DK177567B1 - System or a method for measuring flow of fluid or gas - Google Patents

Abstract

The present invention relates to a system or a method for measuring flow in a flow duct, comprising at least two ultra sound transducers. It is the object of the pending application to measure the flow of air in a duct by one or more transducers transmitting beams of ultra sound controlled by a microcontroller based electronic system. The object can be achieved if the microcontroller stores a vector of data samples for each direction of transmission, which vector comprises an appropriate number of N samples forming a frame, which microcontroller multiply each value of the frame which a complex number, which microcontroller based on the result caiculates the flow in the duct. By the invention according to the present patent application an efficient flow measurement of air flowing in a duct can be achieved.

Description

i DK 177567 B1

System or a method for measuring flow of fluid or gas Field of the Invention

The present invention relates to a system or a method for measuring flow of fluid or 5 gas, which system comprises a flow duct, which flow duct comprises at least two transducers, which transducers generate at least one beam of ultrasound in the flow duct, which Transmitter/receiver circuits are connecting through a transmitter/receiver switch, which switch in a first position connect first Transmitter/receiver circuit to a transmitter circuit and second transducers are connected to a receiver circuit, which 10 receiver circuit comprises at least a band pass filter, which band pass filter is further connected to a microcontroller, which microcontroller comprises an analogue to digital converter, which digital converter converts the analogue signal into digital data samples representing at least transit times and time difference, which microcontroller stores the data samples in a memory.

15 Background of the Invention W02010/122117 discloses a ventilation system, wherein a fan draws air from an exterior of a building or similar closed construction for circulation into an interior of the building, and produces supply airflow through a ventilator unit, which is able to cool and heat air. Moreover, the ventilator unit may be equipped with a humidify-20 ing/dehumidifying device. A controller controls the position of a valve or indirectly the speed of the fan, thereby adjusting the supply airflow in response to an input received from an ultrasound device placed in the supply air duct, wherein the ultrasound device measures the air flow and temperature.

Object of the Invention 25 It is the object of the pending application to measure the flow of air in a duct by one or more transducers transmitting beams of ultra sound controlled by a microcontroller based electronic system.

Description of the Invention

The object can be achieved if the transmitter circuit and a receiver circuit are con-30 trolled by a switch, which switch perform continuous switching of transmitter circuit 2 DK 177567 B1 into a receiver circuit and the receiver circuit into a transmitter circuit, which microcontroller stores a vector of data samples for each direction of transmission, which vector comprises an appropriate number of N samples forming a frame, which microcontroller multiply each value of the frame which a complex number with the magni-5 tude one and a phases representing the transmitted frequency and phase, which microprocessor generates imaginary values and a real values, which imaginary values and a real values low pass filtrated in a digital filter, which filtrated values are sent to an amplitude function and to a phase detection function, which microcontroller based on the result of the amplitude function and the result of the phase detection function cal-10 culates the flow in the duct.

By the invention according to the present patent application an efficient flow measurement of air flowing in a duct is achieved. It is realized that the damping of signals between two transducers in air is sufficiently higher than the damping that is found in 15 traditional ultrasound flow measuring systems simply because a liquid is better matched to the impedance of the transducers. Because of the relatively weak and long signals that are transmitted and received, it is necessary to isolate the receiver from the transmitter in order to get good results of the signals received. By treating the received signals as an oscillating signal which has a curve form which is mostly a sinus curve it 20 is possible to handle that sinus form mathematically and to divide it into the imaginary and real parts of the signal. This can lead to a situation where the amplitude can be very accurately detected, but also very precise phase detection can be made. By performing an efficient low pass filtration of the signals it is achieved that all signals having higher frequencies are reduced so their influence on the measurement is non-25 existing. This can lead to a much higher precision of the measured signals.

The result of the amplitude function can further be processed by the microcontroller in a digital constant fraction discriminator (CFD). Hereby it is achieved that arrival times T up and T down are measured by the constant fraction discriminator. The output from 30 the CFD is used to place the sampling frame so it preferably starts where the pulse would have been if there had been no dead delay. The dead delay is the delay due to signal transmissions through cables and influence from solid parts of the transducers and also if there is a delay in one of the end filters used. The start of the sampling 3 DK 177567 B1 frame can partially equal the transmission time in the air T UP and T DWN. The frequency of the time clock will limit the resolution, but the precision is sufficient for the T UP and DWN values in the denominator but not sufficient for the various delta-Ts since this requires a higher resolution than that by which the sampling frame is adjust-5 ed.

The result of the phase detection function can further be processed by the microcontroller in a digital signal representing the time shift relative to the frame. Hereby an increase is achieved of the resolution of the time difference T UP -T DWN = Delta T.

10 Hereby highly efficient phase detection can be performed. Hereby a relative decrease can be achieved in standard deviations with the above method compared to the deviation obtained from a zero-crossing detection of the same signal as that known from the prior art.

15 The system comprises a transmitter circuit comprising a band pass amplifier to limit the bandwidth of a transmission burst from the microcontroller, which band pass amplifier transmit signal through the switch and further to one of the transducers. Hereby it is achieved that the band pass amplifier can limit the band width of the transmission first from the microcontroller. Signals generated by the microcontroller do have a high 20 slew rate That high slew rate must be reduced in order not to overload the transmitting amplifiers because these amplifiers have a maximum slew rate to which the slew rate of the output signal becomes limited. With the amplifiers in an active state it can be achieved that the transducers are connected to the same node in the circuit both during transmission and reception, and are always loaded with the same impedance.

25

The system comprises a transducer amplifier, which transducer amplifier connects the transducer to the same node in the circuit both during transmission and reception. The transducer amplifier comprises a voltage follower. Hereby it can be achieved that the voltage on the positive input terminal is followed by the negative input terminal, and if 30 a signal appears on the negative input terminal, it is amplified and inverted on the output of the amplifier. When the amplifier is used for transmitting the signal, the amplifier is a power amplifier of a large signals, and when receiving the amplifier acts as a pre-amplifier of small signals. The diodes coupled anti parallel over the feedback re- 4 DK 177567 B1 sistor prevents the amplifier output from becoming into saturation and allow for appropriate amplification of the received signal.

The system comprises at least one transmitter receiver switch, which transmitter re-5 ceiver switch comprises at least three contacts, by which contacts not use terminals are ground connected for reducing noise. Hereby it is achieved that an extra switch is coupled between the contacts in the first switch and perform capacitive coupling to the ground instead of the transmission signal from the other channel. This is especially important in air transducers where the received signal is normally 60 dB lower than 10 the transmitted signal and the transmitting transducer continues ringing for a long time after excitation.

In a preferred embodiment for the invention the band pass amplifier comprises automatic gain control, which automatic gain control comprises an operational amplifier 15 connected with a variable feed back resistance, which feed back resistance is controlled by the microcontroller. Hereby it is achieved that not only the band width is reduced by the band pass filtration but also the gain control of the signal is under control. Depending on reflection or non-reflection or depending on the diameter of the duct, or maybe the change from one type of transducer to another type, there can be an 20 extremely high difference in signal level that is received. Highly efficient transducers in a very small duct can produce a relatively good signal even in air but in a different situation where the diameter of the tube is much higher, and maybe a rather less efficient transducer is used, the damping of the signal can be several thousands times, and a much higher amplification is necessary. Only in this way it can be achieved that ana-25 log digital conversion can be made in a perfect way. The upper frequencies are limited to half the sampling frequency of the digital converter, and the gain has to be controlled so that also the signal level is inside the level for the digital conversion.

The pending patent application further concerns method for operating a system for 30 measuring flow of fluid or gas, as previous disclosed, which method concerns the following steps of operation: 5 DK 177567 B1 a: generate at least one beam of ultrasound in the flow duct by a transmitter circuit connected to a first transmission transducer, b: receive the beam of ultrasound in the flow duct by the second transducer and the 5 receiver circuit, c: perform a band pass filtration of the received analog signal in the receiver circuit, d: convert the analog signal in a digital signal in the microprocessor into digital data 10 samples representing at least transit times and time difference, e: store the data samples in a memory f: stores a vector of data samples for each direction of transmission, 15 g: forming a frame based on the vector comprising an appropriate number of N samples, h: multiply each value of the frame with a complex number with the magnitude one 20 and a phases representing the transmitted frequency and phase, i: generates imaginary values and a real values, which imaginary values and a real values low pass filtrated in a digital filter, 25 j. send the filtrated digital signal to an amplitude function and to a phase detection function, k: calculate the flow in the duct based on the result of the amplitude function and the result of the phase detection function.

30

By this method a very efficient flow measurement in an air duct is achieved. Transducers can be placed across the duct, they can be placed so a reflection takes place, or they can be placed inside the duct so that the transmission of sound is effected directly 6 DK 177567 B1 between the transducers. By the invention according to the present application it can be achieved that relatively simple and inexpensive transducers can be applied. Therefore, by the invention according to the present application it is possible to achieve an inexpensive but yet highly efficient system for measuring air flow.

5 Description of the Drawing

Fig. 1 shows a possible embodiment of the invention.

Fig. 2 shows a first alternative embodiment of the invention.

Fig. 3 shows a principle diagram of the various technical features necessary for computer analysis of received signals.

10 Fig. 4 shows a digital constant fraction discriminator CFD.

Fig. 5 shows a pre-amplifying system connected to the transducers.

Fig. 6 shows a possible embodiment of the same amplifier used both as transmitter and receiver.

Fig. 7 shows a possible embodiment of a switch.

15 Fig. 8 shows a possible embodiment of an automatic gain control.

Fig. 9 shows an alternative embodiment of an automatic gain control.

Fig. 10 shows a curvature of gain vs. binary digits Fig. 11 shows a possible embodiment of a transducer.

Fig. 12 shows a transducer placed in housing.

20 Fig. 13 shows a possible embodiment of a reflective system with two transducers and a reflecting mirror.

Fig. 14 shows examples of various disadvantages of the reflecting system.

Fig. 15 shows alternative embodiment with reflections.

Detailed Description of the Invention 25 Fig. 1. shows a system 2 which system comprises a flow duct 4 where a first transducer 6 and a second transducer 8 are transmitting acoustic waves 10 across the duct 4. Both transducers 6 and 8 are connected to a switch 12 by which switching for receiving and transmission is effected. The transmitter receiver switch 12 is connected to a transmitter circuit 14 and to a receiver circuit 16. The receiver circuit 16 comprises a 30 band pass filter 18 which is connected to a microcontroller 20, and in this microcontroller 20 it is connected to a digital converter 22. The microcontroller 20 further comprises a digital filter 38, and an amplitude function 42 and phase detection function 44.

7 DK 177567 B1

The microcontroller 20 further comprises a digital constant fraction discriminator 46. Further, in the microcontroller 20 a voltage follower 50 is shown. The microcontroller 20 is further connected to the transmitter circuit 14 which transmitter circuit 14 comprises a band pass amplifier 48.

5

In operation the flow will be measured between the transducers 6 and 8 where in one situation, the transducer 6 acts as transmitter and the transducer 8 acts as receiver, and in the next situation the transmission occurs in the opposite direction where the transducer 6 is the receiver and the transducer 8 is the transmitter. On the basis of these 10 signals and by the means provided inside the microcontroller 20 the system can calculate the flow in a highly efficient manner.

Fig. 2 and 3 discloses that the received and band pass amplified signal is analog to digital converted by a build in analog to digital converter and stored in memory 123.

15 The measurements of transit times and time difference are solely done by the micro controller 120 based on these stored values.

After transmission in either direction a number of samples are stored the memory 123. The sampling starts after a time determined by an internal timer in the microcontroller 20 120 such that the received pulse is sampled from the beginning. The time for the first saipple is stored in the memory 123 one value for transmission against the flow and another value for transmission with the flow.

A vector 130 of samples is stored for each direction of transmission. Each vector 130 25 contains an appropriate number N of samples, in the actual embodiment the vector 130 contains 512 samples, but less may work well.

The first step in the signal processing is to multiply each value in said frames 132 with a complex number 164 with the magnitude one and a phase corresponding to the 30 transmitted signal: K = X» eJaH‘ "=X„' (cos(<u -ts-n)+j sin {ω · ts ·«)) j 64 8 DK 177567 B1

Where X„ is the stored value at the n,h location, ja> the angular frequency of the transmitted signal, ts the sampling time interval and n is the sample number. 0<n<N The complex result Y„ is filtered by two low pass filters, one for the real part a„ and one for the imaginary part b„.

5

After low pass filtering 116im, 116re the result is a complex sequence (a„+jb„) with n = {0, 1,.. N-l, N}. The amplitude 142 of the received signal can sample for sample be found as the square root of the sum of squares.

K 2 42 10 Fig 4 shows a possible embodiment for an amplitude signal is used to determine the arrival times Tup and 7^v„ by means of a digital constant fraction discriminator CFD 146.

The output from the CFD 146 is used to place the sampling frame so it starts where 15 the pulse would have been if there has been no dead delay. The dead delay is the delay due to cables, solid parts of transducers and delay in the band pass filter. The start of the sampling frame shall ideally equal the transmission time in the fluid Tup and Τ^„. The frequency of the timer clock limits the resolution, but the precision is sufficient for the Tup and Tdwn values in the denominator, but not sufficient for the difference At 20 since it requires more resolution than the sampling frames are adjusted with.

The purpose of the phase detection 144 is to increase the resolution of the time difference Tup - Town = At. The output from the low pass filters 116im,116re represents the phase difference between the frame 132 with its reference sequence and the received 25 signal. The filtered output can be further filtered to decrease the standard deviation on the phase measurements. The phase is the argument to the last complex number (a^-i + jbN-i) in the vector 130 where the amplitude and phase are stable.

Call the signal from the upstream measurement for (a^-i + j&v-y) and the downstream 30 measurement for (cn-i + jd^.i) then the phase difference is: Αφ = arg ——f-111 VCN-1 + j^N-l > 9 DK 177567 B1

The argument can be found as shown below: aN-l j^N-1 _ aN-l j^N-1 CN-1 + J^N-1 _ aN-l ' CN-1 + ^N-l ~^N-1 X^N-l ~CN-1 ~&Ν-1 '^N-l) CN-l+j^N-l CN-l+j^N-l CN-l"^j^N-l CN-1 A^ = arctani^-1'CN-1~aN-1'dN-1

VaN-l ‘CN-I +^N-1 '^N-ly 5 Since the angle is small if the frame is adjusted with steps much smaller than the sample time the arctan is easily calculated by the series: x3 xs x7 arctan(x) = x--+---...

3 5 7

The actual length of the series depends on the required accuracy and the range of values of x, but the shown length will normally suffice.

10

If noise is present the CFD signal 146 may fluctuate with a few samples and the nominal value of the division may be larger than the range of the arc tan function, but a few numbers out of range can be discarded without offsetting the mean value of measurements since the deviations are expected to be symmetrical.

15

Practical measurements have shown a three to five times decrease in standard deviation with the above method compared to the deviations obtained from a zero crossing detection on the same signal.

20 The band pass amplifier is necessary to limit the bandwidth of the transmission burst from the microcontroller. A square wave like the burst from the microcontroller have too high slew rate which will bring the operational amplifier in the T/R switch 112 into slew rate limit and ruin the essential reciprocity of the T/R switch 112.

25 The transducer amplifiers can be coupled as voltage followers 150 or as current generators.

The main difference to common practice is that the transducer is connected to one node in the circuit without switches, the input of the transmission signal to another and 30 the received signal appears on a third. The transducer is connected to the same node in 10 DK 177567 B1 the circuit both during transmission and reception and hence always loaded with the same impedance. Hereby it obeys the “reciprocity theorem” making the time delay difference (Difference in transmission time with or against the flow.) unchanged by transducer changes due to temperature, contamination or aging.

5

Fig. 5 shows a possible embodiment for an amplifier circuit for connoting the transducers 20,208. The voltage on the positive input terminal of one of the operational amplifiers 214,216 is followed by the negative input terminal and if a signal appears on the negative terminal it is amplified and inverted on the output of the amplifier 10 214,216. When the amplifier 214,216 is used for transmitting a signal the amplifier 214,216 is a “power amplifier” of a large signal and when receiving the amplifier act as a preamplifier of a small signal. The anti parallel diodes 260,262 serve as low impedance during transmission and as high impedance during receive mode preventing the amplifier 214,216 from going into saturation under transmission of a large signal.

15 Under reception of small signals they act virtually as disconnections compared to the resistor they are parallel to.

The reciprocity theorem requires the transmitting transducer to be driven with the same impedance as the transducer used as receiver. In fig. 5 the impedances are virtu-20 ally zero, but any impedance can be used.

The circuit in fig. 6 has virtually infinite impedances as a current generator both as power amplifier and as preamplifier.

25 In principle the T/R switch 312 can look like in fig. 5. But in most cases it is too simple. In practice a more elaborate scheme must be followed to avoid over coupling of the transmission signal via the off capacitance of the switch 312.

An example of minimizing the influence of the off capacitance in the switch is shown in fig. 7.

An extra switch 314, 316 is coupled between so the capacitive coupling is to ground 318, 320 instead of to the transmission signal from the other channel. This is especially important in air transducers where the received signal is normally 60 dB lower than 30 DK 177567 B1 the transmitted signal and the transmitting transducer continues “ringing” long time after the excitation.

π

Fig. 8 discloses a band pass amplifier 402 with AGC. The signal 404 from the front 5 end is for air transducers in the range of few millivolts and hence too small to be analog to digital converted by the build in analog to digital converters in standard microcontrollers. At the same time the sampling rate is with present technology of low cost microcontrollers 1 to 2 million samples per second. In order to avoid aliasing all frequencies above half the sampling frequency must be removed before digitizing. The 10 frequencies used in air flow meters are up to 250 kHz and if 500 kHz shall be damped say 60 dB it requires a low pass filter with very sharp cutoff or a more than 10th order filter.

Alternatively and much better is a band-pass filter. The required bandwidth is 5 to 10 15 kHz depending of the transducer used. Sallen Key or multiple feedback active filters are appropriate, but other filter types as passive LC filters, switched capacitor filters or even mechanical filters can be used.

Due to fabrication tolerances and temperature variation the signal amplitude will 20 change from transducer to transducer and during operation. In order to minimize the digitizing noise the analog to digital converter shall have the full dynamic range utilized, hence the controller shall be able to adjust the amplification. In order to keep the dynamic range of the analog to digital converter utilized the amplification must be changed in appropriate small steps preferable in a converter within certain limits the 25 AGC shall work in a way that gives the same percentage gain increase pr step. Depending on the gain variation the necessary number of steps and the size of each step shall be chosen.

A simple 1 of 8 multiplexer 4051 type number can do the task by selecting feedback 30 resistance 406 in an amplifier as shown fig. 8

If larger gain variation is necessary due to the same electronic unit be used for many different tube diameters, or if there exist possibility for contamination that may damp- 12 DK 177567 B1 en the signal, a digital resistor 406 with 1024 steps may be preferred. Since the steps are linear in most commercial versions a scheme like the one shown in fig. 9 may be used.

5 This circuit gives a total gain variation of nearly 30 dB distributed over the 1024 steps as shown hereunder:

Fig. 10 shows a coordinate system where a curve indicates the correlation between the gain and the binary digits. As can be seen from the curvature, increasing binary num-10 bers will achieve a much better gain.

By using an analog signal from either a digital to analog output or a filtered pulse width or rate modulated signal from the processor analog variable gain amplifiers or circuits with diodes or voltage dependent resistors can be used. Also use of PTC resis-15 tor circuits or the like that thermally changes attenuation on high signal amplitude may be used.

Fig 11 discloses the preferred transducers which are common piezoelectric transducers with a piezoelectric element 604 exited at the lowest radial resonance frequency and 20 approximately a quarter wavelength silicon rubber disk 606 as impedance alignment is glued to the front surface of the piezoelectric element 604.

Alternatively transducers used for parking sensors in cars can be used. These normally work at 40 kHz and have a wide angle transmission pattern. If these are used only direct transmission between transducer are used, due to the risk of direct transmission of 25 a spurious signal during the transmission of the reflected signal.

Fig 12 shows a possible embodiment for a transducer and transducer housing. The preferred embodiment is shown in fig. 12 and the scheme below, but other forms either preformed or molded in place may be used.

13 DK 177567 B1 608 Transducer housing_ 610 Silicon rubber foam (1,5 mm 6 x 58,5)_ 604 Silicon rubber foam (1,5 mm 0 16) 612 Silicon rubber foam (1,5 mm 0 20) 606 PZT (2x015,5)_ 614 Silicon rubber (1,6 x 0 16)_ 616 Metal screen (solder able)_ 618 I Screened cable_

Since the sound more readily goes through solids than through air the transducer must be isolated acoustically from the duct 4,104 otherwise some sound would be transmitted through the tube wall and arrive at the receiving transducer 6,8,106,108 and inter-5 fere with the flow signal and create non linearity. The isolation can be performed with foam of silicon rubber, parts 610, 604 and 612 in the fig. 12. The material of the said impedance alignment disk can be various other materials with low acoustic impedances and loss, e.g. resin filled with hollow glass spheres or hard foams.

10 Fig. 13 discloses a flow duct 704 with one reflection 710 whereby the two transducers 706.708 become placed on same side next to each other. The distance between the transducers is the same for all tube diameters; hereby the time difference At for same flow and temperature becomes the same for all sizes.

Sensitivity of the flow duct 15 Let C be the sound velocity, D the tube diameter, Lx the distance between the transducers.

Transducers alternately transmit ultrasonic pulses and alternately receive said transmitted pulses. Hereby the transmission goes with the flow and against it alternately.

20

By Pythagoras the expressions can be written: (CTvpy =(2D)7 +(Lx + VTvry and (C · TD„ )2 = (2 · Df + (Lx - V TmK )!

First solve both equations for C and set the results equal to each other and solve for V: 14 DK 177567 B1

γχ _ 4 · D + Lx TUP — TDWN 2- Lx κ Tup · TDWN J

or r, 4-d2+Lx2 f At Λ

Vx —---

2· Lx ^ TUP · TDWN J

Where At is equal to Tup-Tdwn and is found with high precision due to the coherent 5 detection principle.

Benefits with the reflection in the flow duct configuration:

The flow meter will be inserted in tubes with nominal bore according to standards. Hence from Tup and Tdwn it can be determined which standard diameter the flow meter is inserted in and the appropriate calibration constant can be selected from a 10 table stored in memory.

Since sound traverse the diameter twice in opposite direction secondary flow orthogonal to tube axis is partially canceled, hence some common flow disturbances have limited influence. The flow profile due to laminar and turbulent flow does have influence, 15 but since the fluid always is air the Reynold number influence can be compensated for by calculation based on Tup and Tdwn.

The shift from laminar to turbulent flow creates a known shift in flow profile the shift appears where Reynolds number is: 20 · laminar when Re < 2300 • transient when 2300 < Re < 4000 • turbulent when Re > 4000

Reynolds number Re is:

V

25 p is the density of air: P = ~^

RT

Where p is the pressure in pascal, R =287.05 J/(kg-K) the specific gas constant and Tk the temperature in kelvin. (The density for this purpose can be assumed to be 1.2 kg/m3) 15 DK 177567 B1 dh is the hydraulic diameter of the pipe and vis the kinematic viscosity of air.

v = (0.OOO2-TK2 + 0,0053·TK -0.0327)-10-6 —1

Where Tk is temperature in kelvin.

5 The necessary correction will be determined for each dimension by flow tests, but there exists theoretical/empirical formulas for corrections.

By these measures it is possible to minimize deviations to a few percent even with flow disturbances as close as a few diameters from the inlet.

10 For two reflections two parabolic mirrors shall be used and for three reflections two parabolic and one flat mirror preferably shall be used.

Normal direct transmission with one or more tracks is possible and will be used as alternative. The calibration constants of this are found by standard well proven princi-15 pies.

Claims (9)

16 DK 177567 B116 DK 177567 B1 1. System (2) til måling af strøm af væske eller gas, hvilket system (2) omfatter en strømningskanal (4), der omfatter mindst to transducere (6, 8), hvilke transducere (6, 5 8) genererer mindst en stråle (10) af ultralyd i strømningskanalen (4), hvilke transdu cere (6, 8) er tilsluttet et sendekredsløb (14) og et modtagekredsløb (16), hvilket modtagekredsløb (16) omfatter mindst et båndpasfilter (18), der yderligere er forbundet med en mikroprocessor (20), hvilken mikroprocessor (20) enten internt eller eksternt omfatter en analog-til-digital omformer (22), som omformer det analoge signal til di-10 gitale data samples (24), som i det mindste repræsenterer transittider (26) og tidsforskel (28), hvilken mikroprocessor (20) lagrer data samples (24) i en hukommelse (28), kendetegnet ved, at et sendekredsløb (14) og et modtagekredsløb (16) styres af en kontakt (12), som udfører kontinuerlig tilslutning af sendekredsløbet (14) til et modtagekredsløb (16) og modtagekredsløbet (16) til et sendekredsløb (14), hvilken mikro-15 processor (20) lagrer en vektor (30) af data samples (24) for hver senderetning, hvilken vektor (30) omfatter et passende antal af N data samples, som danner en båndkolonne (32), hvilken mikroprocessor (20) ganger hver værdi af båndkolonnen (32) med et komplekst tal med fast størrelse og en fase, der repræsenterer den sendte frekvens svarende til det sendte signal:A system (2) for measuring the flow of liquid or gas, said system (2) comprising a flow channel (4) comprising at least two transducers (6, 8), which transducers (6, 5 8) generate at least one beam (10) of ultrasound in the flow channel (4), which transducers (6, 8) are connected to a transmitting circuit (14) and a receiving circuit (16), the receiving circuit (16) comprising at least one bandpass filter (18) further connected with a microprocessor (20), which microprocessor (20) comprises, either internally or externally, an analog-to-digital converter (22) which converts the analog signal to digital data samples (24) representing at least transit times (26) and time difference (28), which microprocessor (20) stores data samples (24) in a memory (28), characterized in that a transmit circuit (14) and a receive circuit (16) are controlled by a switch (12), which performs continuous connection of the transmit circuit (14) to a receiving circuit (16) and receiving circuit t (16) to a transmit circuit (14), which micro-processor (20) stores a vector (30) of data samples (24) for each transmitter direction, which vector (30) comprises an appropriate number of N data samples which forming a band column (32), which microprocessor (20) multiplies each value of the band column (32) by a complex number of fixed size and a phase representing the transmitted frequency corresponding to the transmitted signal: 20 Yn = Xn = Xn {cos(co ts n)+jsm(o> ts ·«)) m hvor X„ er den lagrede værdi ved den n,e placering, vinkelfrekvensen af det sendte signal, ts tidsintervallet for data samplingen, og n antal samples 0 <rt<N f hvilken mikroprocessor (20) genererer imaginære værdier (34) og reelle værdier (36), hvilke imaginære værdier og reelle værdier lavpasfiltreres i et lavpasfilter (38), hvilke 25 filtrerede værdier (40) sendes til en amplitudefunktion (42) og til en faseregistreringsfunktion (44), hvilken mikroprocessor (20) på basis af resultatet af amplitudefunktionen (42) og resultatet af faseregistreringsfunktionen (44) beregner strømningen i kanalen (4).20 Yn = Xn = Xn {cos (co ts n) + jsm (o> ts · «)) m where X„ is the stored value at the n, e location, the angular frequency of the transmitted signal, ts the time interval for the data sampling, and n number of samples 0 <rt <N f which microprocessor (20) generates imaginary values (34) and real values (36), which imaginary values and real values are low pass filtered in a low pass filter (38), which filtered values (40) are sent for an amplitude function (42) and for a phase detection function (44), which microprocessor (20), based on the result of the amplitude function (42) and the result of the phase detection function (44), calculates the flow in the channel (4). 2. System ifølge krav 1, kendetegnet ved, at resultatet af amplitudefunktionen (42) yderligere behandles af mikroprocessoren (20) i en digital konstantfraktiondiskrimina-tor (CFD) (46). DK 177567 B1 17System according to claim 1, characterized in that the result of the amplitude function (42) is further processed by the microprocessor (20) in a digital constant fraction discriminator (CFD) (46). DK 177567 B1 17 3. System ifølge krav 1, kendetegnet ved, at resultatet af faseregistreringsfunktionen (44) yderligere behandles af mikroprocessoren (20) i et digitalt signal, der repræsenterer tidsforskydningen i forhold til båndkolonnen (32). 5System according to claim 1, characterized in that the result of the phase detection function (44) is further processed by the microprocessor (20) in a digital signal representing the time offset relative to the band column (32). 5 4. System ifølge et af krav 1-3, kendetegnet ved, at systemet (2) omfatter et sendekredsløb (14) omfattende en båndpasforstærker (48) til at begrænse båndbredden af en transmissionsbyge fra mikroprocessoren (20), hvilken båndpasfilterforstærker sender signal gennem kontakten (12) og videre til en af transducerne (6, 8). 10System according to one of claims 1-3, characterized in that the system (2) comprises a transmit circuit (14) comprising a bandpass amplifier (48) for limiting the bandwidth of a transmission burst from the microprocessor (20), which bandpass filter amplifier sends a signal through the switch (12) and on to one of the transducers (6, 8). 10 5. System ifølge et af krav 1-4, kendetegnet ved, at systemet (2) omfatter en transducerforstærker (16), som forbinder transduceren til samme knudepunkt i kredsløbet (14, 16), både under sending og modtagelse.System according to one of Claims 1-4, characterized in that the system (2) comprises a transducer amplifier (16) which connects the transducer to the same node in the circuit (14, 16), both during transmission and reception. 6. System ifølge krav 5, kendetegnet ved, at transducerforstærkeren (16) omfatter en spændingsfølger (50).System according to claim 5, characterized in that the transducer amplifier (16) comprises a voltage follower (50). 7. System ifølge et af krav 1 -6, kendetegnet ved, at systemet (2) omfatter mindst en sender-modtager omskifter (12) omfattende mindst tre kontakter, ved hvilke kontakter 20 ikke anvendte terminaler jordforbindes for reducering af støj.System according to one of claims 1 to 6, characterized in that the system (2) comprises at least one transmitter-receiver switch (12) comprising at least three contacts, at which contacts 20 unused terminals are grounded for noise reduction. 8. System ifølge krav 4, kendetegnet ved, at båndpasforstærkeren (48) omfatter automatisk forstærkningskontrol, hvilken automatiske forstærkningskontrol omfatter en operationsforstærker tilsluttet en variabel feedback modstand, hvilken feedback mod- 25 stand styres af mikroprocessoren.System according to claim 4, characterized in that the bandpass amplifier (48) comprises automatic gain control, which automatic gain control comprises an operational amplifier connected to a variable feedback resistor, which feedback resistor is controlled by the microprocessor. 9. Fremgangsmåde til drift af et system (2) til måling af strømning af fluid eller gas, hvilket system er beskrevet i krav 1-7, hvilken fremgangsmåde angår følgende trin under drift: a: generere mindst en stråle (10) af ultralyd i strømningskanalen (4) ved et sendekredsløb (14) tilsluttet en første sendetransducer (6,8), 30 DK 177567 B1 18 b: modtage ultralydstrålen (10) i strømningskanalen (4) med den anden transducer (6, 8. og modtagekredsløbet (16), c: udføre båndpasfiltrering af det modtagne analogsignal i modtagekredsløbet (16), 5 d: konvertere analogsignalet i et digitalt signal i mikroprocessoren (20) til digitale data samples (24), som repræsenterer i det mindste transittider (26) og tidsforskel (28), e: lagre data samples (24) i en hukommelse (28), 10 f: lagre en vektor (30) af data samples (24) for hver transmissionsretning, g: danne en båndkolonne (32) baseret på vektoren (30) omfattende et passende antal af N signalelementer, 15 h: gange hver værdi af båndkolonnen (32) med et komplekst tal med størrelsen en og en fase, der repræsenterer den sendte frekvens svarende til det sendte signal: γη = χη· eJa,‘ ”=Xn- (cos(<y · ts · n) + j sin (ω ts ·«)) j 64 hvor X„ er den lagrede værdi ved den nte placering, jco vinkelfrekvensen af det sendte 20 signal, ts tidsintervallet for data samplingen, og n antal samples 0 <n<N, i: generere imaginære værdier (34) og reelle værdier (36), hvilke imaginære værdier og reelle værdier lavpasfiltreres i et digitalt filter (38), 25 j: sende det filtrerede digitale signal til en amplitudefunktion (42) og en faseregistreringsfunktion (44), k: beregne strømningen i kanalen på basis af resultatet af amplitudefunktionen (42) og resultatet af faseregistreringsfunktionen (44). 30A method of operating a fluid or gas flow measurement system (2), described in claims 1-7, which relates to the following operating steps: a: generating at least one beam (10) of ultrasound in the flow channel (4) at a transmit circuit (14) connected to a first transmit transducer (6,8), receive the ultrasonic beam (10) in the flow channel (4) with the second transducer (6, 8 and the receive circuit (16) ), c: perform bandpass filtering of the received analog signal in the receiving circuit (16), 5 d: convert the analog signal of a digital signal in the microprocessor (20) to digital data samples (24) representing at least transit times (26) and time difference ( 28), e: storing data samples (24) in a memory (28), 10 f: storing a vector (30) of data samples (24) for each transmission direction, g: forming a band column (32) based on the vector (30) ) comprising an appropriate number of N signal elements, 15 h: times each r value of the band column (32) with a complex number of magnitude one and a phase representing the transmitted frequency corresponding to the transmitted signal: γη = χη · eYa, '”= Xn- (cos (<y · ts · n) + j sin (ω ts · «)) j 64 where X„ is the stored value at the nth location, jco the angular frequency of the transmitted 20 signal, ts the time interval for the data sampling, and n the number of samples 0 <n <N, i: generating imaginary values (34) and real values (36), which imaginary values and real values are low pass filtered in a digital filter (38), 25 j: transmitting the filtered digital signal to an amplitude function (42) and a phase detection function (44), k : calculate the flow in the channel based on the result of the amplitude function (42) and the result of the phase detection function (44). 30