Voltage control system

This invention relates to a system for deriving a d.c. voltage from an alternating supply of variable frequency and amplitude.

It is known to derive a regulated output voltage from a variable alternating supply by suppressing the whole or part of selected cycles of the supply. In US-A-3982175 a load is supplied with alternating current and the power applied to the load is regulated by periodically interrupting positive half cycles of the alternating supply. This reference is concerned only to regulate an alternating current, and where the load is required to be supplied with direct current, the voltage control system must include means for rectifying the interrupted alternating supply.

It is known from US-A-,3991358 to derive a substantially constant dc output voltage from an alternating input voltage of varying frequency and amplitude. This result is obtained by arranging that, if a positive-going pulse of the input voltage exceeds a limiting value, the input voltage is short-circuited for a time which is less than half of the duration of that positive-going pulse. It is a disadvantage of this prior art that the range of input frequency over which the output voltage can be controlled is limited.

It is an object of the invention to derive a substantially constant dc voltage from an alternating supply of widely variable frequency and amplitude.

In particular the invention relates to a system for supplying a voltage to a spark igniter arrangement for a gas turbine engine, where this voltage is to be derived from an alternator which is driven by the engine. In such an arrangement the output of the engine-driven alternator will vary considerably in frequency and amplitude. If the arrangement is to provide an adequate spark discharge at low voltage, such as when the engine is started, the spark discharge at higher engine speeds may damage the igniter.

According to the invention there is provided a system for deriving a substantially constant d.c voltage from an alternating supply of variable frequency and amplitude, comprising a circuit responsive to a detectable point in each cycle of the supply for generating a control signal whose duration depends on the number of said detectable points which occur in a predetermined time, a switching arrangement responsive to said control signal for interrupting said alternating supply for the duration of said control signal, and a circuit for deriving said substantially constant d.c voltage from the interrupted supply.

An embodiment of the invention will now be described by way of example only and with reference to the accompanying drawings in which:-

Figure l is a block diagram of a spark discharge arrangement incorporating a voltage control system according to the invention,

Figure 2 shows a control signal generating circuit forming part of Figure l,

Figure 3 shows a switching circuit forming part of Figure l,

Figure 4 shows a spark voltage output circuit forming part of Figure l, and

Figures 5-8 show wave forms generated at a plurality of frequencies of alternating supply.

As shown in Figure l a generator l0 which is driven by a gas turbine engine (not shown) supplies an alternating voltage on lines ll to a spark voltage output circuit l3 by way of a switching circuit l4 and a transformer l5. The circuit l3 supplies a voltage pulse to each of two spark igniters l6. The switching circuit l4 is responsive to control signals T on pairs of lines l7, l8 and l9, 20 from respective identical control signal generating circuits 2l, 22. The circuits 2l, 22 are responsive to signals on a line 23 from one of the lines ll and also to a signal on a line 24 indicative of a requirement to energise the igniters l6, the signals on line 24 being derived from a +5v supply in a manner to be described.

As shown in Figure 2 the circuit 2l includes a Schmidtt trigger circuit 30 and a zener diode 3l to both of which the signal on line 23 is applied by way of a resistor 32 and an amplifier circuit 29. The threshold of the trigger circuit 30 is close to 0v so that a positive pulse is provided on a line 33 at a detectable point on each cycle of the supply on lines ll, this point being the negative-going zero crossing point of the alternator output. The amplifier circuit 29 ensures that if the supply on line ll becomes erratic in such a manner that the magnitude of a negative-going excursion is small, that excursion will nevertheless be detected and the pulse on line 33 be provided. The pulse on line 33 is applied to a clock terminal of D type bistable 34 whose D terminal is maintained at +5v, giving an output signal QA which goes positive in response to the clock pulse on line 33. This output signal QA is applied through a timing circuit 35 and inverter 36 to the clear terminal CLR of the bistable 34, the arrangement being such that the output QA on a line 37 persists for 0.6 milliseconds in response to the clock signal CK on line 33.

The signal on line 37 is applied to the D input of a further D-type bistable 38 which is also clocked by the signal on line 33 and is enabled when the signal on line 34 is at +5v. The arrangement is such that the bistable 38 provides an output signal QB which is low only when the signal on line 24 is +5v and when a clock signal on the line 33 occurs outside one of the 0.6 msec signals on line 37. The QB signal is applied by way of a line 39 and a diode 40 to an oscillator circuit 4l which provides a l.5MHz output when the signal QB is low. Three parallel amplifiers 42 are responsive to signals from the oscillator 4l to provide an increased output by way of a capacitor 43 to the primary of a step-up isolating transformer 44.

The output voltage of the transformer 44 is half-wave rectified by a diode 45 to provide a d.c. voltage across lines l7, l8. A depletion-mode FET 46 is connected between the lines l7, l8. Alternate half cycles from the transformer 44 pass through a resistor 47 and a diode 48 and maintain a voltage on the gate of the FET 46 which render it non-conductive. Voltage on the gate of the FET 46 is smoothed by a capacitor 49.

As shown in Figure 3 the circuit l4 includes two enhancement-mode FETs 50, 5l having their sources commonly connected to the line l8 and their gates connected to the line l7. Two further enhancement mode FETs 52, 53 have their sources commonly connected to the line 20 and their gates to the line l9. In the presence of the rectified signals T on lines l7, l9 the FETs 50, 5l and the FETs 52, 53 respectively are biased to permit current flow to the primary of the transformer l5 (Figure l). When FET 46 and the corresponding FET in circuit 22 are conductive the source-gate capacitance of the FETs in the circuit l4 is shorted out, ensuring that these devices have a rapid switch-off response when the signals T on line l7, l9 are removed. The source-gate capacitance of these FETs is effective to smooth the half-wave rectified d.c. on the lines l7, l9.

As shown in Figure 4 the secondary winding of the transformer l5 is connected through a known form of rectifier and voltage doubling circuit 60 to an earth rail 6l and to a high voltage rail 62. A spark discharge device 63 is connected between the rails 6l, 62. Capacitors 64, 65 are connected between the rail 62 and lines 66 and 67 which are in turn connected to the rail 6l through respective identical diode arrangements 68, 69, such that the lines 66, 67 are prevented from becoming negative with respect to the earth rail 6l. Identical resistor chains 70, 7l of 3k ohms each are connected in parallel with the respective diode arrangements 68, 69. Lines 66, 67 communicate by way of respective chokes 72, 73 whith respective ones of the spark igniters l6 (Figure l). In use the voltage doubling circuit 60 responds to alternating output from the transformer l5 to increase the voltage on the rail 62 to -3kV, charging the capacitors 64, 65 correspondingly, the lines 66, 67 being at this stage close to earth potential by virtue of the low resistance of the chains 70, 7l. The spark discharge device 63 breaks down at -3kV, connecting the rail 62 to earth and bringing the potential on the lines 66, 67 to +3kV which is discharged through respective ones of the igniters l6 (Figure l). Since the diode arrangements 68, 69 prevent the rails 66, 67 from going negative the spark discharge does not continue after the voltage on these rails has dropped to earth potential.

Referring back to Figure l identical isolating circuits 80, 8l are responsive to respective identical control signals Cl, C2 from a controlling computer (not shown). The circuit 80 includes a light-emitting diode 82 energised from a +5v source only when the signal Cl is low, in which circumstance a photo transistor 83 is switched on and connects the line 24 to earth, maintaining the output QB of the bistable 38 permanently high, whereby the FETs in circuit l4 permanently interrupt current flow. When the signal Cl is high the transistor 83 is off and the signal on line 24 is +5v, enabling the bistable 34 to respond to clock signals on the line 33. Since the line 24 from circuit 80 is connected to the corresponding line from circuit 8l, it will be apparent that both signals Cl, C2 must be high to permit the igniters l6 to operate.

Figure 5 shows the signals QA, QB from the bistables 34, 38 respectively when the frequency of the alternating supply S is less than l667Hz, that is when the time for one alternating cycle exceeds the 0.6msec output of the signal QA. As described with respect to Figure 2 the negative-going cross over of the alternating supply S provides a clock signal CK to bistables 34, 38. The output QA goes high for 0.6msec, and because of propagation delay, indicated at d, in bistable 34 the D input of bistable 38 is not high during the initial clock signal CK. The output QB thus remains low. The same condition obtains at the subsequent clock signals CK so that QB remains low. The oscillator 4l is permanently energised and maintains a signal T on line l7 to keep the FETs 50, 5l switched on, so that alternating supply to the transformer l5 is not interrupted. The foregoing also applies to the circuit 22 and the FETs 52, 53.

Figure 6 shows signals QA, QB when the interval between clock signals CK is just less than 0.6msec, that is with a supply frequency just over l667Hz. Initially QA goes high and QB stays low, as before. At the next succeeding clock signal CK QA is high and QB is thereby set high, and remains high until a clock signal occurs when QA is low. The effect is that when only one clock signal occurs during the time when QA is high the signal QB is high for one cycle of the supply S, which is therefore interrupted for that time.

Figure 7 shows signals QA, QB when two clock signals occur in the 0.6msec high level QA signal, that is when the frequency of supply S is greater than 3333Hz. It will be seen that QB goes high to interrupt the alternating supply for two cycles in every three thereof.

Figure 8 shows four clock signals occurring in the 0.6msec high level of signal QA, that is when the supply frequency is greater than 6667Hz. The supply S is interrupted for four cycles out of every five.

The arrangement thus reduces the duration of input to the transformer in a stepwise fashion with the increase in supply frequency which results from an increase in alternator speed. Since the rms voltage of the supply S also increases with alternator speed the general effect is that the product of rms voltage and duration of its application to the transformer l5 remains within relatively narrow limits at frequencies over l667Hz.