GB2423363A

GB2423363A - A flow meter

Info

Publication number: GB2423363A
Application number: GB0603269A
Authority: GB
Inventors: Michael Langley Sanderson
Original assignee: Cranfield University
Current assignee: Cranfield University
Priority date: 2005-02-18
Filing date: 2006-02-20
Publication date: 2006-08-23
Anticipated expiration: 2026-02-20
Also published as: GB0503422D0; GB0603269D0; GB2423363B

Abstract

An acoustic flow meter has two transducers spaced along a fluid flow path which emit ultrasonic signals in opposite directions. Each signal is detected by the other transducer than the one it was sent from. The transmitted signals have exactly equal amplitudes and are either in phase with each other or exactly out of phase with each other. The signals travel at different speeds in the upstream and down stream directions, so the received signals are no longer in a pre-determined phase relationship. In the case of in-phase signals, one of the received signals is subtracted from the other to form a resulting new signal. In the case of anti-phase signals (as in figure), the received signals are added together to make a new signal. The amplitude of the new signal is related to the flow rate due to constructive or deconstructive interference when the received signals are combined. The flow rate can be calculated from the amplitude of this signal, therefore allowing simpler circuitry to be used than in conventional flow meters.

Description

A FLOWMETER

The present invention relates to the field of flowmeters and in particular to acoustic and ultrasonic flowmeters of the transit-time variety.

In an acoustic flowmeter of the transit-time variety, a sound wave is made to pass through the fluid flowing in a channel and the velocity of the fluid flow and hence the fluid flowrate is determined on the basis of the difference between the rate of propagation of the sound wave in the upstream and downstream directions of fluid flow. Thus the flowrate of the fluid through the channel is determined by the difference between the two propagation times. The frequency of the sound waves used in an acoustic flowmeter may be (and usually is) an ultrasonic frequency. The acoustic flowmeter is then commonly referred to as an ultrasonic flowmeter. The ultrasonic type of acoustic flowmeter is accepted broadly in the art.

Figure 1 shows the operation of a diametrical flowmeter in which the cross section of the channel is circular. An ultrasonic wave is transmitted from transducer 1 and received by transducer 2 to measure the transit time upstream. A second ultrasonic wave is transmitted from transducer 2 and received by transducer 1 to measure the transit time downstream.

The transit time difference between the downstream and upstream waves is given by: (1) where iT is the difference in the transit times; T2 is the propagation from transducer 1 to transducer 2; T21 is the propagation from transducer 2 to transducer 1; v is the average axial velocity measured along the beam; 8 is the angle between the direction of propagation and the pipe axis; 1 is the length of the path over which the integration is made and c is the speed of sound in the fluid. Thus: 2AT v=_C (2) 2.cosO./ The average velocity measured along the beam is then used to compute the average velocity across the cross section of the channel and hence the flowrate of the fluid through the channel.

Diametrical ultrasonic flowmeters employing a single beam are known to be sensitive to velocity profile changes which occur as a consequence of the change in Reynolds Number in fully developed flow or by upstream flow disturbances caused by pipe work such as bends or reducers and valves. Multi-chordal or multi-path flowmeters use the transit time measurements over one or more paths in the cross section of the flowmeter to improve the velocity profile averaging across the whole of the of the cross section of the flow and thus reduce the effects of varying velocity profiles.

At low flowrates, particularly in small diameter tubes, the transit time differences become difficult to measure. For such applications, multiple reflections or axial designs are commonly used. In the reflecting methods as shown in figure 2 the ultrasonic beam transits the flow tube several times (typically 2 or 4 times) by reflecting the beam off the walls of the flowtube and thus increasing the transit times and the transit time differences. The number of reflections is usually limited to either 1 or 3 since the width of the beam may lead to interference between reflections.

Axial designs as shown in figure 3 are used to decouple the length over which the transit time difference is measured from the diameter of the pipe. However, even for such flowmeters the time measurement is difficult. For an axial design with a flow velocity of 0.lm/s, the separation of the transducers is 0.lm, the transit time in water is 66EJs and the transit time difference is 8.8ns and for 1% accuracy it is necessary to have a method capable of resolving time to 88ps. This can only be eased by increasing the axial length over which the transit time difference is measured and this makes the flowmeter impractical.

The methods known in the art for measurement of the transit time difference include pulse methods in which the two waves are transmitted sequentially or simultaneously. In the case of sequential measurements the propagation times T12 and T21 are measured sequentially using a digital clock as shown in figure 4. In simultaneous measurements both transducers are used as transmitters and then both as receivers. A pulse whose width is equal to the differences in the propagation times is generated. This is then measured digitally. For small time differences multiple measurements and averaging are generally required which increases the flowmeter response time.

In methods which employ a burst of a single frequency, the time difference is converted to a phase difference between the two receive signals as shown in figure 5. The phase difference in degrees between the two receive signals is thus:- = LT.360.f (3) where f is the frequency of the single frequency. For a flowmeter operating at a frequency of 1MHz, having an axial length of 0. lm the phase difference is approximately 30 at a flow velocity of 0.lm/s and has to be resolved to 0.03 for an accuracy of 1%. The measurement of zero phase difference is also difficult.

Compensation in all of the above methods is also required for changes in the speed of sound in the fluid as shown in equation 2. In both time and phase difference methods these are made by measuring the transit times of the waves T21 and T12. Methods for compensating for the speed of sound are known in the art.

An object of the present invention is to provide a new method which obviates the need for high-speed time measurement circuitry. It is applicable to diametrical, multi-chordal/multjpath, reflecting and axial flowmeters. It can be used in pulse or single frequency and sequential or simultaneous operation. The approach enables the measurement to be made with shorter path lengths between the two transducers and/or at lower velocities.

According to a first aspect the invention provides a flow meter comprising a first transducer and a second transducer, each located at a spaced locations along a fluid flow path; a drive circuit which applies a first drive signal to the first transducer to cause the first transducer to emit radiation and a second drive signal to the second transducer to cause the transducer to emit radiation; a detecting circuit adapted to detect a first output signal produced by the first transducer which is at least partly due to receipt of the radiation from the second transducer and a second output signal produced by the second transducer which is at least partially due to receipt of radiation from the first transducer; and a processing means adapted to process the first and second output signals to generate a flow rate signal indicative of the velocity of flow of fluid along the fluid flow path; characterised in that: the processing means combines the first and second output signals to produce a time of flight signal whose amplitude is dependent upon the difference in time of flight between radiation transmitted from the first transducer to the second transducer and radiation transmitted from the second transducer to the first transducer, the flow rate being derived from the time of flight signal.

By producing a signal whose amplitude depends on the time of flight, it becomes possible to determine the flow rate by simply measuring the amplitude of the signal as opposed to making actual time of flight measurements as before. This allows a simpler circuit to be provided and may also lead to increases in performance.

The apparatus may drive the first and second transducers simultaneously or in sequence. In the first instance, the detection of the two output signals may also be made simultaneously.

The drive signals may comprise identical signals, for example sinusoids of the same frequency and amplitude. The processing means may then determine the amplitude signal by subtracting one output signal from the other. At zero flow rate this will, ideally, produce a zero output.

Alternatively, the drive signals may be of the sinusoids of the same amplitude but exactly in anti-phase. The processing means may then add the two output signals together to produce the time of flight signal.

It is important that the two drive signals are identical. The apparatus may therefore use one signal generator to produce both drive signals.

Of course, the apparatus need not produce sinusoidal drive signals. They could be of some other waverform, or could even be pulsed signals.

The apparatus may include means for determining the amplitude of at least one of the drive signals. The processing means may then use this to calibrate the time of flight signal. For example, the greater the magnitude of the drive signals, the greater the amplitude of the time of flight signal for a given non-zero flow rate.

In the case of an apparatus which drives both transducers at the same time, the signals received at transducers may include unwanted noise caused by the transducer reacting to a reflection of its own drive signal.

These non-reciprocal effects lower the accuracy of the time of flight signal.

The apparatus may therefore be adapted to transmit at least two further signals. A first one of the signals may be produced by applying a calibration signal to the first transducer equal to the first drive signal at a time when no drive signal is applied to the second transducer. The apparatus may then be adapted to detect any output signal from the first transducer. This signal will represent the unwanted noise generated by the first transducer.

The same may then be performed for the second transducer, e.g. applying a second calibration signal to the second transducer and detecting any output signal from that transducer.

The apparatus may then combine the output signals that result from the calibration signals with the time of flight signal and any calibration signal to produce a more accurate time of flight signal which is free from non- reciprocal effects.

The apparatus may include a transformer having two bifilar windings, one terminal of each winding being joined to the other in such a way as to signals which are in anti-phase, the output of the two windings providing the drive signals for the first and second transducers.

According to a second aspect the invention provides a method of measuring flow rate in a fluid employing at least two ultrasonic transducers in which the transit time difference is measured between ultrasonic beams transmitted in opposite directions between the two transducers.

The time difference may be measured by simultaneously driving both transducers and measuring the difference in the transit times.

The method may comprise measuring the time difference by transmitting a burst of a single frequency or a continuous signal, such as a sinusoidal waveform.

The signals may be transmitted and received reciprocally, i.e. from the first to the second transducer and then the second to the first and so on.

Or they may be transmitted simultaneously.

It is preferred that the method comprises transmitting signals of the same amplitude and phase from both transducers. The method may then combine the measure signals from both transducers by subtracting one from the other so that the difference between the two receive signals is generated. Alternatively, the signals may be transmitted in anti-phase and combined by subtracting one from the other.

Where they are sent alternately, the method may comprise creating a measure of the receive amplitudes for each of the transducers in turn when Only one transducer is being driven and the receive signal is detected from only the transducer being driven.

The method may also, in a refinement includes steps for using the estimates of the amplitudes of the individual receive signals to remove the non-reciprocal elements of the difference of the two receive signals and create an estimate of the difference in two reciprocal receive signals.

The method may be adapted to use the ratio of the estimate of difference in the reciprocal receive signals to the estimates of the individual receive signals to estimate the phase difference between the two signals.

It may comprise using the phase difference between the receive signals to estimate the transit time difference and hence the flow velocity and flowrate.

The method relies, in at least one arrangement, on driving both transducers with identical signals (or identical other than being in antiphase). The method may generate the drive signals from the output of two bifilar windings of a toroidal transformer, one terminal of each winding being joined to the other in such a way as to signals which are in antiphase The method of the invention may therefore include 4 phases: a drive phase, a calibration phase and two compensation phases. In the drive phase a signal is sent from one transducer to the other and the output of the second transducer measured, and also from the second to the first and the output of the first measured. These two signals are then combined to produce a time of flight signal dependent upon the difference in the time of flight of the two signals. In the calibration phase, the amplitude of at least one of the drive signals is determined by measuring the output of one of the transducers whilst only the other one is driven. In the calibration phases each of the transducers is driven whilst the other is not driven and the output from the driven transducer is measured. The signals from each of the four phases are then combined to determine the time of flight and hence flow velocity.

Where the method drives only one transducer at a time during the measurement phase, the compensation phases may be omitted as there will be no unwanted reciprocal effects that are to be compensated for.

There will now be described by way of example only, various embodiments of the present invention with reference to and as illustrated in the accompanying drawings of which: Figure 1 is an illustration of a simple time of flight ultrasonic flowmeter to which the present invention is applicable; Figure 2 is an illustration of an alternative flow meter which employs reflections to increase the path length; Figure 3 is an illustration of a an axial flow meter to which the present invention is applicable; Figure 4 is a diagram explaining the method and apparatus for flow

measurement of the prior art;

Figure 5 is a diagram of an alternative prior art method based upon phase measurement; Figure 6(a) shows the apparatus and (b) the signals produced from an apparatus in accordance with a first embodiment of the present invention Figure 7 is an illustration showing the paths from both reciprocal (wanted) and non-reciprocal (unwanted) signals in the apparatus of Figure 6; Figure 8 illustrates the signals present in an alternative embodiment of the present invention which provides sufficient information when combined to provide compensation for non-reciprocal signals; Figure 9 is an illustration of a first embodiment of a drive circuit of a flow measurement apparatus in accordance with one aspect of the invention; and Figure 10 is an illustration of a second embodiment of a drive circuit of a flow measurement apparatus in accordance with one aspect of the invention.

The basic operation with respect to an axial flowmeter is shown in figure 6. In figure 6a two ultrasonic transducers 1 and 2 are located at opposite ends of the measurement tube. The transit time difference is measured along the length of the flow tube separating the two transducers. The transit time difference is thus defined by equation 1 above.

The method of operation will now be described with respect to a method employing a burst of a single frequency and simultaneous transmit and receive. Figure 6a show the transmitter/receiver circuits 3 and 4. These transmitter/receiver circuits enable the transducers 1 and 2 to be operated in a reciprocal manner, i.e. drive from and receive into the same electrical impedance. A simple form of reciprocal drive is voltage drive/current detect, i.e. drive from and receive into a zero impedance source, although any circuit which achieves the same drive and detection impedance can be employed.

It is known that employing reciprocal drive and driving both transducers with identical waveforms, given the transducers are operated in their linear regime and that the receive signal only comes from the other transducer, that the two receive waveforms will be identical at zero flow conditions. This is generally employed in conventional measurement systems to ensure that the detection of time differences upstream and downstream on the two transducers is undertaken on signals which are identical and also to guarantee the zero stability of the flowmeter.

The new method has two phases: a measurement phase and a calibration phase. In the measurement phase the drive signals applied to transducers 1 and 2 are identical in magnitude but exactly in anti-phase. After receiving each of the signals they are summed. Therefore at zero flow the two receive signals sum identically to zero as shown in figure 6b assuming that the only received signals are reciprocal. Under flow conditions the burst from transducer 1 is phase advanced and that from transducer 2 is phase retarded. Thus the signals no longer sum to zero.

The output is then related to the amplitude of the individual received signals and the phase difference between them. Figure 6b shows the output amplitude for different flows and hence different phase differences. If the amplitude of the individual receive signals is Rcaj (and they will both be identical as a consequence of the reciprocity theorem) and then the amplitude of the sum, Rmeas, for a phase difference D is given by Rmcas Rcai Sjfl c1 (4) From which c1 can be estimated since Sjfl D = / Rcai (5) For small D, sin D cD in radians and thus the time difference and hence the velocity can be measured by measuring the amplitude of one or other of the individual receive signals and their sum signals. For the velocity identified in the example given above with a phase difference of approximately 3 between the two waveforms, if the individual receive waveforms are converted to voltage signals of 1V amplitude, the sum signal will have an amplitude of approximately 5OmV. This can be easily amplified and measured. The amplitude of the received waveform can be measured either by measuring the peak amplitude of the received waveform or digitally sampling and storing the received signal at an appropriate sampling rate and amplitude resolution and using the stored signal to provide an estimate of the amplitude of the received signal.

Analog to Digital converters with appropriate speed and resolution are known in the art.

The calibration phase of the measurement consists of measuring the amplitude of either of the receive signals (which will be identical because of the reciprocal drive/receive method). This is achieved by driving only one transducer and receiving on the other. Measurement of the two signals Rmeas and Rca enables the phase difference D and hence ET and subsequently v to be measured using equations (5), (3) and (2).

Differences in receive signals can occur as a consequence of nonreciprocal action. These can occur as a consequence of non-linearity in the transducer or the transducer receiving signals which were transmitted by itself. For piezoelectric transducers which are commonly used, the first effect can be made negligible by restricting the level of drive voltage applied to the transducer. Figure 7 shows the non-reciprocal signals in an axial flowmeter. These can Occur as a consequence of either the transducer continuing to ring down after the drive signal has been stopped or from reflection occurring as a consequence of discontinujtjes or reflections which occur in the piping system. Should these occur during the period of the received reciprocal signals, then they can cause errors in the measurement method identified above. These effects may be more significant in flowmeters having a short path length.

The scheme shown in Figure 8 accounts for the non-reciprocal signals and enables the phase difference to be measured accurately in the presence of non-reciprocal signals. The method extends the number of phases in the measurement from 2 to 4. These phases are the measurement phase, compensating phase 1, compensating phase 2 and a calibration phase.

In the measurement phase the two drive signals to transducers 1 and 2 are of the same amplitude and exactly in anti-phase. After receiving each of the signals they are summed. This summed received signal now contains signals which arise as a consequence of reciprocal action and signals from each of the transducers which occur as a consequence of non- reciprocal action. The signals will therefore not sum to zero at zero flow and the amplitude will not be an accurate measure of the phase difference between the two signals. During this phase the summed receive signal is sampled at an appropriate sampling rate and resolution. A sample sequence Smea, (T0 +nT); n = 1....N is obtained where T0 is the start time of the sampling, n the sample number, T the sampling period and N the number of samples taken. T0, T and N are chosen to capture the receive signal with the required time resolution. A sequence corresponding to the difference of the two receive signals when the transducers are driven with the same transmit signal is thus created and stored.

The compensating phase 1 consists of driving the transducer 1 only and receiving the output from transducer 1. This output consists therefore only of signals generated by transducer 1 and is therefore its nonreciprocal Contribution made to the output in the measurement phase.

This receive signal is sampled and stored at exactly the same rate and resolution as during the measurement phase. This creates a sample sequence Scompi (T0 +nT) using the same T0, T and N as in the measurement phase.

The compensating phase 2 consists of driving the transducer 2 only and receiving the output from transducer 2. This output consists therefore only of signals generated by transducer 2 and is therefore its nonreciprocal contribution made to the output in the measurement phase. The receive signal is sampled and stored at exactly the same rate and resolution as during the measurement phase. A sample sequence Scomp2 (T0 i-nT) is thus created using the same T0, T and N as in the measurement phase.

The calibration phase consists of driving either transducer 1 and receiving on transducer 2 or driving transducer 2 and receiving on transducer 1 in order to measure the amplitude of either one of the receive signals. This measurement will not contain any non-reciprocal signals since the same transducer is not being used for transmission and reception. The received signal is sampled and stored at the same rate and an appropriate resolution during the calibration phase. This creates a sample sequence S0 (T0 i-nT) employing the same T0, T and N as in the measurement phase.

After collection of data from the four phase, the following computation is undertaken. The sampled data from the compensating phases 1 and 2 is subtracted on a sample-bysample basis from the sampled data taken during the measurement phase, i.e: Srec (T0 + nT) = Smeas (T0 + nT) - Scompi (T0 + nT) - comp2 (T0 + nT) (6) The output of this computation is therefore a sampled signal which has only the sampled reciprocal data. This is used to the estimate the amplitude of the output caused by the phase difference between the two receive waveforms The sampled data from the calibration phase is then used to estimate the amplitude of the individual receive signals. This is then used in conjunction with the estimate of the amplitude of the output caused by the phase difference between the two receive waveforms and hence to compute the velocity using equations (5), (3), (2).

These estimates can be obtained either by making peak measurements or by averages across the whole cycles of the receive waveforms.

Compensation for the variation of speed of Sound in the fluid can be undertaken using one or more of the receive signals obtained when a single transducer is driven and the signal received from the other transducer. Methods for undertaking this are known in the art.

The method can also be employed for sequential operation. In sequential operation one transducer is driven and the other used as a receiver and then the roles reversed. The effect of non-reciprocal action in terms of reflections and ring down are eliminated in this method of operation.

There are two modes of operation that can be employed. The first is to apply the same drive signal to both transducers and generate the difference of the receive signals by subtraction. The second is to generate an anti-phase drive for one of the transducers and generate the difference signal by addition of the two signals. Both methods lead to the sequence Srec(To + nT) being created without the need for the compensating phases, although with a requirement for a greater number of bits in the analogue to digital converter in order to undertake the addition or subtraction accurately and the need for the time jitter on the analogue to digital converter to be low. Calibration of the difference signal is undertaken by using one of the receive signals.

In addition to the sinusoidal operation identified above the method can also be applied to pulsed operation where the receive signal is a pulse with ringing. If the reciprocally received waveform at each transducer at zero flow is given by R(t) then the reciprocally received waveform at flow conditions will be given by R12(t) and R21(t) where R12(t) is the signal received on transducer 1 from a transmission from transducer 2 and R21(t) is the signal received on transducer 2 from a transmission from transducer 1.

For small transit time differences the signals R12(t) and R12(t) can be represented as a Taylor expansion of the zero flow receive signal, i.e: R (7) 12 dt 2 2.dt2 4 and: R21(t)= R(t) -(O.11+ 2)*fl2 (8) where UT is the transit time difference and therefore the difference in the two waveforms under flow conditions is given by: R21(t)-R12(f) dR(t)T (9) Thus the transit time can be estimated from: (10) dR(t) dt The sampled received waveforms employing either the simultaneous or sequential method identified above can be used to determine the transit time difference. This requires measuring the difference signal between the two receive signals and the derivative of the receive signal from one or both of the receive signals. The effects of any non-reciprocal signals in the simultaneous method can be removed by using the four phase method identified above. Methods for obtaining the derivatives of these signals are known in the art, the simplest being achieved by taking the difference between adjacent samples.

The following describes an embodiment as shown in Figure 9 of the simultaneous single frequency measurement method. In this embodiment, the summation of the two receive signals is undertaken in the receive electronics. The measurement method depends on producing two drive waveforms identical in amplitude but opposite in phase; creating a reciprocal drive and receive signal and summing the resultant receive signals. In the case of sinusoidal operation the drive signals will typically consist of a number of cycles of a sinusoid of frequency within the frequency band of the transducers. The circuit shown in figure 9 meets these requirements. The two drive signals which are in anti-phase are generated by the phase splitter PSi. These are then applied to the drive/receive electronics. Transistors Tn andTr2 act as both transmitters and receivers. In the transmission period they act as emitter followers providing a voltage drive with low output impedance. In the receive period they act as common base amplifiers and therefore provide a current detector with a low input impedance. The current flowing in the collector circuit for a transistor with a high beta ratio is equal to the current flowing in the emitter. The collector resistor converts the current to a voltage for subsequent signal processing. The common collector resistor R of the circuit shown in figure 9 sums the currents from Tn and Tr2 and since the transducers Ti and T2 are driven in anti-phase the resultantcurrent is the difference current. The capacitors ci and C2 are used to balance the circuit during the phases when only one transducer is being driven. This limits the net current through the resistor R and thus the excursion of the collectors during the drive phase thus preventing the transistors Ti and T2 going into saturation and becoming non-linear. The capacitances of Ci and C2 are chosen to be approximately equal to the capacitance of the transducers Ti and T2 although since they are only acting as balancing components they do not have to be accurately matched. The switches Si, S2, S3 and S4 are digitally controlled analogue switches selected such that their ON resistance is small compared to the impedances of the transducers or capacitances at the The phases of operation of the system are as follows: Phase 1: Switches Si and S4 closed and switches S2 and S3 open during both the transmit and receive periods. Under such conditions both transducers Ti and T2 are driving and receiving. The signal across R is the difference between the reciprocal signals together with the non- reciprocal elements. This receive signal is sampled and stored and provides the systems with the digital sample sequence Smeas(To + nT).

Phase 2: Switches Si and S3 closed and switches S2 and S4 open during both the transmit and receive periods. Under such conditions only Ti is being driven and the receive signal is the non-reciprocal signal from Ti.

This receive signal is sampled and stored and provides the system with the digital sample sequence S0 (T0 +nT).

Phase 3: Switches S2 and S4 closed and switches Si and S3 open during both the transmit and receive periods. Under such conditions only T2 is being driven and the receive signal is the non-reciprocal signal from T2.

This receive signal is sampled and stored and provides the system with the digital sample sequence Scomp2(To +nT).

Phase 4: Switches Si and S3 closed and switches S2 and S4 open during transmit period and S2 and S4 closed and switches Si and S3 open during the receive phase. (Alternatively, switches S2 and S4 closed and switches Si and S3 open during transmit period and Si and S3 closed and switches S2 and S4 open during the receive phase). Under such conditions Ti or T2 is being driven and the receive signal is being detected on T2 or Ti.

This receive signal is stored and provides the system with the digital sample sequence Scai (T0 +nT).

The four digital sampled data sequences can then be used to effect the algorithms identified above from which the phase difference and hence the transit time difference can be obtained. Combined with the transit time measurements, obtained from S (T0 +nT) and or S2 (T0 +nT), the flow velocity which can be estimated using equations (6), (5) and (2).

The above provides one embodiment of the simultaneous method. The method requires the phase splitter to generate two signals which are identical in amplitude and exactly in ant-phase. The two transistors which are used as drivers and receivers must be well matched. The following provides the preferred embodiment for the simultaneous method. The circuit is shown in Figure 10. In the preferred embodiment shown in Figure 10 the power amplifier Al drives the primary of a toroidal transformer, TR1. The toroidal transformer has two secondaries which are wound as a bifilar winding. This ensures that the coupling of the two secondaries is very high. Opposite ends of the two windings are joined at the centre tap. This configuration ensures that the voltages induced across the windings are equal in magnitude and Opposite in phase.

Employing a low impedance driver together with the bifilar winding ensures that any loading effects occurring on one side of the secondary are reflected on to the other side such that the output from the two windings always remain equal in magnitude and opposite in phase. The transducers Ti and T2 are driven from the output of the transformer and the receive signal is detected by a single transimpedance amplifier A2.

This amplifier has low input impedance and converts the current fed into its input into an output voltage for further processing. The resistor Rre( sets the transimpedance value for the amplifier. Employing a single amplifier obviates the need for any matching between the two sides. The capacitors Cl and C2 act as balancing impedances during the drive phase to ensure that the transimpedance amplifier A2 is not driven into a nonlinear regime. The capacitance of these capacitors is selected to be approximately that of the transducers Ti and T2 although exact matching is not required since their purpose is purely balancing. The switches Si, S2, S3, and S4 are digitally controlled analogue switches selected such that their ON resistance is small compared with the impedances of the transducers Ti and T2 or capacitors Cl and C2 at the operating frequency.

The operation of the circuit as shown in Figure 10 is in four phases: Phase 1: Switches Si and S4 closed and switches S2 and S3 open during both the transmit and receive periods. Under such conditions, both transducers Ti and T2 are driving and receiving. The signal at the output of A2 is the difference between the reciprocal signals together with the non-reciprocal elements. This receive signal is amplified sampled and stored and provides the systems with the digital sample sequence Smea, (T0 + nT).

Phase 2: Switches Si and S3 closed and switches S2 and S4 open during both the transmit and receive periods. Under such conditions, only Ti is being driven and the receive signal is the non-reciprocal signal from Ti.

The output of A2 is amplified sampled and stored and provides the system with the digital sample sequence Scompi(To +nT).

Phase 3: Switches S2 and S4 closed and switches Si and S3 open during both the transmit and receive periods. Under such conditions, only T2 is being driven and the receive signal is the non-reciprocal signal from T2.

The output of A2 is amplified sampled and stored and provides the system with the digital sample sequence Scomp2(To + nT).

Phase 4: Switches Si and S3 closed and switches S2 and S4 open during transmit period and S2 and S4 closed and switches Si and S3 open during the receive phase. (Alternatively, switches S2 and S4 closed and switches Si and S3 open during transmit period and Si and S3 closed and switches S2 and S4 open during the receive phase). Under such conditions, Ti or T2 is being driven and the receive signal is being detected on T2 or Ti.

The output signal from A2 is amplified stored and provides the system with the digital sample sequence Scaj (T0 +nT).

While a preferred embodiment of the invention has been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope of the invention. The embodiment shown and described is for illustrative purposes only and is not meant to limit the scope of the invention as described by the claims.

Claims

1. A flow meter comprising a first transducer and a second transducer, each located at a spaced locations along a fluid flow path; a drive circuit which applies a first drive signal to the first transducer to cause the first transducer to emit radiation and a second drive signal to the second transducer to cause the transducer to emit radiation; a detecting circuit adapted to detect a first output signal produced by the first transducer which is at least partly due to receipt of the radiation from the second transducer and a second output signal produced by the second transducer which is at least partially due to receipt of radiation from the first transducer; and a processing means adapted to process the first and second output signals to generate a flow rate signal indicative of the velocity of flow of fluid along the fluid flow path; characterised in that: the processing means combines the first and second output signals to produce a time of flight signal whose amplitude is dependent upon the difference in time of flight between radiation transmitted from the first transducer to the second transducer and radiation transmitted from the second transducer to the first transducer, the flow rate being derived from the time of flight signal.

2. A flow meter according to claim 1 in which the first and second transducers are driven simultaneously or in sequence.

3. A flow meter according to claim 2 in which the detection of the two output signals is made simultaneously.

4. A flow meter according to any preceding claim in which the drive signals comprise identical signals, for example Sinusoids of the same frequency and amplitude, and in which the processing means determines the amplitude signal by subtracting one output signal from the other.

5. A flow meter according to any one of claims 1 to 3 in which the drive signals are sinusoids of the same amplitude but exactly in anti- phase, and in which the processing means is adapted to add the two output signals together to produce the time of flight signal.

6. A flow meter according to any preceding claims in which a single signal generator produces both drive signals.

7. A flow meter according to any preceding claim which further includes means for determining the amplitude of at least one of the drive signals.

8. A flow meter according to any preceding claim which is further adapted to transmit at least two additional signals, a first one of the signals being produced by applying a calibration signal to the first transducer equal to the first drive signal at a time when no drive signal is applied to the second transducer.

9. A flow meter according to claim 8 which is adapted to detect any output signal from the first transducer during the time of generation of the calibration signal.

10. A flow meter according to claim 8 in which a second one of the signals is produced by applying a second calibration signal to the second transducer and detecting any output signal from that transducer at that time.

11. A flow meter according to claim 11 which is further adapted to combine the output signals that result from the calibration signals with the time of flight signal and the calibration signal to produce a more accurate time of flight signal which is free from non-reciprocal effects.

12. A flow meter according to any preceding claim which includes a transformer having two bifilar windings, one terminal of each winding being joined to the other in such a way as to signals which are in antiphase, the output of the two windings providing the drive signals for the first and second transducers.

13. A method of measuring flow rate in a fluid employing at least two ultrasonic transducers in which the transit time difference is measured between ultrasonic beams transmitted in opposite directions between the two transducers.

14. The method of claim 13 in which the time difference is measured by simultaneously driving both transducers and measuring the difference in the transit times.

15. The method of claim 13 or 14 which further comprises the step of measuring the time difference by transmitting a burst of a single frequency or a continuous signal, such as a sinusoidal waveform.

16. The method of any one of claims 13 to 15 in which the signals are transmitted and received reciprocally, i.e. from the first to the second transducer and then the second to the first and so on.

17. The method of any one of claims 13 to 16 which comprises transmitting signals of the same amplitude and phase from both transducers and combining the measured signals from both transducers by subtracting one from the other so that the difference between the two receive signals is generated.

18. The method of any one of claims 13 to 17 which further includes a step of using the estimates of the amplitudes of the individual receive signals to remove the non-reciprocal elements of the difference of the two receive signals and create an estimate of the difference in two reciprocal receive signals.

19. The method of any one of claims 13 to 18 in which the drive signals are generated from the output of two bifilar windings of a toroidal transformer, one terminal of each winding being joined to the other in such a way as to signals which are in anti-phase

20. The method of any one of claims 13 to 19 which comprises four phases: a drive phase, a calibration phase and two compensation phases.

21. A flow meter substantially as described herein with reference to and as illustrated in the accompanying drawings.