor devices.

The embodiment shown in FIG. 4 of the present invention provides a winding configuration for a coupled IJ vector device with small leakage inductance, and as shown in FIG. A double conductive band coil is formed by sandwiching a conductive (copper, aluminum) strip with an insulating film or a green tape (not shown).

If the insulating film is an aluminum conductive band, an oxidized coating on the surface of the aluminum conductive band treated with alumite can be used. Moreover, in the case of a copper strip, in addition to the same insulating film as that for copper wires (round wires, rectangular wires), an insulating film formed by a drape coating method or the like can be used. Further, various types of insulating tape such as Mylar tape and mica tape can be used. For the lead wire, an IJ-type conductive band plate or the like is connected in advance to the end of the former conductive band. The double conductive band coil wound as shown in FIG.

In this way, a reactor coil with extremely low leakage inductance can be obtained. On the other hand, in high frequency devices, losses due to eddy currents and overheating become a problem.

FIG. 4b is a cross-sectional view of the cylindrical coil, where a magnetic flux 0i passing through the coils 4 and 8 is generated, and due to this magnetic flux 0i passing through the conducting band, eddy currents are generated at the ends of the conducting band (both sides of the band) A, A large amount flows to B, causing this region to overheat. On the other hand, as shown in the two-part sectional view of FIG. 4c, magnetic shunt plates ga and 9b are provided at both ends of the coil (both magnetic pole surfaces) to adjust the magnetic field.

There may be a hole in the center of the magnetic field shunt plate, and it is sufficient if there is a magnetic material near the end of the coil. The magnetic adjustment plates 9a, gb may be made of a magnetic mixture containing magnetic powder, such as epoxy resin or rubber material mixed with iron powder or ferrite powder.

Alternatively, a plate-shaped or thin donut-shaped ferrite core may be used. In the cross-sectional magnetic flux line diagram as shown in c of the same figure, the magnetic flux passing through the conductor (conducting band) decreases through the magnetic flux path mediated by the magnetization plate, as shown by ◇o in the figure d.

This can prevent overheating of the bottle current and thus the coil. Regardless of whether it is air-core or solid-core, a conductive band coil as shown in FIG. 4a is highly effective in reducing leakage inductance. Furthermore, it is effective as a coupled reactor device that performs the operation of the present invention described above.

Also, if the solid state switch 2 is a self-off type, the self-off
The surge energy at the time is reduced, and the current is transferred between the solid state switch 2 and the rectifying element 5.

In the case of the self-off type switching element 2, it goes without saying that the arc-extinguishing thyristor 3 and the commutating capacitor are unnecessary.
As described above, according to the present invention, by using a coupling reactor in which band-shaped conductors are wound overlappingly, the responsibility of the solid-state switch is reduced.

[Brief explanation of drawings]

Figures 1a and b are circuit configuration diagrams and operation waveform diagrams showing an example of a conventional device; Figures 2a and b are circuit configuration diagrams and operation waveform diagrams illustrating an embodiment of the present invention; A circuit configuration diagram showing another embodiment of the invention, FIG. 4 is a conceptual diagram of a coil structure of a reactor device used in this invention. In the figure, P is the first potential terminal of the current power source, N is the second potential terminal of the DC power source, 2a to 2f are solid state switching elements, 4a and 4b are reactor windings, 5a to 5f are rectifying elements, 7 is a load,
8a, 8b are coupled windings, A, , A2 are bidirectional current-carrying terminals,
D, , D2 are unidirectional current-carrying terminals, and 0 is a variable potential output terminal. Note that the same reference numerals in the figures indicate the same or equivalent parts. Figure 4 Figure 1 Figure 2 Figure 3

Claims (1)

[Claims] 1. A series body of at least two solid-state switches having at least one bidirectional current-carrying terminal and at least one pair of unidirectional current-carrying terminals and connected in series in the forward direction between these terminals;
A first reactor winding connected between one terminal of the unidirectional current-carrying terminal and the first potential terminal of the DC power source, and a first reactor winding connected between the other terminal of the unidirectional current-carrying terminal and the second potential terminal of the DC power source. and a second reactor winding connected thereto, and furthermore, the bidirectional current-carrying terminal is a fluctuating potential output terminal that provides a fluctuating potential by switching both solid state switches of the solid state switch series body, a first rectifying element connected between the first potential terminal of the power source and a second rectifying element connected between the output terminal and the second potential terminal of the DC power source; A first coupling winding connected in series and magnetically coupled to at least the first reactor winding; and a second coupling winding connected in series to the first rectifying element and magnetically coupled to at least the second reactor winding. , A power control device characterized in that the reactor winding and the coupling winding are each formed by winding band-shaped conductors in an overlapping manner.