DE4042254A1 - Electromechanical compensation balance with mass-spring-damper system - derives compensation signal from weighing pan deflection with and without article to be weighed - Google Patents

Abstract

The electromechanical compensation balance produces a compensation signal used to counter the deflection of a weighing pan or load carrier resulting from the wt. of weighed material. The compensation signal is thus a direct measure of wt. of the material. The compensation signal is a current through an electromagnetic coil derived electronically from the deflection of the pan. The relative displacement of the vibration deflection of the pan with and without the measurement material is measured. USE/ADVANTAGE - For precision weighing appts. Accuracy of electromagnetic compensation balance is increased or its measurement time is reduced.

Description

The invention relates to an electromechanical compensation balance with an adaptive digital controller to generate the compensation signal, this compensation signal a measure of the goods to be measured is.

In modern weighing technology with electromechanical scales have next to one direct electrical measurement of the deflection of a weighing pan by the weighing sample or a direct electrical measurement of the weight of the weighing material Kompensationsver drive enforced. The weight or mass determination after the electrodynamic Compensation method is based on the deflection caused by a weighing material by bending elements strictly vertically guided load carrier with an electrically generated Repeal counterforce, see for example R.E. Scheidegger, today's state of Precision weighing technology, Feinwerktechnik 74 (1970), page 162 ff. And S. Schmid, Die Meß Principles of Modern Electronic Analytical and Precision Balances, Electrical Engineering 30 (1979) Issue 2, pages 61-63. This counterforce can be, for example, with a plunger coil generate, with the amount of electricity to be proportional to the compensated weight power is. The measured value for the weight is usually a direct measurement and corresponding evaluation of the compensation current obtained, see for example Th. Gast and W. Seifert, a novel self-compensating barless scale, Feinwerktech nik and Meßtechnik 82 (1974) Issue 6, pages 279-284. The deflection becomes, for example measured by means of strain gauges mounted on the bending elements of the balance see, for example, M. Kochsiek and B. Meißner, Weighing Cell Principles, Accuracy, Practical application for verifiable scales, PTB report PTB-MA-4 (1986), Braunschweig, ISSN 0179-0595. The load carrier must usually with a damping element be provided to counteract disturbing vibrations. In one embodiment The analogue signals are used for evaluation, display or further processing of the measured values Compensating currents digitized with analog-to-digital converters, see for example H. Schröder, Microprocessor Assisted Automatic Weighing, Technical Measurement 53 (1986) Heft 4, pages 152-153. A microprocessor usually serves for further processing of the digital Measured values and for controlling the weighing process.  

Now scales are also known in which a measurement can be carried out while the material being measured is still vibrating, d. H. the balance is not that strong damped that the material to be measured with the balance pan is at rest. Here is the electric Measurement signal obtained by electrical filtering, so that only the static signal is evaluated.

The quality, accuracy and measuring time of a balance is by damping the electronic Libra and the accuracy and time constant of the force or wegaufnehmenden Weighing cells determined. The more accurate the load cells, the stronger the damping of the Balance or the electronic filtering of the measuring signal. This usually means a longer one Measuring time or increased electronic or mechanical effort. Since the elek Tronische effort because of the further processing of the measuring signal usually no longer is negligible, this electronic means also for the Gewin tion of the measuring signal to be used, as in the approach in an electronic filtering of the Measuring signal is already performed.

The invention is therefore based on the object, the accuracy of an electromechanical Increase compensation scale or reduce their measurement time, without damaging the damping specify the mechanical load carrier with the sample, so that a measurement at schwin gendem load carrier with the sample is possible.

The invention solves the task in the first line by the displacement of the vibrations of the Load carrier with the material to be measured against the vibrations of the load carrier without the material to be measured is detected. For this purpose, the asymmetry of the two half periods of a full oscillation period with Meßgut compared to the symmetrical vibrations without Meßgut as Determined starting position by the duration of the two half periods are measured. The load carrier performs a controlled continuous oscillation with amplitudes in the micrometer Range that are tracked, for example, with a position difference optical pickup can. Its output signal is digitized and thus the oscillation periods in time measured. From the result of the time difference measurement of the two oscillation half periods a statement about the deflection of the load carrier can be made. In a weighing  The method according to the compensation principle becomes the result of the time difference measurement obtained a control signal, which counteract the displacement of the load carrier by the Meßgut acts. The control signal is so large until the two oscillation half periods of the load carrier are the same size, d. H. Symmetry is present or time difference measurement is zero. at electromechanical compensation scales, the control signal is for example an electrical Electricity that uses a coil to generate a proportional electromagnetic force on the coil Load carrier generated to compensate for the deflection of the load carrier due to the material to be measured ren. The control signal or the compensation current is known to be a direct measure of the Measured material to be measured.

The invention will be described below with reference to the drawings.

Fig. 1 shows the representation of the system parameters of a spring-mass Dämp Fung system (FMD).

Fig. 2 shows the definition of the force introduced into the spring-mass-damping system.

FIG. 3 shows the relationship between the time difference measurement and the mechanical oscillation of the spring-mass-damping system described by y (t).

Fig. 4 shows the dependence of the period of spring constant and mass.

FIG. 5 shows an example of how the spring-mass-damping system is guided into the measurement position and summing up the mechanical vibrations generated thereby at the discontinuities to the resulting amplitude.

Fig. 6 shows the structure and components of an embodiment.

Fig. 7 shows for the embodiment of FIG. 6, the flow chart of the software for a control interval consisting of time measurement and manipulated variable change.

The relationship between the system parameters can be derived with the aid of the illustration in Figure 1. The system-writing differential equation applies to the sum of the forces acting in the system.

y (t) describes the deflection of the system from the support S _A. The substitution y (t) ≡ s (t) -S _{A is used} for s (t) <S _A since the system can not develop its own momentum until s (t) <S _A. S _M is the measuring position determined by the resolver to which the time difference measurement refers. m _n stands for the total mass in the spring-mass-damping system, ie for the net weight of the weighing mechanism and the mass of the sample, which is variable but constant during the measurement; therefore the index n was used here. k _F is the spring constant of the parallel guide, c the damping constant, ω _n the natural frequency of the spring-mass-damping system. The residual travel S _R corresponds to the Federvor voltage in the parallel guide, which offers the opportunity to mechanically compensate for much of the net mass of the weighing mechanism.

The stimulating force and manipulated variable

F (t) = k _mag · i (t) with k _mag = 2πr · N · B (2)

is generated by electromagnetic means. k _mag stands for the parameters of coil and magnet. The time course of F (t) is defined as the sum of non-equidistant ramp functions and is defined for intervals [t _i , t _{i + 1} ] with a length of approximately 1.5 · T _n . The interval length is therefore dependent on the mass m _n present in the system during this interval. T _n is the period of the mechanical oscillation dependent on the mass m _n .

Regarding the regulation to regulate the system due to a time difference measurement in the Kom pensationsfall, corresponding parameters should be introduced equal to the approach for F (t). See also the illustration in FIG. 2.

For each interval [t _i , t _{i + 1} ], the spring-mass-damping system becomes a force change by feedback

fed. t _diff stands for the result of the time difference measurement from the past period T _n and is positive for Δs <0, negative for Δs <0.

A change in the manipulated variable is in each case initiated at time t _i , based on the time difference measurement from the last period T _{n of} the mechanical oscillation. The direction of the ramp-shaped manipulated variable change (or change in force) depends on the sign of the measured time difference, the duration after the value of the measured time difference evaluated with α and the height after the force increase per time determined with F _M / T _M. In this case, F _{M is} the force which is necessary to lift the system from the support S _A to the measurement position S _M , and T _{M is} the time required for this. Decisive is primarily the control factor α, because with it the force increase per interval [t _i , t _{i + 1} ] with the slope F _M / T _{M is determined} quantitatively by feedback from the time difference t _diff ( FIG. 2). The manipulated variable change takes place within a half oscillation of the mechanical oscillations, to which a time difference measurement is immediately made again following. Therefore, the interval length is about 1.5 T _n .

Since s (t) <S _A can only be if F (t) <F _A = k _F · S _A , the substitution (t) ≡ F (t) -k _F S _A for s (t) < S _A makes sense. The bearing force F _A = k _F · S _A corresponds to a force offset that must be implicit in F (t), but has no effect on the dynamics of the spring-mass-damping system for the considered area.

The moment of reaching or exceeding this power offset by the introduced force F (t) thus defines the time t ₀ (zero point) for the substituted system.

For the solution of the differential equation in the time domain one can in very good Write approximation:

The time difference t _diff results from the residence times T 0 and T 1 of the oscillating spring-mass-damping system above or below the measuring position S _M determined by a position sensor and is dependent on the deviation ± Δs relative to the measurement position. FIG. 3 shows in which way T 0 and T 1 are digitized. The times T 0 and T 1 are mapped by means of comparators into square-wave signals X 0 and X 1 , which then serve as gate time. In addition, even with a very small comparator window symmetric to the electrical equivalent u (y (t), S _M ) = 0 detected by S _M when the system is in the immediate vicinity T2 (T2 «T _n ) of the measurement position. This Informa tion is also detected in the form of a square wave X 2 .

One obtains only information about the ratio of Δs / A, of two unknown quantities. It would of course be reasonable to measure the height of the amplitude A; then one could directly calculate the quantity Δs of interest. If one were to do so, one would have to digitize a voltage (or current) with great accuracy. Exactly the necessary effort but should be circumvented with the proposed weighing method. Since the degrees of freedom within the mechanical system are known, the maximum possible mechanical amplitude A can be determined. The measurement of the difference time t _diff is possible as long as the oscillation still touches the measurement position S _M , ie A Δs. Accordingly, a maximum deflection Δs S _M / 2 at an amplitude A <S _M / 2 could still be detected by the time measurement. With a minimum deflection Δs → 0, an amplitude of A <S _{M would} still be permissible. From a mechanical point of view, the amplitude reserve increases inversely proportional to the decreasing deflection Δs.

The system is then capable of measuring, if applicable

The transmission behavior of the spring-mass-damping system according to equation (5) can also be written in the sum form, separated according to the statically acting mean value y _stat and the dynamically acting oscillating component y _dyn for individual discrete t _i . The mechanical oscillation is, so to speak, carried by the introduced force, which determines the mean in interaction with the mass m _n present in the spring-mass-damping system.

y (t _{i + 1} ) = y _stat (t _{i + 1} ) + y _dyn (t _{i + 1} )

= y _stat (t _i ) + Δy _stat (t _i ) + y _dyn (t _i ) + Δy _dyn (t _i ) (7)

The statically acting proportion y _stat of y (t) at a time t _{i + i} is determined only by the force balance:

and thus

Of particular interest, for the reasons mentioned above, is the amplitude behavior of the dynamically acting component y _dyn .

For each interval, two points of discontinuity occur with the time interval α · t _diff , between which the system is supplied by the control with a force change ± ΔF = k _F .Δs at the speed S _M / T _M. The sum of the opposing oscillations arising depends only on α · t _diff . If the compensation case is reached, ie t _diff = 0 and thus Δs (t _i ) / A = 0, one obtains for (9.1):

Here the influences of the damping come into play, which in the Gln. (9.1) and (9.2) are not taken into account. Because during the compensation case no more force changes take place, no new vibrations are generated in the system. This leads, despite very much low attenuation, after several intervals to the loss of mechanical vibration and consequently loss of measurement capability. So it must be ensured that the Oscillation also maintains a minimum amplitude in the long term.  

In general, the problem arises that Δs (t _i ) / A = 1 can also be when Δs → 0 and A → 0. For this case, α → 0 would also have to be used if the deflection ΔS should not increase. For these reasons, it is necessary to counteract the damping effects in order to maintain a minimum amplitude mechanical vibration. If one introduces into the spring-mass-damping system a force change + ΔF, which is immediately reversed afterwards with a force change -ΔF, then the value for y _{stat has} not changed after this process. Due to the vibrations generated at the points of discontinuity, however, the value for y _dyn must have changed.

The amplitude of the generated vibration is superimposed on the existing vibration. In Interaction with the system acting in low damping must be in the system forcibly set a continuous oscillation, in such a way that the losses in the Amplitude by damping just the increase of the amplitude by the new impulse correspond.

The amplitude of the continuous oscillation is thus clearly determined by the Schwingungsregenerie tion and will converge in each case against a limit. By the vibration regeneration on the one hand ensures that the system remains capable of measuring, on the other hand, now a lower limit for A can be selected, which is necessary in terms of the choice of control parameters. The choice of the control parameters depends primarily on the desired measuring range and the vertical degree of freedom, which is determined decisively by the choice of S _M , which determines the maximum possible amplitude.

For 0.1 <α <0.5, an acceptable control behavior can be achieved. It can α within this range, the larger the higher the amplitudes that occur of mechanical vibration. Trials have confirmed these values.

All in all, it makes sense to experimentally determine the most favorable control factor α on the active system. In addition, there are further extensive possibilities of intervention by varying the speed of the force increase over T _M as a function of the current state of the system.

The achievable resolution for the time measurement is determined primarily by the clock rate. For example, a maximum resolution of 1 μs is achieved at a clock of 12 MHz of the microprocessor used. Since the magnitude of the measurable time difference t _diff depends on the ratio Δs / A, two representative average values are given.

m = 20 g ± 0.1 g ⇒t _diff ≈ ± 210 μs

m = 200 g ± 0.1 g ⇒t _diff ≈ ± 70 μs.

Based on the steady state system, a mass change of ± 0.1 g at a sensitivity of 10 μm / g causes a vertical deflection of ± 1 μm. An interesting aspect is the fact that better measurements for t _diff can be achieved for small masses due to the smaller amplitudes than for large masses. In addition, a control cycle is shorter, the smaller the mass in the system (see Fig. 4). Thus, the proposed method for weight determination is particularly suitable for the measurement of small masses.

Using the general form of the compensation weighing system shown above can be realized according to the invention many different embodiments. from Embodiments of the invention will be described below with reference to drawings.

For a constructed weighing system, for example, the following dimensioning can be realized, which then leads to a control behavior, as shown in FIG. 4.

Sensitivity: E = 10 μm / g, Total weight: _mn = 20 g, attenuation: D = 0.00544 (amplitude drop = 5% / 1.5 T _n ), Rise time: T _M = 20 μs, measuring position: S _M = 100 μm, Control Factor: α = 0.5.

The selected values are calculated

Spring constant: k _F = 0.981 N / mm, Angular frequency: ω _n = 221.5 s ^-1 , Period: T _n = 28.4 ms ⇒ f = 35.2 Hz.

A preferred embodiment of the described method for the determination of small masses is particularly suitable to discretize the essential processes as possible and digitize. This means a substantial shift of the measured value determination and Control in a digital computer and its programming. Furthermore, the remaining analog modules are made as simple as possible. Accordingly the sensor used can be designed so that the information supplied by it as simple as possible to be evaluated by a microprocessor. The fact that on the one hand a time measurement can be very accurate without much effort, on the other hand from used microprocessor with on-chip timers is directly feasible, leads in one preferred embodiment, the compensation case in the electrodynamic weight measuring on the basis of time measurements.

If a horizontal position is determined by means of a differential position sensor, with respect to which the mechanical deflection is measured, the symmetry axis coincides the mechanical vibration with the measuring position if and only if the required force bi Lanz in the spring-mass-damping system of the introduced electromagnetic force is shaped. It can be set for this state of compensation case. A change the mass causes a change in the average position of the spring-mass-damping system and as a result, a measurable time difference at the position transducer. The time difference leaves to be used as the basis for the scheme, provided that it is discreet, d. H. interval wise measures and regulates.  

This is done in the described method exclusively in a program (software), which calculates the necessary changes of the exciting force due to the results of Zeitmes solution. The stimulating force is directly proportional to the value of a software counter which counts directly into a D / A converter via a data bus. If the system has adjusted, ie if the time difference is zero, then the state of the software counter is a measure of the mass present in the system. In order to ensure the measuring capability, a Dauerschwin must be excited movement to compensate for the loss of kinetic energy through the even without Dämp tion element still existing low attenuation. This is an essential difference to common methods that are often capable of measuring only with a damping element. Fig. 6 shows the structure of the cooperating hardware and software components as designed for the weighing method of the invention.

In Fig. 7 is a flowchart of the program for successive Regelungsinter valle, consisting of a time measurement and the manipulated variable, reproduced.

Claims (14)

1. Electromechanical compensation balance as a spring-mass damping system in which the compensation signal counteracts the deflection of a balance pan or a load carrier as a result of the weight of a sample and thus is a direct measure of the weight or mass of the sample, and in the Com pensationssignal is a current through an electromagnetic coil and is obtained with measuring means and an electrical control device from the deflection of the weighing pan, characterized in that means are provided for the displacement of the vibration deflection of the load carrier with the Meßgut opposite the vibration deflection of the load carrier without the To be measured. 2. Electromechanical compensation balance according to claim 1, characterized in that the transducers are suitable, the Asymmetry of the two half periods of a full period of Schwingungsaus Steering of the load carrier with the material to be measured against the two symmetrical Half periods of the vibration deflection of the load carrier without Meßgut as off to measure the situation. 3. Electromechanical compensation balance according to claim 2, characterized in that the control device and the elec tromagnetic coil contains electrical means to allow the load carrier with and without Measured material with an electromagnetic force to constant vibration with constant Amplitude is excited, so that the damping of the spring-mass damping system is being compensated. 4. Electromechanical compensation balance according to claim 2, characterized in that it comprises electrical means for controlling the Time difference of the two half periods and thus the asymmetry of the two Half periods of the vibration deflection of the load carrier to measure.   5. Electromechanical compensation balance according to claim 2, characterized in that electrical means are provided, so that the output signal of the measuring means is a digital signal. 6. Electromechanical compensation balance according to claim 2, characterized in that the measuring means is a known per se is opto-electrical position sensor. 7. Electromechanical compensation balance according to claim 2, characterized in that the measuring means is a known per se inductive pickup is. 8. Electromechanical compensation weigher according to claim 2, characterized in that the measuring means is a known per se capacitive transducer is. 9. Electromechanical compensation balance according to claim 2, characterized in that the measuring means known per se Strain gauge measuring bridge is. 10. Electromechanical compensation balance according to claim 6 or 7 or 8 or 9, characterized in that the transducer is a differential Pickup is. 11. Electromechanical compensation scale according to claim 3 and 4, characterized in that the control device has electrical means, so that the increase .DELTA.F of the electromagnetic force for exciting the spring-mass-damping system to continuous vibrations proportional to Zeitdiffe difference t _{diff of} the two half periods of the vibration deflection of the load carrier is. 12. Electromechanical compensation balance according to claim 11, characterized in that the control means electrical means has, so that the force increase .DELTA.F ramped in the direction of the oscillation amplitude at each odd or even odd follow the zero crossing, d. H. at each or an odd subsequent half period, so that the force increase .DELTA.F is generated in alternating direction. 13. Electromechanical compensation balance according to claim 11 or 12, characterized in that the control device consists of a programmed microprocessor exists as a digital controller. 14. Electromechanical compensation balance according to claim 5 and either 7 or 8 or 9 and 13, characterized in that the analog electrical output signal the measuring means with an electronic analog-to-digital converter in a digital electrical signal is implemented.