High Efficiency Quasi-Square Wave Drive Circuit for Switching Power Amplifiers

1. Field of the Invention

The present invention relates to drive circuitry of the type of used for switching power amplifiers, and in particular to drive circuitry capable of delivering, to a capacitive load, a voltage substantially in the form of a square wave (quasi-square wave) at high frequency and high efficiency.

2. Description of the Prior Art

In recent years not only has the number of different versions of MOSFET devices and MOS-controlled devices increased, but also the applications of such devices have increased. Examples of such devices are insulated-gate bipolar transistors and MOS-controlled thyristors. Each of these devices has the potential of operating at higher frequencies than earlier technology, and each is a voltage-controlled device. The control input of such devices usually has a capacitance associated therewith, and thus appears as a capacitive load to the drive circuit. The control input, for example, may be the gate of an MOS transistor.

For the most efficient operation of these devices as switches, a voltage square wave must be supplied to this input. This means that the drive circuit must deliver pulses of current to alternatingly charge and discharge the capacitance. Prior art drive circuitry of this type have made use of a so-called totem pole, push-pull arrangement to deliver these current pulses.

A drive circuit of this type exhibits losses equal to F.multidot.C.multidot.V.sup.2, wherein F is the operating frequency of the drive circuit, C is the capacitance of the driven device (i.e., the capacitance "seen" by the drive circuit), and V is the peak-to-peak voltage applied to the input of the driven device. At higher operating frequencies, or in applications wherein a number of driven devices are connected in parallel, these losses in the drive circuit can become appreciable.

It is an object of the present invention to provide an improved drive circuit for driven devices which appear as a capacitive load to the drive circuit.

It is a further object of the present invention to provide such a drive circuit for MOS-gated power devices.

Another object of the present invention is to provide a circuit capable of applying a voltage in the form of a quasi-square wave to the input of such devices.

It is a further object of the present invention to provide a drive circuit which eliminates the F.multidot.C.multidot.V.sup.2 losses associated with conventional drive circuits, and thus to permit operation at higher frequencies and at a higher efficiency than has been achievable with conventional drive circuits.

The above objects are achieved in accordance with the principles of the present invention in a drive circuit which includes two switching transistors having respective control inputs (gates) which are supplied with input signals that are 180.degree. out of phase, but are otherwise substantially identical. Each of these switching transistors has a voltage threshold which must be reached before the transistor will be switched to a conducting state. Each transistor respectively contributes to the drive of the driven device for one-half of a complete cycle. The circuit constructed in accordance with the principles of the present invention includes a combination of components which, for each half-cycle, cause the input capacitance of the driven device to be charged during a time when the voltage at the control input of the switching transistor responsible for that half-cycle is increasing, but has not yet reached the threshold voltage. The driven device input capacitance is thus charged during a time when that switching transistor is not conducting. Because at the end of that interval the voltage across the controlled path (drain-source path) is substantially zero, the F.multidot.C.multidot.V.sup.2 switching loss exhibited by conventional drive circuits is eliminated when the switching transistor begins to conduct.

The drive circuit is supplied with a square wave input, and delivers a quasi-square wave input to the driven device. As used herein, the term "quasi-square wave" means a signal having a waveform which closely resembles, but may deviate from, a precise square wave, with the deviations being such that the ability to drive a device which is conventionally driven with a true square wave is not impaired.

In one embodiment of the invention, the out-of-phase input signals which are respectively supplied to the inputs of the two switching transistors are obtained by supplying a true square wave to the primary winding of a transformer having two secondary windings, the secondary windings having the same number of turns, but having respective output terminal pairs of opposite polarity.

FIG. 1 is a schematic circuit diagram of an exemplary embodiment of a drive circuit constructed in accordance with the principles of the present invention.

FIG. 2 is a voltage-time diagram showing the shape of an output signal, supplied to a driven device, obtained with the circuit of FIG. 1.

An example of a drive circuit constructed in accordance with the principles of the present invention is shown in FIG. 1. A low-amplitude square wave drive signal is supplied as an input signal to the primary of a transformer T1. The transformer has two secondaries. One terminal of one of these secondaries is connected to the source of a switching transistor Q1, with the other terminal of this secondary being connected to the gate (control input) of the transistor Q1 through a resistor R1. One terminal of the other secondary is connected to the source of a further switching transistor Q2, with the other terminal of this secondary being connected to the gate of the transistor Q2 through a resistor R2. The resistances of R1 and R2 are equal. The respective terminals of the transformer T1 are numbered in FIG. 1, with conventional polarity designations. The two secondaries have equal turns, but are of opposite polarity, so that the respective signals supplied to the gates Q1 and Q2 are 180.degree. out of phase, but are otherwise substantially identical.

The drain of the transistor Q1 is connected to a source of positive voltage +V. The source of the transistor Q2 is connected to a source of negative voltage -V. The source of transistor of Q1 and the drain of Q2 are both connected to the input of a device to be driven, which in the embodiment of FIG. 1 is shown to be the gate of a further transistor Q3. The gate of transistor Q3 is shunted to ground via an inductor L1. The transistor Q3 may be a part of a separate switching circuit, the remainder of which is not shown in FIG. 1, whose operation is not relevant to the explanation of the present invention. The input of the driven device has a capacitance associated therewith, in this case the gate capacitance of transistor Q3.

In operation, the drive signal supplied to the primary of transformer T1 must be of sufficient amplitude so that the voltages supplied to the gates of switching transistors Q1 and Q2 are of an amplitude to cause those switching transistors to behave as switches. For most MOSFET devices, a peak voltage in the range of 10 to 20 volts is sufficient. It is also necessary that this drive signal be a square wave. Such a drive waveform will cause transistors Q1 and Q2 to alternately be switched to a conducting state. During the time that the switching transistor Q1 is conducting, the gate voltage of transistor Q3 is held at +V, and during the time that switching transistor Q2 is conducting, the gate voltage of transistor Q3 is held at -V. As this voltage is applied to the inductor L1, it has an associated current in the form a triangular sawtooth waveform. The resistors R1 and R2, in combination with the gate-source capacitances of switching transistors Q1 and Q2, form respective low-pass networks which control the voltage slew rate appearing at the respective gates of transistors Q1 and Q2. Because transistors Q1 and Q2, being n-channel MOSFETs, each have a switching threshold voltage on the order of three volts, periods exist twice per drive cycle during which neither transistor Q1 nor transistor Q2 is in a conducting state. During these periods, the current from inductor L1 is diverted into the gate of transistor Q3 in a pulsed waveform, thereby rapidly charging the gate capacitance of Q3. By properly selecting the resistance value for R1 and R2, and the inductance value for L1, the peak current in the inductor L1 will passively charge the gate capacitance of the transistor Q3 from -V to +V during a dead-time in which neither transistor Q1 nor transistor Q2 is in a conducting state the value of L1 is chosen in conjunction with the gate capacitance of Q3 to provide a specific transition time which is equal to this "dead-time" of Q1 and Q2. Thus the inductor L1 provides all the energy to charge the gate capacitance of Q3, and Q1 and Q2 serve only to "recharge" the inductor L1 between transitions. The transition consists of the rise (or fall) of the sinusoid have a peak amplitude in excess of the total transition voltage and also having a natural frequency set by L1 and the gate capacitances of Q3. The transition ends when Q1 or Q2 turns on. As can be seen from the L1 equation below, a reduction in the value of L1 results in an increase in the natural frequency F. This has the adverse effect of creating greater conduction losses in Q1 and Q2 according to the F.multidot.C.multidot.V.sup.2 equation. At the end of this time interval when either transistor Q1 or transistor Q2 is switched "on", their respective drain-source (controlled path) voltages are substantially zero. Thus, the F.multidot.C.multidot.V.sup.2 switching loss is eliminated fulfilling an object of the present invention.

The drive circuit shown in FIG. 1, operating as described above, delivers a quasi-square wave output to the driven device, i.e., to the gate of the transistor Q3, as shown in FIG. 2. In general, the waveform can be described as a sinusoidal waveform having flattened peaks, i.e. somewhat trapezordal, with the sinusoidal portions of the waveform not deviating from a true square wave shape to an extent which impairs the ability of the drive circuit of FIG. 1 to operate conventional devices requiring a square wave input.

The component values for the circuit shown in FIG. 1 are determined by the following equations: ##EQU1## wherein V.sub.gs(th) is the threshold voltage of the transistor Q1, V.sub.d is the peak amplitude of the drive voltage at the gate of transistor Q1, C.sub.gsQ1 is the gate-source capacitance of transistor Q1, C.sub.gsQ3 is the gate-source capacitance of the transistor Q3, and .delta.t is the time during which the gate voltage of the transistor Q3 is charged from -V to +V. The above equations are valid for the conditions V.sub.d >2 V.sub.gs(th).

Although modifications and changes may be suggested by those skilled in the art, it is the intention of the inventor to embody within the patent warranted hereon all changes and modifications as reasonably and properly come within the scope of his contribution to the art.