GB2097606A

GB2097606A - DC to DC converter

Info

Publication number: GB2097606A
Application number: GB8112572A
Authority: GB
Original assignee: Standard Telephone and Cables PLC
Current assignee: STC PLC
Priority date: 1981-04-23
Filing date: 1981-04-23
Publication date: 1982-11-03
Also published as: GB2097606B

Abstract

In a conventional DC-to-DC converter of the single-ended type, it is usual to use what is often called a snubber network to eliminate high- voltage transients due to the fast switching pulse edge which occurs when the power or switching transistor is switched off. Such a network usually involves a capacitor in series with a resistor connected across the transistor, which resistor is shunted by a diode. In one circuit embodying the invention the capacitor (C1) is of the order of ten times the size of that usually used, with the result that in addition to the elimination of leakage transients, the core of the transformer (TF2) is totally reset, which enables the whole of the flux swing possible with the transistor to be used. In addition a much higher pulse duty cycle is usable than in a conventional circuit. In another arrangement (Figure 3) the capacitor is even longer - e.g. of the order of 100 times the size usually used, with a further switched transistor (TR4) in series with it. This eliminates the dissipation of energy which usually occurs in the series resistor. <IMAGE>

Description

SPECIFICATION DC-to-DC converter The present invention relates to DC-to-DC converters of the so-called single-ended type.

In conventional converters of the above type it is usual to use a snubber network to eliminate highvoltage transients due to the fast switching edge which occurs when the power transistor of the converter is switched off. Such a network usually involves a capacitor in series with a resistor connected across the transistor, there being a diode across the resistor. Further, it is necessary in such conventional circuits to use a reset winding on the output transformer, which winding is in series with a diode, and is so arranged as to return flux energy into the transformer core during the off period of the transistor. With this method the maximum pulse width usable is restricted to 50% of the pulse cycle, and the flux excursion usable is limited to about 25% of the flux swing available in that core.

According to the invention there is provided a DC-C converter of the type in which a power transistor has its emitter-collector path in series with the primary winding of an output transformer, wherein a capacitor connected across the power transistor in series with a semiconductor device serves by the current it passes when the power transistor is switched off to reduce or eliminate the effects of switching transients, and wherein the capacitor has a value which is high compared with the value normally used to reduce or eliminate transients, such that on switching the transistor off an overswing occurs as a result of which the core of the output transistor is reset to its off condition such that at least a substantial proportion of the flux swing available with the core material can be used.

Embodiments of the invention will now be described with reference to the accompanying drawing, in which Figure 1 is a simplified circuit diagram of a first example of a DC-to-DC converter embodying the invention, Figure2 is a waveform useful in explaining the operation of the circuit of Figure 1, Figure 3 is a second example of a DC-to-DC converter embodying the invention, and Figure 4 is a wave form useful in explaining the operation of the circuit of Figure 3.

In the circuit shown in Figure 1 we have a switching transientTR1 connected across the DC supply in series with a winding of a regenerative feedback transformer TF1 and a winding of the output transformer TF2. The latter via its secondary winding feeds a rectifier circuit with a choke input filter, the output direct voltage being developed across the output of this filter. Connected across the transistor TR1 and the winding of TF1, we have a capacitor C1 in series with the parallel combination of a resistor R1 and a diode Do. these latter resemble the snubber circuits much used in such converting but as will be seen later they differ therefrom.

In the conventional DC-to-DC converters used, for instance, to convert from 48 volts DC (the usual voltage of a telephone exchange battery) to, for instance 12 volts DC for transistor circuitry, the capacitor which corresponds to C1 has a relativeiy low value, e.g. of the order of 500 pf or 1000 pf. In such a conventional converter the snubber network is used to eliminate high voltage transients generated by the interaction of the fast switching edge which occurs when the transistor is switched off with the leakage inductance of the feedback transformer.

In addition, in such a circuit it is necessary to "reset" the core of the power transformer during the off period of the switching transistor, i.e. to return the flux excursion in the core to zero or remanence flux levels. This is usually done by an additional winding on the power transformer which is connected in series with a diode.

The known technique restricts the maximum pulse width usable for the on period to less than 50% of the cycle time, and also limits the usable flux excursion to 25% of the flux swing available with the core material used.

In the circuit shown, the component values are altered substantially to enable the reset winding (or windings in some cases) to be eliminated. Thus the capacitor C1 in Figure 1 has a value of the order of 10 000 pf, which is a decade higher in value then that used merely as a snubber. With such an arrangement it has been found that the leakage transients are eliminated, and the core is totally reset, which enables the core of a 100% flux swing, i.e. both positive and negative, and enables duty cycles of up to 80% to be attained. This enables the power through-put for a given core size to be greatly increased.

Referring to Figure 2, the on portion of the cycle is shown by the first (rectangular portion), with the remainder showing what happens on switch off.

Thus on switch off there is a region of "overswing" A - in which magnetic energy is removed from the core, and a region B in which magnetic energy is injected in the reverse direction. This is followed by a quiescent period C. The optimum case is where the areas under the waveform in the negative flux excursion equal the positive flux excursion in the "on" condition.

In the arrangement just described it will be appreciated that during the B portion of the curve in which the diode D1 is in its high impedance state some energy, albeit a small amount, is dissipated in the resistor R1. An extension of the principle, as will be seen in Figure 3, is to make the capacitor even larger, so as to eliminate this dissipation. This is desirable if high power has to be handled, when the presence of such dissipation can be inconvenient.

In Figure 3, the circuit includes the switching transistor TR3 whose emitter-collector path is in series with the primary winding of the output transformer TF3. There is also a feedback transformer TF4 which has a winding connected between the base amd the emmiter of another transistor TR4 having the capacitor C2 connected in series with its collector-emitter path, as shown. This capacitor has a value of the order of many microfarads, i.e. much greater than the capacitor used in Figure 1. In fact, it would normally be an electrolytic capacitor.

In this circuit, the switching transistor TR3 is switched on by a control signal applied to the winding W1 of transformer TF4 and hence to the base-emitter circuit of TR4. Note that although regenerative drive is used in this circuit, it would also be possible to apply the "enlarged capacitor" princi ple to a converter using a voltage drive. During the on period of the pulse which forms the control signal, TR4 is switched off. The control signal is also applied via another winding of TF4 to the base circuit of TR3 as a negative pulse, and this switches TR3 off so that its collector voltage goes positive. Avoltage is already present across the capacitor C2 which is larger than the supply voltage, its value being depenent on the duty cycle of the pulse supply unit.

With the collector voltage going positive, the negative side of C2 goes positive, and is clamped to the negative rail by the diode D2. Therefore current flows from the primary winding of TF3 into C2, and this continues until the magnetism energy has been completely removed.

While TR3 is being turned off, TR4 is turned on, but it does not conduct until the magnetising energy has been removed from the core. When this occurs the current direction in C2 reverses, and the current magnetises the core in the opposite direction to give a negative flux level from which the converter aperates.

Figure 4 shows the waveform, with the port A of the TOFF portion corresponding to removal of magnetic energy via D2 and port B corresponding to magnetic energy being input via TR4.

Claims

1. A DC-to-DC converter of the type in which a power transistor has its emitter-collector path in series with the primary winding of an output transformer, wherein a capacitor connected across the power transistor in series with a semiconductor device serves by the current it passes when the power transistor is switched off to reduce or elimin ate the effects of switching transients and wherein the capacitor has a value which is high compared with the value normally used to reduce or eliminate transients, such that on switching the transistor off an overswing occurs as a result of which the core of the output transistor is reset to its off condition such that at least a substantial proportion of the flux swing available with the core material can be used.

2. A DC-to-DC converter as claimed in claim 1, and in which the capacitor is in series with the parallel combination of a diode and a resistor, which diode is the said semiconductor device.

3. A DC-to-DC converter as claimed in claim 1, in which the capacitor is in series with the emitter collector path of a furthertransistorwhose base is controlled from a regenerative feedback transformer for the switching transistor.

4. A DC-to-DC converter substantially as described with reference to Figures 1 and 2 or Figures 3 and 4 of the accompanying drawings.

New claims or amendments to claims filed on 3 Sep 1981 New or amended claims:

5. A DC-to-DC converter of the type in which a power transistor has its emitter-collector path in series with the primary winding of an output transformer, wherein a capacitor connected across the power transistor in series with a semiconductor device serves by the current it passes when the power transistor is switched off to reduce or eliminate the effects of switching transients, and wherein the capacitor has value which is high compared with the value normally used to reduce or eliminate switching transients, such that on switching the transistor off an overswing occurs as a result of which the core of the output transistor is reset to its off condition, enabling at least a substantial proportion of the flux swing available with the core material to be used, and such that resetting the core of the power transistor is effected without the use of a reset winding.