GB2096328A - Apparatus for electrically measuring rotation - Google Patents

Abstract

Apparatus for measuring the rotation of an electrically-conductive disc (especially the disc of a watthour meter) having an aperture 216 therethrough comprises a pair of plates on opposed sides of the disc. A varying voltage is applied to one of the plates 218 and the change in induced voltage on the other plate 220 as the aperture passes between. the plates is sensed by a sensing circuit 226. Two pairs of plates spaced from one another about the axis of the disc may be used to eliminate spurious pulses resulting from axial movement of the disc. The sensing circuit may be arranged to generate a signal which is passed to a signal-processor, which in turn generates a number of output signals proportional to the number of signals received from the sensing circuit. The signal-processor can be a microprocessor, which may also supply the varying voltage applied to the plate(s) adjacent the disc. <IMAGE>

Description

SPECIFICATION Apparatus for measuring rotation The invention relates to apparatus for measuring rotation. The apparatus is especially intended for measuring the rotation of the rotor in a watthour meter and thus for determining the amount of electrical energy passing through the watthour meter. However, the instant apparatus may be used to generate pulses from any type of rotor and may thus be employed in a variety of other applications, for example, odometers.

Induction watthour meters are widely used for the measurement of power and energy comsumption. Such watthour meters incorporate a disc which rotates at a speed proportional to the power being consumed by the load connected to the meter. In order that the rotation of the disc may be measured and the total amount of electrical energy passing through the meter thus determined, it is often desired to generate signal pulses at a rate proportional to the speed of rotation of the disc. It is necessary that the apparatus used to generate such pulses not impose a mechanical load on the rotation of the disc, since the imposition of any such load will reduce the accuracy of the watthour meter.

Moreover, since commercial watthour meters are rugged, stable and can be left to operate unattended to long periods, it is essential that the pulse-generating apparatus also possess these characteristics. One common method of generating the desired pulses is to cut one or more holes or slots in the disc and to dispose an optical emitter and detector on opposed sides of the disc so that as the holes or slots pass between the emitter and detector, thereby allowing light to fall on the detector, the detector will generate the pulses. Such optical pulse-generating apparatus can be made rugged and reliable, but providing the optical emitter and detector with the necessary ruggedness and reliability is expensive and can substantially raise the cost of the watthour meter.

This invention seeks to provide a pulsegenerating apparatus which is rugged and reliable but which can be manufactured more cheaply than prior art optical pulse-generating apparatus.

According to this invention there is provided apparatus for measuring the rotation of an electrically-conductive rotor, said rotor having walls defining an aperture therethrough, said apparatus comprising: a pair of electrically-conductive plates comprising first and second plates disposed on opposed sides of said rotor such that as said rotor rotates said aperture will pass between said pair of plates; a signal generator for applying a varying voltage to said first plate; and sensing means for sensing the change in the voltage induced on said second plate as said aperture passes between said plates.

The term "aperture" used with regard to the opening in the rotor does not necessarily imply that the opening is completely surrounded by the rotor. Obviously, the "aperture" may have the form of a slot, notch or other cut-out extending to the periphery of the rotor and the term "aperture" is used herein to include such slots, notches and other cut-outs extending to the periphery of the rotor.

Because of restrictions on space within watthour meters, in most cases it will be found impracticable to mount the sensing means immediately adjacent the second disc, so that it will be necessary to provide a lead of some length extending from the second plate to the sensing means. Such a lead connected to the second plate acts as an extension of that plate and voltage is induced in the lead by the varying voltage applied to the first plate and the lead connected thereto.To prevent such induced voltages in the lead being sensed by the sensing means and thus producing spurious signals from the sensing means, it is desirable to connect a buffering means between the second plate and the sensing means, this buffering means presenting to the second plate an impedance substantially greater than the input impedance of the sensing means, thereby maximizing the input to the sensing means and reducing the effects of "noise" in the leads connecting the second plate to the sensing means.

Another problem which may be encountered with the instant apparatus is that, as the aperture in the rotor begins or ends its passage between the plates, erratic signals may be detected by the sensing means if the portion of the rotor passing between the plates moves parallel to the axis of the rotor, as may happen if the rotor is not accurately planar or is not accurately mounted in a plane perpendicular to its axis. To overcome this problem, the instant apparatus preferably comprises a second pair of electrically-conductive plates disposed on opposed sides of the rotor so that the aperture will pass therebetween, the first plate of the second pair having applied thereto the varying voltage and the second plate of the second pair being connected to the sensing means.The two pairs of plates are spaced from one another about the axis of rotation of the rotor so that the aperture passes between the two pairs of plates at different times. In this form of apparatus, the sensing means comprises a bistable device (such as a flip-flop) arranged to change from a first state to a second state when the induced voltage changes on the second plate of the first pair and from the second state to the first state when the induced voltage changes on the second plate of the second pair. The bi-stable device produces an output signal during at least one of these changes of state.It will be appreciated that in this apparatus having two pairs of plates, even if multiple changes in induced voltage are produced during the passage of the aperture between one of the pairs of plates, the extra changes will have no effect since whether one or a plurality of changes are generated the effect will only be to change the bistable device from one of its stable states to the other. In this type of apparatus, the buffering means (if present) preferably comprises separate buffering circuits connected between each of the second plates and the sensing means.

Although reference has been made above to two separate pairs of plates, it will be appreciated that the two first plates to which the varying voltage is applied can have the physical form of a single plate sufficiently large to extend across both of the second plates.

In order that a large change in induced voltage on the second plate(s) will occur as the aperture passes between the pair(s) of plates, the rotor must be effectively at ground potential so that when the aperture is not passing between the plates the conductive disc materiai between the two plates will shunt the induced signal to ground. In some applications, no difficulty will be experienced in electrically grounding the rotor, but the rotor in a watthour meter must be insulated from the shaft so that the rotor cannot be grounded directly. Where the rotor cannot be grounded directly, the frequency of the varying voltage applied to the first plate or plates should be sufficiently great that the capacitance between the rotor and the surrounding components is sufficiently large to make the rotor appear grounded.In commercial watthour meters, it is desirable to use an alternating voltage having a frequency of at least about 10,000, and preferably about 50,000, Hz.

Known optical pulse-generating apparatus can only generate an integral number of pulses for each revolution of the disc. In many applications, it is desired to generate a non-integral number of pulses per disc rotation, and in some cases only a fraction of a pulse per disc revolution. Thus, commercial watthour meters are known which require the generation of0.144,0.72, 1.08, 1.44, 10.368, 141.75 and other non-integrai numbers of pulses per disc revolution. Hitherto, to generate a nonintegral number of pulses per disc revolution, it has been necessary to drive a gear train mechanically from the disc, this gear train operating an electro-mechanical switch or driving another disc provided with apertures which permit the generation of pulses by the abovedescribed optical pulse-generating apparatus.Not only is such apparatus complex, but a different design is necessary for each desired ratio of pulses to disc revolutions (which increases the supplier's inventory and reduces possible economies of scale) and the resistance to rotation of the disc imposed by the gear train causes the accuracy of the watthour meter to be reduced.

In the apparatus of this invention the sensing means may be arranged to generate a signal each time the aperture passes between the plates, and the apparatus may include signal-processing means for receiving the signals from said sensing means and for generating output signals, the number of said output signals being equal to the number of signals received from said sensing means times a predetermined constant not equal to one. The signal-processing means may be capable of operating in at least two modes, the constant being different in each of these modes.

Conveniently, a single microprocessor may be employed as both the signal generator and the signal-processing means; the microprocessor can then correlate the responses received from the second plate or plates with the signals which it applies to the first plate or plates and can thus reject any spurious pulses produced by outside interference.

Apparatus in accordance with this invention will now be described, by way of example, with reference to the accompanying drawings, in which: Fig. 1 shows schematically a first embodiment of the invention having only a single pair of plates; Fig. 2 shows schematically a second embodiment of the invention having two pairs of plates; Fig. 3 shows the electrical circuitry of the apparatus shown in Fig. 2; Fig. 4 is a schematic diagram of a watthour meter incorporating a modified form of rotationmeasuring apparatus; Fig. 5 is a circuit diagram of the circuitry of the apparatus shown in Fig. 4; Figs. 6 and 7 are block diagrams showing the main program used in the microprocessor shown in Fig. 5; and Fig. 8 is a block diagram of the "Interrupt Routine" used by the microprocessor shown in Fig 5.

In Fig. 1 is shown an electrically-conductive disc 210 mounted for rotation about a shaft 212.

The disc 210 is insulated from the shaft 212 by an insulating portion 214 surrounding the shaft 212. The disc and shaft form part of a watthour meter, the remaining parts of which are omitted from the figure for the sake of clarity. The disc 210 has walls defining an aperture 216 therethrough, this aperture having the form of a sector of an annulus.

On opposed sides of the disc 210 are disposed a pair of electrically-conductive plates comprising a first or upper plate 218 and a second or lower plate 220. Each of the plates 218 and 220 has the same size and shape as the aperture 216 and the plates are arranged so that as the disc 210 rotates the aperture 21 6 will pas between the plates 21 8 and 220. The spacing of the plates 21 8 and 220 from the disc 210 is exaggerated in Fig. 1 for the sake of clarity.

A source 222 of alternating voltage is connected between the upper plate 21 8 and ground and thus a varying voltage is applied to the upper plate 218. The lower plate 220 is connected via a buffer 224 to a sensing circuit 226. The buffer 224 and the sensing circuit 226 may be of the type conventionally employed in connection with optical rotation-measuring apparatus, or may be as described below with reference to Fig. 3.

The alternating voltage applied to the upper plate 21 8 induces an alternating voltage in the lower plate 220. At the high alternating frequency used (about 50,000 Hz.) the capacitance between the disc 210 and the surrounding components of the watthour meter is sufficiently large to make the disc appear grounded. Accordingly, when the conductive material of the disc lies between the plates 218 and 220 the alternating voltage induced on the lower plate 220 is small. However, when the aperture 216 passes between the plates 218 and 220, the voltage induced on the lower plate 220 increases substantially and the voltage pulse thus produced is passed to the sensing circuit 226. Thereupon, the sensing circuit 226 produces an output pulse which is recorded by the watthour meter.The buffer 224 is arranged to present to the second plate an impedance substantially higher than the input impedance of the sensing circuit 226, thereby reducing the effects of "noise" induced in the lead connecting the second plate 220 to the sensing circuit 226.

Although the embodiment of the invention just described is capable of accurately measuring the rotation of the disc 21 0, problems arise if the disc is not completely flat or if it is not mounted perpendicular to its axis of rotation. In either of these circumstances, the position of the portion of the disc between the plates 21 8 and 220 moves parallel to the axis of the disc, and this axial movement may give rise to erratic action of the sensing circuit 226 and consequent spurious output pulses therefrom. To overcome this problem, the embodiment of the invention shown in Figs. 2 and 3 may be employed.

As shown in Fig. 2, this second embodiment of the invention again comprises an electricallyconductive disc 210 mounted for rotation about a shaft 212 from which it is insulated by an insulating portion 214. An aperture 216 having the form of a sector of an annulus passes through the disc 210. A first or upper plate 21 8 is disposed above the disc 210 and connected to a source 222 of alternating voltage. A second or lower plate 220 is disposed below the disc 210 adjacent the upper plate 218 and connected by a buffer 224 to a sensing circuit 226. The plates 218 and 220 are of the same size and shape as the aperture 21 6 and are disposed so that the aperture 21 6 will pass therebetween as the disc 210 rotates.

The aforementioned components of the second embodiment of the invention are identical to those of the first embodiment described with reference to Fig. 1. However, the second embodiment further comprises a second pair of plates comprising a first or upper plate 228 and a second or lower plate 230. The plates 228 and 230 are identical in size and shape to the plates 21 8 and 220 but are displaced 1 80C therefrom around the shaft 212. The upper plate 228 is connected to the voltage source 222, while the lower plate 230 is connected via a buffer 234 to a sensing circuit 236. The two buffers 224 and 232 together comprise the buffering means of the apparatus.

The outputs of the sensing circuits 226 and 236 are connected to the first and second inputs respectively of a flip-flop 238. (The two sensing circuits 226 and 236 and the flip-flop 238 together comprise the sensing means of the apparatus.) The flip-flop 238 is arranged so that an output pulse from the sensing circuit 226 will cause the flip-flop 238 to change from its first stable state to its second stable state, while an output pulse from the sensing circuit 236 will cause the flip-flop 238 to change back from its second stable state to its first stable state. The flip-flop 238 generates an output pulse as it changes from its first stable state to its second.

It will be seen that the apparatus shown in Fig.

2 will accurately record the number of revolutions of the disc 210 even if spurious pulses are generated as the aperture 216 passes between the two pairs of plates. Even if multiple pulses are generated during the passage of the aperture 21 6 between the plates 218 and 220, thus leading to a plurality of pulses from the sensing circuit 226, these plurality of pulses will only cause the flipflop 238 to change from its first to its second stable state. Similarly, even if a plurality of pulses are generated from the sensing circuit 236 as the aperture 21 6 passes between the second pair of plates 228 and 230, the flip-flop 238 will only change from its second to its first stable state.

Thus, the flip-flop 238 will generate only a single output pulse for each revolution of the disc 210, regardless of how many pulses may be generated as the aperture 216 passes between the two pairs of plates 218, 220 and 228, 230.

Fig. 3 shows the circuitry employed in the voltage source, buffers, sensing circuits and flipflop shown in Fig. 2. The voltage source 222 comprises a square wave generator comprising a NOR gate U 1", one input of which is connected to ground. The output of gate U1 " is connected to one input of a second NOR gate U2", and to one terminal of a resistor R2". The other input of NOR gate U2" is connected to ground and its output is connected via a capacitor C1" and a resistor R1" to the non-grounded input of gate Ul". The opposed terminal of resistor R2" is connected between capacitor Cl" and resistor R1".

The square wave output of gate U2" is fed to the upper plates 218 and 228. The lower plate 220 is connected to ground via a resistor R3" and is also connected to the gate of an N-type FET Ql". The source of Q1" is connected to ground via a resistor R4" and is also connected to one side of a capacitor C2". The FET Q1" and the resistors R3" and R4" together comprise the buffer 224.

The opposed side of capacitor C2" is connected to ground via a resistor R5" and is also connected to the positive input of a comparator U3". The negative input to comparator U3" is supplied from the mid-point of a voltage divider comprising resistors R6" and R7". The output of comparator U3" is connected to the positive supply line via a resistor R8" and is also connected via a diode Dl" to an RC circuit formed by capacitor C3" and resistor R9". The output from the RC circuit is fed to the positive input of a comparator U4". The negative input of comparator U4" is connected to the mid-point of a voltage divider comprising resistors Rl011 and R1 1".The output of comparator U4" is connected via a resistor R12" to the positive supply line and is also connected to one input of a NOR gate U5". The part of the circuitry from the resistor R5" to the resistor R12" comprises the sensing circuit 226.

The buffer 234 and the sensing circuit 236 are identical in construction to the buffer 224 and the sensing circuit 226 respectively and corresponding parts thereof are denoted by the same reference numerals with a third prime. The output from the sensing circuit 236 is fed to one input of a NOR gate U5"'; the NOR gates US" and U5"' are interconnected to form a flip-flop. The output of NOR gate U5"' is connected via a resistor R1 3" to the base of an NPN transistor Q2". The collector of transistor Q2" is connected to a positive supply line via a resistor R14" and an LED D2", while the emitter of transistor Q2" is grounded.

The circuitry shown in Fig. 3 operates as follows. As already mentioned, the square wave generator 222 applies a high frequency square wave to the upper plates 218 and 228. The signal from lower plate 220 is buffered by the buffer circuit comprising Q1", R3" and R4". When the pulses from the buffer 224 are greater than the reference voltage established by the voltage divider formed by resistors R6" and R7, a pulse wave form will be produced at the output of comparator U3". This pulse wave form is rectified by the diode D1 n and filtered by the RC circuit formed by capacitor C3" and resistor R9".The resultant rectified, filtered pulse is applied to the comparator U4", which produces a pulse at a level suitable to drive the NOR gate U58. The signal from the other lower plate 230 is processed in precisely the same manner and applied to the NOR gate U5"'. The pulses from the sensing circuit 226 delivered to U5n act as set pulses for the flip-flop 238, while those delivered from sensing circuit 236 to U5"' act as reset pulses for the flip-flop 238. The output of U5"' is a pulse which occurs once per disc revolution. The LED D2" produces a visual representation of the pulse.

The apparatus described above may be modified by substituting for the square wave generator 222 a microprocessor which supplies pulses to the upper plates 218 and 228. The same microprocessor can then sense the outputs from comparators U3" and U3"' directly, thus eliminating the diodes D1" and D1"' and all subsequent circuitry.

An example of apparatus modified in this way will now be described with reference to Figure 4.

In Fig. 4 is shown an electrically-conductive disc 10 mounted for rotation about a shaft 12.

The disc 10 is insulated from the shaft 12 by an insulating portion 14 surrounding the shaft 12.

The disc and shaft form part of a watthour meter, the remaining parts of which are omitted from the figure for the sake of simplicity. The disc 10 has walls defining an aperture 1 6 therethrough, this aperture having the form of a sector of an annulus.

Two pairs of plates 18, 20 and 22, 24 lie adjacent the disc 10, the two plates of each pair lying on opposed sides of the disc and being disposed so that as the disc 10 rotates the aperture 1 6 will pass between the two plates of each pair. The two pairs of plates 1 8, 20 and 22, 24 are spaced from one another 1800 about the axis of the shaft 12. Each of the plates 18, 20, 22 and 24 is of the same size and shape as the aperture 16 and lies in a plane parallel to the plane of the disc 10.

The first or upper plates 1 8 and 22 are connected via a line 26 to a microprocessor 28 which serves as both the signal generator and the signal-processing means of the apparatus. The microprocessor 28 applies to the upper plates 18 and 22 a pulsed wave alternating voltage having a pulse duration of about 70 microseconds and a repetition rate of about 1000 Hz. The lower plates 20 and 24 are connected via separate buffering circuits 30 and 32 respectively (which together form the buffering means of the apparatus) to comparators 34 and 36 respectively (which together form the sensing means of the apparatus). The comparators 34 and 36 are connected to separate inputs of the microprocessor 28.

Turning to Fig. 5, the apparatus is powered by a transformer T1 driven from a normal 120 V. 60 Hz. mains supply. The secondary coil of the transformer T1 is connected to a full-wave bridge comprising diodes D1, D2, D3 and D4. The outputs from this full-wave bridge are connected across a smoothing capacitor C1 and also to a commercially-available voltage regulator No. 78MO5, available from Fairchild Camera and Instrument Corporation, 464 Ellis Street, Mount View, California 94042. This voltage regulator produces a substantially ripplefree 5V supply.

The single microprocessor 28 acts as both the signal generator and the signal processing means of the apparatus. The microprocessor 28 is a Fairchild Type F3870 microprocessor, available from the aforementioned Fairchild Camera and Instrument Corporation. Further details concerning this microprocessor and the programming thereof are given in a pamphlet entitled "F3870 MicroMachine 2 Fairchild Microcomputer Family" published by Fairchild Camera and Instrument Corporation and dated June 1978.

To provide a clock for the operation of the microprocessor 28, the Time Base pins 1 and 2 thereof are connected to an external RC circuit, pin 1 being connected directly to ground, pin 2 being connected via a resistor R1 to the positive supply line and a capacitor C2 being connected between pins 1 and 2. Both the Ground Pin 20 and the Test Pin 21 are connected to ground, while the Power Supply Pin 40, the External Reset Pin 39 and the External Interrupt Pin 38 are connected in common to the positive supply line and are also connected in common to one plate of capacitor C3, the opposed plate of which is grounded.Since the External Reset and External Interrupt pins of the F3870 microprocessor 28 only operate when the inputs thereto go low, the connection of pins 39 and 38 to the positive supply line disables the external interrupt and external reset functions of the microprocessor.

The lower plate 20 is connected to ground via a resistor R2 and is also connected to the gate of an N-type FET 01. The drain of Q1 is connected to the positive supply line, while its source is connected to ground via a resistor R3 and is also connected to one plate of a capacitor C4. The opposite plate of capacitor C4 is grounded via a resistor R4 and is also connected to the positive input of a comparator U1. The negative input of U1 is supplied from the midpoint of a voltage divider comprising resistors R5 and R6 and a capacitor C5 connected in parallel with R6. The output of comparator U 1 is fed to pin 25 of the microprocessor 28.

The lower plate 24 is connected to pin 24 of the microprocessor 28 via exactly the same circuitry as that between the lower plate 20 and the pin 25 of microprocessor 28, the various components in this circuitry being distinguished from those in the circuitry associated with the plate 28 by primes.

Pins 26-33 of microprocessor 28 are grounded via severable links. The microprocessor 28 incorporates a read-only memory (ROM) having stored therein 256 constants. These constants are selected by severing one or more of the links grounding pins 26-33. For example, if all the links are left intact, the microprocessor 28 operates on a constant of 1.44, this being the constant most popularly employed in commercial watthour meters. The constants are arranged in the microprocessor ROM in such a way that the more popular a constant is in commercial watthour meters, the fewer links need to be severed to bring that constant into operation.

The output of microprocessor 28 is provided on pins 22 and 23. Pin 23 is connected to the base of PNP transistor Q2. The emitter of Q2 is connected via an LED D5 and a resistor R7 to the positive supply line, while the collector of Q2 is grounded. The LED D5 forms part of an optical isolator also including an NPN phototransistor Q3.

The emitter of Q3 is connected to the base of a further NPN transistor 04. The collectors of Q3 and Q4 are connected in common to one vertex of a diode bridge formed by diodes D6, D7, D8 and D9, while the emitter of Q4 is connected to the opposed vertex of the bridge. Lines Y and K are connected to the remaining vertices of the bridge.

Preciseiy similar circuitry is associated with the pin 22 of the microprocessor 28; the components of the circuitry associated with pin 22 are designated by the same reference numerals as those associated with the pin 23, but with the addition of primes. Note that resistor R7 is connected to the positive sides of both diodes. D5 and D5' and that line K is connected to both the diode bridges.

The operation of the circuitry shown in Fig. 5 is as follows. As already mentioned, the microprocessor 28 acts as a signal generator and applies a high frequency pulsed voltage to the upper plates 1 8 and 22 via the line 26. The FET 01, in conjunction with the resistors R2 and R3, act as a buffer circuit and presents a high impedance to the plate 20. This high impedance maximizes the voltage pulse obtained from the plate 20 as the aperture passes between the plates 1 8 and 20, thereby reducing the effects of "noise" induced in the line from the plates 20 to the FET Q1. The output from Q1 passing through the capacitor C4 then feeds the relatively low input impedance of the comparator U1, determined by the resistor R4.The comparator U 1, together with its companion comparator U1', acts as the sensing means of the apparatus.

When the voltage induced on the positive input of the comparator U1 is greater than the reference voltage established by the voltage divider R5, R6 and C5, U1 sends an output to pin 25 of the microprocessor 28. Thus, each time the aperture 1 6 in the disc 10 passes between the plates 18 and 20, a pulse is received at pin 25 of microprocessor 28. Similarly, whenever the aperture 16 passes between the plates 22 and 24, a pulse is received at pin 24 of the microprocessor 28. Thus, as the aperture passes between the plates 18 and 20, pin 25 is at logic 1 and pin 24 at logic 0, whereas when the aperture passes between plates 22 and 24 pin 25 is at logic 0 and pin 24 at logic 1.Should "noise" within the watthour meter cause both pins 24 and 25 to go to logic 1 simultaneously, the microprocessor 28 will reject the resultant pulses as spurious, i.e. the microprocessor 28 only acknowledges as genuine those pulses in which one of the pins 24 and 25 is at logic 1 and the other is at logic 0. When pin 25 is at logic 1 and pin 24 at logic 0, a bi-stable device incorporated within the microprocessor 28 changes from a first stable state to a second stable state, whereas when pin 24 is at logic 1 and pin 25 at logic 0 the bi-stable device changes from its second stable state back to its first stable state. Thus, the bi-stable device will undergo two changes of state for each revolution of the disc 10.

The output pins 22 and 23 are always held at opposite logic levels and an output pulse from the microprocessor is signalled by changing simultaneously the logic state of each pin. When pin 23 is at logic 0, transistor 02 allows current to flow through photodiode D5. This renders phototransistor Q3 conductive, which in turn renders transistor Q4 conductive, thereby interconnecting lines Y and K. Since pin 22 is simultaneously at a logic high, transistor Q2' does not conduct and line Z is isolated from the line K.

On the other hand, when pin 23 is at a logic high and pin 22 at a logic low, lines K and Z will be interconnected but line Y will not be connected to line K. Thus, the solid state AC/DC switches formed by Q4, D6, D7, D8, D9 and Q4', D6', D7', D8', D9' simulate a conventional three-wire (form C) relay contact and the output on lines K, Y and Z can be fed to a conventional recording device.

A block diagram of the software used in the microprocessor 28 is shown in Figs. 6, 7 and 8.

As shown in Fig. 6, upon commencing operation, the program begins with a Reset block 40 which clears the registers to be used in the program, and then clears a Remaining Integer Pulses register at block 42. The clearing of the Remaining Integer Pulses register ensures that any pulses stored in this register from previous operation do not give rise to spurious output pulses at the beginning of the new operation.

After the clear remaining integer pulses block, the program proceeds to a Disable Interrupt block 44. As explained in more detail below, the microprocessor 28 generates internally interrupt pulses at a rate of about per millisecond which cause the microprocessor to perform the interrupt routine shown in Fig. 8. Because of the limited amount of memory in the microprocessor (the F3870 microprocessor has only 64 bytes of scratch-pad random access memory) some of the random access memory is utilized in both the main program and the interrupt routine.

Accordingly, at various points during the main program it is necessary to disable the input from the internal interrupt function prior to carrying out arithmetical processing in the main program in order to ensure that this arithmetical processing is not interrupted by an interrupt routine, which would cause overwriting of certain registers being used in the arithmetical processing, so giving rise to spurious values stored in certain registers and consequent errors when the microprocessor returns to the main program after the interrupt routine is finished. At block 44, the program stops the operation of the timer which generates the interrupt pulses.

After disabling the interrupt, the program proceeds, at block 46, to add a Pulses Per Edge register to a Pulse Total register. Since both these registers have been zeroed during the reset routine, at this initial stage of the program the sum in the Pulse Total register will still be zero after the addition. Thus, when the program proceeds to the next decision block 48 and tests the sum in the Pulse Total register to see if it is greater than or equal to lithe program will proceed immediately via the "NO" output from block 48 to block 54, which begins an edge routine in which the microprocessor tests for an "edge", a term which is used herein to refer to the complete passage of the aperture 1 6 between one of the pairs of plates 18, 20, 22, and 24.

At block 54, the program writes the integer 5 into a Verification Counter register. The program then proceeds to a Disable Interrupt block 56 which prevents the microprocessor from acting on an interrupt while the succeeding block is being carried out (the period which the interrupt function will not be recognized is so brief-a few microseconds-that the recognition of the interrupt signal will merely be delayed by a short interval-there is no danger of the microprocessor failing to respond to that interrupt pulse). Although at this initial stage of the program the interrupt function will already be disabled because the program has just passed through the block 44, the block 56 is necessary because in later cycles the interrupt function may still be enabled when the program reaches block 56.

After disabling the interrupt function of block 56, the program proceeds, at block 58, to apply a voltage pulse to the upper plates 1 8 and 22 and to test the inputs received on pins 24 and 25 to see if an edge has been detected. The interrupt function is then re-enabled at block 62, thence the program proceeds to a decision block 62 where a decision is made as to whether the inputs on lines 24 and 25 indicate an edge. As previously mentioned, for valid detection of an edge a signal must be received on one of pins 24 and 25 but not on the other; if inputs are received on both pins 24 and 25 simultaneously, the microprocessor rejects the input as spurious.

Furthermore, because of the arrangement using a single aperture and two pairs of plates, inputs should be received alternately from each of the lower plates 20 and 24. Accordingly, if the previous valid input was received on pin 24, the microprocessor will only accept as valid an input on pin 25, and vice versa.

If no valid input indicating an edge is detected, the program loops back from decision block 62 to block 54. Thus, until a valid edge signal is detected, the program will merely cycle through the blocks 54 and 62 inclusive. When a valid edge input is received, the program exits decision block 62 via the "YES" output thereof and enters a further decision block 64 in which the validity of the output state of the microprocessor is checked.

As previously mentioned, the output from the microprocessor occurs on pins 22 and 23 which should always be of opposite polarity. If at block 64 it is determined that pins 22 and 23 are at the same polarity, (the NO output of block 64), the program proceeds to an initialize output block 66 which ensures that the pins and 22 and 23 are made of opposite polarity, and thence loops back to the block 54. On the other hand, if at block 64 it is determined that the output state on pins 22 and 23 is valid, the program proceeds to block 68 where the Verification Counter register is decreased by 1. The Verification Counter register is then tested at block 70 to determine whether it is equal to zero. If the Verification Counter register is not equal to zero, the program loops back from block 70 to block 56.Thus, a valid edge response must be detected on five successive loops through the blocks 56-68 before the program will proceed beyond block 70. This redundancy in the program greatly reduces the possibility that a spurious input on one of the pins 24 and 25 (caused, for example, by a sudden spike in the noise affecting the lines connecting the lower plates 20 and 24 to the respective FET's Q1 and Q1 ') will give rise to spurious outputs from the microprocessor 28. Obviously, a sudden fluctuation in the noise level might cause a single valid input to the pins 24 and 25 but such a random event is hardly likely to generate five successive valid inputs and would thus not allow the program to proceed beyond the decision block 70.

When a decision is reached at block 70 that the Verification Counter register is equal to zero, thus showing that five successive loops of the program have indicated a valid edge detection, the program proceeds via a Disable Interrupt block 72, which is identical to block 56, to an Input Ratio Code block 74. At block 74, the microprocessor 28 tests the continuity of each of the lines attached to pins 26-33. The continuity or non-continuity of each line is used to generate a single digit of a binary number so that after the continuity of all the lines attached to pins 26-33 has been checked, the microprocessor 28 has derived an 8-digit binary number presenting the address of one of the 256 locations in the ROM where the various possible ratios are stored.The program then proceeds to a block 76 where the 8-digit binary number thus produced is used to obtain from the ROM the pulse ratio previously set by the operator. The pulse ratio emerges from the ROM in the form of the 6-(decimal) digit number expressed in floating-point form, the most significant digit of the 6-digit number representing an order of magnitude. In the particular program shown in Fig. 7, this most significant digit can only be 0,1 or 2 so that the pulse ratios to be used must lie in the range 0.1 to 99.999.

The magnitude digit of the pulse ratio derived at block 76 is then tested, at block 78, to see whether it is equal to 0. If the magnitude digit is equal to zero (thus indicating that the pulse ratio is less than one) the program exits block 78 via output A and loops back to block 44. After disabling the interrupt routine at block 44, the program then adds the pulse ratio, which is held in a Pulses Per Edge register into a Pulse Total register at block 46. Since the Pulse Total register has hitherto been 0, after the summation at block 46 the Pulse Total register will still have a value less than 1 and the program will exit at block 48 along the NO output as previously described and await the verification of a further pulse. However, on later cycles of the program, the total in the Pulse Total register will exceed one.In such later cycles, when the program reaches block 48, and the total in the Pulse Total register is greater than 1, the program will leave block 48 via the YES output and enter block 50 where the Pulse Total register is decreased by 1 and then enters a Toggle Outputs block 52 where an output pulse is indicated on the pins 22 and 23. From the block 52, the program returns to the previously described part of the program at block 54. Thus, the blocks 44 to 52 form a "fraction routine" which serves to generate the necessary output pulses when the program is operating with a pulse ratio less than unity. Also, because of the presence of the block 44 in the fractional routine, while the program is operating with a pulse ratio less than unity, no interrupt pulses will be received from the timer and thus the interrupt routine shown in Fig. 8 will not be activated.

Thus, when the microprocessor is operating with a pulse ratio less than unity, the program simply adds the relevant fractional pulse ratio to the Pulse Total register after each edge has been detected and each time the Pulse Total register equals or exceeds unity, the Pulse Total register is reduced by 1 and an output pulse is generated.

Accordingly, the output from the microprocessor under these circumstances comprises a substantially equally spaced series of pulses, the intervals differing between pulses differing by no more than the period between the two edges. It should also be noted that, when the microprocessor is operating with a fractional pulse ratio, the program does not proceed beyond block 78 and no pulses are outputted on the interrupt routine, as will be described in more detail below.

On the other hand, if at block 78 the program finds that the magnitude digit is not 0, the program proceeds to block 80 where it adjusts the pulse ratio for the magnitude digit 1 by moving the pulse ratio one decimal place to the left. The program then proceeds, at block 82, to test the magnitude digit to see if it is equal to one.

If the magnitude digit is equal to 1, the program proceeds directly to block 88 to be described below. If, however, the magnitude digit is not equal to one, the program proceeds from block 82 to a further adjustment block 84 similar to block 80, where the pulse ratio is shifted one further decimal place to the left. The magnitude digit is then tested, at block 86, to confirm that it is equal to 2; if the magnitude digit is not equal to 2, there has been an error in the reading and the program returns to the reset block 40. If, however, no error has occurred, the program proceeds from block 86 to block 88.

Although in the specific program illustrated in Fig. 7, the magnitude digit can only be 0,1 or 2, it will be appreciated that the range of possible pulse ratios can easily be extended by allowing the magnitude digit to vary over a wider range, the only adjustment to the program needed being insertion of further blocks similar to blocks 78,82 and 86 for testing the magnitude of the magnitude digit and the insertion of further adjustment blocks similar to blocks 80 and 82.

As already mentioned, when the pulse ratio received at block 76 is fractional, the fraction is merely added to a single Pulse Total register at block 46. However, for reasons which will appear below, when the pulse ratio received at block 76 is unity or greater, the integral and fractional parts of the pulse ratio are handled separately.After the necessary adjustment of the pulse ratio has been effected at block 78-86, the fractional part of the pulse ratio, which is stored in a "Decimal" register is added to a "Remaining Decimal Pulses" register at block 88, and then the integral part of the pulse ratio, which has been stored in a "Integer" register is added to a "Remaining Integer Pulses register at block 90. (Naturally, the program incorporates a carry-over arrangement so that if the total in the Remaining Decimal Pulses register at block 88 becomes greater than unity, that total is reduced by 1 and the Remaining Integer Pulses register is increased by one).

From block 90 the program proceeds to block 92, where the total just generated in the Remaining Integer Pulses register is copied into both a Remaining Integer register and a Remaining'lnteger Decrementer register. Next, at block 94, a Number Of Groups Per Toggle Incrementer (NOGPTI) register is copied into both a Previous Number Of Groups (PNOG) register and into a Number Of Groups Per Toggle Decrementer (NOGPTD) register. The NOGPTI register is then cleared at block 96 and then the block 98 the timer is initialized and interrupt function is re-enabled. The program then loops back, as shown by connection B, to box 54 and begins the edge routine again, waiting for the next valid edge detection.

The blocks 80-98 constitute a multiple routine which operates only when the program is operating with a pulse ratio not less than one. It will be seen that the multiple routine itself does not cause output pulses to be sent out from the microprocessor but only accumulates in the registers employed various data related to the number of integral pulses which are required to be sent out but have not yet been sent out. After each passage through the multiple routine, this total number of pulses to be sent out is present in the Remaining Integer Pulses register, the Remaining Integer register and the Remaining Integer Decrementer register, while the NOGPTI register is clear. The actual outputting of the pulses, when the pulse ratio is not less than unity, is effected by the Interrupt Routine shown in Fig.

8.

As previously mentioned, the microprocessor 28 generates internally interrupt pulses at a frequency of approximately 1 per millisecond.

When one of these interrupt pulses is received during a part of the program where the interrupt function is enabled, the progress of the main program shown in Figs. 6 and 7 is interrupted and the microprocessor carries out the interrupt routine shown in Fig. 8. In Fig. 8, the receipt of an interrupt pulse is shown at block 100. Upon receipt of the interrupt pulse, the program proceeds to save the contents of a Status register and the Accumulator of the microprocessor. The F3870 microprocessor does not automatically store the contents of the Status register and the Accumulator and thus the insertion of block 102 is necessary to insure that the contents of the Status register and the Accumulator are temporarily stored so that the correct values are present in the Status register and Accumulator to allow recommencement of the main program when the interrupt routine is finished.

From block 102, the program proceeds to block 104 and decrements the Remaining Integer Decrementer register by 1. The program then tests the Remaining Integer Decrementer register at block 106 to see if it is equal to 0. If the Remaining Integer Decrementer register is equal to 0, the program increments the NOGPTI register at block 108 and then, at block 110 loads the contents of the Remaining Integer Register into the Remaining Integer Decrementer Register.

From block 110 the program proceeds to block 112. If at block 106 the remaining Integer Decrementer register is not equal to 0, the program proceeds from block 106 directly to block 112.

At blocks 112 and 114, the NOGPTD register is checked to see if it is equal to 0. If the-NOGPTD register is equal to 0, the program proceeds to block 11 6 and loads the contents of the PNOG register into the NOGPTD registers Then, at blocks 11 8 and 120 the Remaining Integer Pulses register is checked to see if it is equal to 0. If this register is not 0, the program decrements the Remaining Integer Pulses register by one at block 122 and then toggles the outputs of the microprocessor on pins 22 and 23, at block 124, thereby outputting a pulse. The program then proceeds to block 126, where the contents of the Status register and the Accumulator, which were temporarily stored at block 102, are restored to their former locations.

If at block 120 it is found that the Remaining Integer Pulses register is equal to 0, the program proceeds from block 120 directly to block 126.

Also, if a block 114 the NOGPTD register is found to be equal to 0, the program proceeds from block 114 to block 128 where the NOGPTD register is decremented by 1, and then directly to block 126.

After restoring the original values of the Status register and Accumulator at block 26, the interrupt routine re-enables the interrupt function at block 130 and then at block 132 returns to the main program at the point from which it left the main program.

The foregoing program enables the microprocessor to space its output pulses substantially uniformly even when a large number of pulses have to be generated for each edge.

Basically, for each cycle, defined by the period between successive edges, the microprocessor notes the number of pulses to be sent out (in the Remaining Integer Pulses register) and also determines the duration of that cycle by counting the number of interrupt functions received during the cycle. At the end of the cycle, the number of interrupt pulses are divided (by repeated subtraction) by the integral number of pulses to be sent out (from the Remaining Integer pulses register); the quotient is placed in the NOGPTI register so that during the next cycle of the device after each of the appropriate number of interrupts one output pulse will be sent out. In other words, on each cycle of the program, the pulses being sent out are the appropriate number of pulses for the preceding cycle.Obviously, if the number of pulses per edge is non-integral, the integral number of outward pulses generated on each cycle will vary above and below the non-integral number. For example, if the pulse ratio is 8.25, on some cycles 8 output pulses will be generated and on others 9, but the flow of pulses will not vary by more than that caused by variations of one pulse per cycle.

It will be obvious to those skilled in the art that the apparatus described may be modified in several ways. For example, more than one aperture may be provided in the disc in order to increase the number of pulses detected by the comparators U1 and U1'. Note, however, that if more than one aperture is employed, the apertures should not be spaced at 1 800 from each other since there must not be apertures between both pairs of plates simultaneously because that would cause inputs to appear at pins 24 and 25 of the microprocessor simultaneously, whereupon the microprocessor would reject the inputs as spurious. In practice, we have found it convenient to operate with three apertures spaced at 1200 intervals from one another. With three apertures and two pairs of plates, six edges are generated for each revolution of the disc.

The apparatus described above may be modified by the use of only a single pair of plates, the plates 22 and 24 and the circuitry associated with plate 24 being eliminated. However, such an apparatus is more susceptible to error due to noise within the watthour meter and spurious pulses due to axial movement of the disc between the plates.

Claims (23)

Claims

1. Apparatus for measuring the rotation of an eiectrically-conductive rotor, said rotor having walls defining an aperture therethrough, said apparatus comprising: a pair of electrically-conductive plates comprising first and second plates disposed on opposed sides of said rotor such that as said rotor rotates said aperture will pass between said pair of plates; a signal generator for applying a varying voltage to said first plate; and sensing means for sensing the change in the voltage induced on said second plate as said aperture passes between said plates.

2. Apparatus according to claim 1 further comprising a second pair of electricallyconductive plates comprising first and second plates disposed on opposed sides of said rotor such that as said rotor rotates said aperture will pass between said second pair of plates, said second pair of plates being spaced from said first pair of plates about the axis of rotation of said rotor, said first plate of said second pair being connected to said signal generator and said second plate of said second pair being connected to said sensing means, and wherein said sensing means comprises a bi-stable device arranged to change from a first state to a second state when said induced voltage changes on said second plate of said first pair, and from said second state to said first state when said induced voltage changes on said second plate of said second pair, said bi-stable device producing an output signal during at least one of said changes of state.

3. Apparatus according to claim 2 wherein said bi-stable device comprises a flip-flop.

4. Apparatus according to claim 2 wherein said signal generator applies to said first plates an alternating voltage having a frequency of at least about 10,000 Hz.

5. Apparatus according to claim 2 wherein a buffering means is connected between each of said second plates and said sensing means, said buffering means presenting to said second plates an impedance substantially greater than the input impedance of said sensing means.

6. Apparatus according to claim 1 or 2 wherein said signal generator supplies a square wave voltage.

7. Apparatus according to claim 1, wherein said sensing means generates a signal each time said aperture passes between said plates; and including signal-processing means for receiving the signals from said sensing means and for generating output signals, the number of said output signals being equal to the number of signals received from said sensing means times a predetermined constant not equal to one.

8. Apparatus according to claim 7 wherein said signal-processing means is capable of operating in at least two modes, said constant being different in each of said modes.

9. Apparatus according to claim 7 further comprising a second pair of electricallyconductive plates disposed on opposed sides of said rotor such that as said rotor rotates said aperture will pass between said second pair of plates, said second pair of plates being spaced from said first pair of plates about the axis of rotation of said rotor, the first plate of said second pair being connected to said signal generator and the second plate of said second pair being connected via said sensing means to said signalprocessing means.

10. Apparatus according to claim 9 wherein said signal generator applies to said first plates pulsed voltage having a pulse width of not more than about 100 microseconds.

11. Apparatus according to claim 9 wherein said signal generator applies a square wave pulsed voltage to said first plates.

1 2. Apparatus according to claim 9 wherein a buffering means is connected between each of said second plates and said sensing means, said buffering means presenting to said second plates an impedance substantially greater than the input impedance of said sensing means.

1 3. Apparatus according to claim 8 wherein said signal-processing means comprises a readonly memory having said constants stored therein.

1 4. Apparatus according to claim 8 or 1 2 wherein the mode in which said signal-processing means operates is determined by the electrical continuity of a plurality of conductors associated with said signal-processing means.

15. Apparatus according to claim 7 or 8 wherein said output signals from said signal processing means are passed through an optical isolator.

1 6. Apparatus according to claim 8 wherein in at least one of said modes said constant is greater than unity, and wherein, when said apparatus is operating with a constant greater than unity, said signal-processing means is arranged to space the output pulses substantially uniformly.

1 7. Apparatus according to claim 7 wherein said signal processing means sums the number of output pulses to be sent out during an interval and measures the duration of said interval, and then sends out the said number of output pulses evenly spaced during the succeeding interval.

1 8. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figure 1 of the accompanying drawings.

1 9. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figure 2 of the accompanying drawings.

20. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figures 2 and 3 of the accompanying drawings.

21. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figure 4 of the accompanying drawings.

22. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figures 4 and 5 of the accompanying drawings.

23. Apparatus for measuring the rotation of an electrically-conductive rotor, substantially as hereinbefore described with reference to Figures 4 to 8 of the accompanying drawings.