EP1757809A1 - Electromagnetic metering pump with motion and speed control - Google Patents

Abstract

The pump has a thrust member (20) fixed to a connecting rod (19), where the member is axially movable in a longitudinal axis in a magnet shroud (17). A magnetizing coil magnetizes a magnet in the shroud, for drawing the thrust member and the rod into the shroud against the force of a recuperating spring (23), so as to make an air gap between the member and the shroud small. A control device influences an actual stroke path executed during a forward motion of the thrust member independent of a measured dead region due to the elastic deformation of a diaphragm (13) connected to the rod.

Description

The invention relates to a magnetic metering pump according to the preamble of claim 1.

Such magnetic metering pumps are well known and are adapted by additional devices to the respective requirements. They work according to the volumetric principle, in which the dosing process consists of the transport of a sealed chamber volume. The dosing volume per stroke corresponds to the volume difference of the membrane movement.

In such a Magnetdosierpumpe a movable pressure piece is mounted so that it is drawn in electrical control of the magnetic coil in the magnetic sheath while shortening its air gap and is pushed back after switching off the electrical control by a return spring back to its original position in a fixed magnetic jacket. Fixedly connected to the pressure piece is a push rod which transmits the movement and the force to the metering diaphragm.

In the simple case, the solenoid is turned on to perform a Dosierhubs for a certain time. Other embodiments impose a regulated current profile on the magnetic coil according to a predetermined time profile, as a result of which the magnetic force and thus the dosing output are more reproducible and independent of electrical parameters, such as, for example. the current level of mains voltage.

The stroke frequency is specified by the repetition frequency of the electrical drive pulses. The stroke length may e.g. be changed by a mechanically adjustable spindle, which sets the starting point of the lifting movement; the end position results when the magnet is fully tightened. A possible embodiment is to screw a Hubverstellbolzen accessible from the device operating side knob and scale in a thread of a Hubdeckels, which in turn is attached to the back of the magnetic jacket and whose position is invariable with respect to the magnetic jacket.

The movement of the membrane results from the interaction of the effective forces. After switching on, the magnet current, and thus the generated force, first increases due to self-induction. is the through the diaphragm and the return spring on the push rod overcome the effective force, sets the pressure piece in motion. The air gap decreases as the path progresses, the magnetic force continues to increase accordingly. The result is a fast accelerated movement with hard abutment of the pressure piece on the jacket, only damped by a usually existing damping ring (O-ring). The entire movement takes place within a few milliseconds, resulting in very high instantaneous speeds of the metering medium and high pressure peaks up to twice the working pressure and above.

The membrane is not rigid, but deforms elastically in the flexing area by a certain amount when the pressure of the metering medium acts on them. The amount of this deformation is lost to the effectively executed lifting movement and causes the dosage decreases with increasing working pressure. This falling characteristic is much more pronounced in normal applications than the required dosing accuracy would permit. Magnetic metering pumps therefore usually can not be operated in a general setting over a wide range of working pressure with the desired accuracy; rather, the occurring error is detected by a calibration measurement and included in the further calculations. However, this calibration measurement must be carried out under actual working conditions in the specific application and is, especially in connection with aggressive chemicals, a work step which entails considerable expense.

Although the currently widely used magnetic drive can handle a few simple parts and is therefore inexpensive to manufacture, it remains limited to relatively low powers and has drawbacks with respect to the hydraulic properties of the dosing operation over a motor-driven pump. The motor drive e.g. by means of gear or eccentric is more efficient and has more favorable dosing properties for many processes, but is much more expensive to produce.

The object of the present invention is in particular to eliminate the known disadvantages with respect to the hydraulic properties of the dosing process and thereby to achieve a variable, wider range of application of the magnetic metering pumps, without negatively affecting their advantages, namely the simple and inexpensive production. Furthermore, the movement process of the pressure piece and the associated push rod should be adapted to the desired information that both the dosing itself is adjustable, as well as resulting from manufacturing technology or the property of the elastic membrane errors are taken into account by the scheme and can be corrected. The position sensor should be designed so that manufacturing and / or occurring inaccuracies in the position measurement can be compensated by the electronics used.

The solution of the problem is that with the unit of pressure piece and push rod, a reference element is connected, the position of which is scanned by a position sensor, wherein the position sensor outputs an actual signal which is in a fixed relationship to the position of the reference element and about a control loop as part of its control accuracy, the movement of the unit of pressure piece and push rod so influenced that it follows a predetermined setpoint profile.

With the help of the control device and the position sensor, the movement of the pressure piece is detected with the push rod and guided according to a predetermined motion profile. For this purpose, the control device determines, on the basis of the boundary conditions, the respectively suitable sequence of movements as a default and regulates the actual movement determined by the measured values of the position sensor by influencing the magnetic coil current in such a way that it follows the specification as well as possible, so that otherwise e.g. be eliminated by the properties of the membrane resulting inaccuracies.

If the position sensor operates according to a non-contact principle, a wear-free operation of the sensor is ensured, which is advantageous and ultimately necessary in view of the high number of strokes during the life of a metering pump.

If the position element connected to the push rod is arranged at the end facing away from the dosing head and outside the dosing head, greater flexibility is achieved with respect to the mounting space for the position sensor.

If the reference element influences the beam path of a light source and the position sensor cooperating with it, which is arranged on the magnet jacket, works according to a photosensitive receiver principle, wear-free operation is ensured on the one hand, which is indispensable in view of the high number of strokes during the lifetime of a metering pump, and the moving parts are scanned without contact. Another advantage of this arrangement is that such a design of a position sensor is in principle insensitive to stray magnetic fields, which are not to be avoided when operating the sensor close to the magnet.

If the reference element is a shadow body or a shadowing contour and if the position sensor interacting with it, which is arranged on the magnet jacket, consists of a row of photosensitive charge-coupled receiver cells, such an arrangement has an optical design Basic important properties that the position sensor must meet. On the one hand, the arrangement works wear-free due to the optical principle of operation and is insensitive to stray magnetic fields, on the other hand, such a trained sensor has virtually no linearity error.

If the position sensor continues to be arranged on its own sensor carrier, which is firmly connected to the magnet sheath, such an arrangement can be pre-assembled and tested as a structural unit and thus facilitates the assembly. If the sensor carrier is designed as a part made of non-conductive plastic, this additionally simplifies the electrical insulation of the sensor components against the magnet sheath.

If the position element, the shadow body or the shadowing contour and the position sensor represent a light barrier-like arrangement and the measured values are fed continuously or intermittently to the control loop, such an arrangement provides the position data to the control loop at a speed that is commensurate with the requirements.

If the position sensor consists of a number of linearly arranged receivers (pixels), preferably 128 pixels, such an arrangement can easily determine the position by counting the shadow boundary between illuminated and unlit cells and achieves a resolution according to the distance of the Cells of the receiver module used.

If the light source is a light-emitting diode (LED) arranged opposite the position sensor so that its light beam is not hindered by the push rod on the direct route to the receiver, this has the advantage that the inexpensive LED has an approximately punctiform spot that is suitable for a high optical resolution is essential, and practically has an almost infinite life. The arrangement opposite the position sensor past the push rod results in a large distance between the light source and the receiver, which makes the projection angle of the relevant light beam relatively independent of the mounting position of the elements.

If the output value of the position sensor (36) is formed by interpolating the brightness values of a plurality of pixels lying in the shadow transition region, a finer resolution is achieved for the output signal of the position sensor than predetermined by the mechanical grid of the cells of the CCD receiver.

If filter measures are used during the processing of the signals of the position sensor, then the interference immunity of the position sensor is improved.

The sensitivity of the position sensor to assembly deviations and mechanical displacements during operation, e.g. due to heating or bearing wear is reduced if zero position errors of the position sensor are eliminated by means of a reference memory or scaling error of the position sensor by approaching one or more reference positions.

If exposure fluctuations of the position sensor are compensated for by control of the light source based on the obtained brightness values of the pixels, this reduces the sensitivity of the position sensor to variations in component parameters.

Compensating for brightness variations between individual pixels by including a reference memory for the sensitivity of each pixel reduces the effects of contamination of the optical receiver.

If the signal read out from the position sensor is further processed in a control device and compared with a setpoint input, the control device influencing the current flow to the magnetic coil and thus bringing about a correction of the motion sequence, this targeted influenceability of the membrane movement can achieve or improve advantageous hydraulic properties of the dosing , eg be used in the slow dosing, the pressure compensation and / or the dosing accuracy in Teilhubbereich.

If the control device alternatively influences the position, the speed or the acceleration of the pressure element via a control device by changing the coil current, the advantages of the respectively more suitable control method can be used specifically to suit the requirements of a specific dosing task. A control of the membrane velocity allows a direct control of the actual flow rate of the dosing medium, which is required, for example, for the slower suction to avoid cavitation. On the other hand, regulation of the diaphragm position makes it possible to control situations near standstill, in which the speed information which is formed by differentiating the path signal becomes very small and can no longer be usefully processed by the control device. The control of the membrane position avoids this difficulty and is advantageous to apply for example in the electronic stroke length limitation or slow dosing. The regulation of the acceleration of the diaphragm is advantageous for easy controllability of the control, since the acceleration of the moving masses represents a direct image of the magnetic force and thus indirectly the magnetic current.

If the control device deliberately lowers the speed of the pressure piece in the suction phase and / or in the pressure phase, pressure losses caused by flow resistance or the occurrence of cavitation are thus counteracted. When dosing high viscosity media, e.g. of lecithin, arise at bottlenecks such as in the valves at high flow velocity high pressure losses. These pressure losses must be applied in the form of an additional force by the drive and can be kept low when using the control of the membrane speed. In addition, flow noise is effectively reduced at reduced flow rate. When dosing slightly outgassing media, e.g. of sodium hypochlorite, especially during the suction at too high flow velocity by falling below the vapor pressure of the metering medium cavitation occurs, which has increased mechanical wear. In a regulation of the membrane velocity in the intake phase and / or in the pressure phase, this is advantageously avoided.

If the desired stroke length is communicated by an operator input of the control device and the control device limits the movement of the pressure piece by correspondingly activating the magnetic coil electronically to the stroke length to be executed, the associated mechanical adjustment elements can basically be dispensed with. If the movement of the pressure piece is limited electronically even at maximum stroke length, without reaching the mechanical stop, in principle, the O-ring for damping can be omitted.

If the detection to which value the stroke adjustment is set, by measurement during the dosage directly above the position sensor, the otherwise necessary additional sensor for the mechanical position of the associated adjustment elements can be omitted.

Limits the controller, the speed of the pressure piece at the beginning and / or end of the printing phase, eg in the first or in the last third of the stroke, by driving the solenoid, so that pressure peaks caused by rapid changes in speed of Dosiermediumsstroms or by hard start of the mechanical stop would occur, can be omitted, otherwise necessary additional resources such as Pulsationsdämpfer omitted.

Limits the control device, the speed of the pressure piece at the end of the printing phase by driving the solenoid, so that the effect of over-promotion is avoided, the dosing accuracy is significantly improved, especially at low back pressure.

Distributes the control device, the forward movement of the pressure member during the printing phase by driving the solenoid to the predetermined by the repetition frequency of Dosierhübe time that the application of the metering medium as evenly as possible, up to very slowly running Dosierhüben of e.g. a few minutes, concentration fluctuations of the dosing medium can be largely avoided.

Calculates the control device in operation at quasi-continuous dosage, ie without significant rest period between suction and the following Dosierhub, the lifting movement to operate with reduced stroke length and increased stroke frequency under approximate retention of the membrane speed in Dosierhub, so that the desired dosing results in time average , And it ends the suction by driving the solenoid before the pressure piece was pushed by the return spring all the way to the front mechanical (rest) stop, so that the movement of the pressure piece takes place only in the region of the stroke, in which the air gap and thus the magnetic current requirement is small, a reduction in the required electrical drive power and the resulting heat loss is achieved in the time average.

The dosing accuracy is improved if, during the initial phase of the controlled forward movement of the pressure piece, either the controller itself or another control unit observes the magnetic flux, closes it to the force curve and thus detects the opening of the outlet valve and, using this observation, the dead zone due to the elastic deformation the membrane is formed, measures and the actually executed stroke by targeted termination of the stroke movement depending on the determined membrane deformation influenced so that caused by the membrane deformation error contribution (relative to the stroke or the metered volume) and thus eliminates the dependence of the dosage of the back pressure is significantly reduced. This improvement is achieved by eliminating the error caused by the elastic deformation of the membrane under the effect of the working pressure, that the amount of this deformation does not contribute to the dosage. Due to the reduced dependence of the dosing amount of the working pressure recalibrations, which are otherwise required for significant change in operating parameters such as the working pressure, omitted. The derivation of the membrane deformation from observation of the magnetic current is advantageous because it represents a good image of the actual power requirement, especially in the case of magnetic metering pumps and can be derived directly from already existing signals of the control device and thus requires no additional metrological effort.

Is influenced by the control device in operation at a reduced stroke length of actually executed stroke during the forward movement of the pressure piece depending on the measured dead zone by the elastic deformation of the membrane by deliberate termination of the stroke, so that caused by the membrane deformation error contribution eliminated and so the linear dependence of Dosing amount of the percentage value of the set stroke length is significantly improved, so this increases the dosing accuracy in this case. This improvement is achieved by eliminating the error caused by the elastic deformation of the diaphragm under the effect of the working pressure, that the amount of this deformation does not contribute to the metering and thus the effective stroke length is not strictly proportional to the mechanically adjusted. For the rest, what has been said in the previous section applies.

Overpressure conditions can be advantageously recognized and limited during the dosing process, when the control device measures the dead zone during the forward movement of the pressure piece, which results from the elastic deformation of the membrane, and can make an estimate of the working pressure based on this measured dead zone and when exceeding a predetermined maximum value of Pressure to avoid further pressure increase adjusts the dosage. The otherwise necessary additional resources such as e.g. Overpressure limiters can be saved as a result, provided that the metering pump is the only pressure-increasing unit in the process.

The heat loss inside the device of the magnetic metering pump is dissipated efficiently when the inside of the housing, including the magnet and the electronics, is cooled. As a result, highly lossy modes such as e.g. allows continuous dosing with slowed membrane movement.

If, for cooling the components arranged in the interior, a ventilator is arranged in the interior, whose air flow forcibly guides the wall of the magnet and / or the windings of the coil as well as the inner wall of the housing of the magnetic metering pump and other components, the heat loss of the magnet or of the components mentioned becomes forwarded to the internal air in a thermally direct way and subsequently to the housing. The forced air flow improves the heat transfer resistances of the respective lossy components and thus reduces their temperature increase compared to the air temperature inside the housing. Due to the more even distribution of heat over the entire surface of the case a larger proportion of the surface contributes to heat dissipation than without forced cooling. The peak temperature of the housing surface and the components inside the pump is thus lower overall than without cooling.

If a part of the air flow is passed over the position sensor for cooling the position sensor, its temperature is maintained substantially at the level of the air temperature inside the housing. Since the position sensor is expediently mounted relatively close to the magnet to avoid measurement errors, he would almost without this measure take the temperature of the magnet, which would be much higher than the general air temperature inside the housing without cooling by a fan, as the magnet by far the represents the largest source of heat loss in the device.

Are associated with the Hubdeckel guide surfaces and / or channels, which direct a portion of the air flow to the position sensor, this facilitates the targeted introduction of the air flow to the position sensor.

If another part of the air flow is directed to the electronics installed in the housing cover, their temperature can be kept substantially at the level of the air temperature inside the housing. Since the built-in electronics cover is also mounted relatively close to the magnet, it would be heated without this measure by the magnet, the temperature without cooling by a fan would be much higher than the general air temperature inside the housing.

If, in the housing, the magnet jacket is arranged free-standing in the interior in such a way that it can be surrounded by an air flow for cooling at its circumference, this facilitates the cooling of the magnet by a fan.

If the coil winding has a reduced number of turns with increased wire cross-section, this allows rapid changes in the coil current, as required for controlling the movement of the magnetic pressure piece.

Fig. 1:

Längsschnitt durch eine Magnetdosierpumpe mit geregeltem Magneten

Fig. 2:

Explosionsdarstellung des Positionssensors (Vergrößerung des Ausschnitts X aus Fig. 1

Fig. 3:

Komponenten des Positionsregelkreises

Fig. 4:

Komponenten des Geschwindigkeitsregelkreises

Fig. 5:

Draufsicht auf den Positionssensor in Achsrichtung

Fig. 6:

Seitenansicht des Positionssensors quer zur Achse

Fig. 7:

Darstellung des Schattenbereichs des Positionssensors

Fig. 8:

Darstellung der Helligkeitswerte der Pixels, wie sie dem tatsächlichen Schattenverlauf entsprechen

Fig. 9:

Darstellung des Abbildungsmaßstabs des Positionssensors aufgrund geometrischer Anordnung

Fig. 10:

Interpolation der Positionsauflösung

Fig. 11:

Darstellung der Berechnungsgrundlage für die Interpolation der Positionsauflösung

Fig. 12:

Darstellung der Dosierleistung in Abhängigkeit von der mechanischen Hublänge und vom Arbeitsdruck

Fig. 13:

Darstellung des Konzepts der Kühlung

Fig. 14:

Oszillogramm eines Dosiervorgangs mit Kavitationsschutz beim Ansaugen

Fig. 15:

Oszillogramm eines Dosiervorgangs ohne Kavitationsschutz

Fig. 16:

Oszillogramm eines Dosiervorgangs mit elektronisch auf 0,9mm begrenzter Hublänge

Fig. 17:

Oszillogramm eines Dosiervorgangs mit gebremstem Anfahren des Endanschlags

Fig. 18:

Oszillogramm eines Dosiervorgangs mit Langsamdosierung

Fig. 19:

Darstellung der Dosierbewegung und des zugehörigen Magnetstrombedarfs bei Langsamdosierung mit Kavitationsschutz beim Ansaugen

Below is an embodiment of the invention with its various applications described in more detail. It shows:

Fig. 1:

Longitudinal section through a magnetic metering pump with controlled magnet

Fig. 2:

Exploded view of the position sensor (enlargement of the detail X from FIG. 1)

3:

Components of the position control loop

4:

Components of the speed control loop

Fig. 5:

Top view of the position sensor in the axial direction

Fig. 6:

Side view of the position sensor across the axis

Fig. 7:

Representation of the shadow area of the position sensor

Fig. 8:

Display of the brightness values of the pixels as they correspond to the actual shadow

Fig. 9:

Representation of the image scale of the position sensor due to geometric arrangement

Fig. 10:

Interpolation of position resolution

Fig. 11:

Presentation of the calculation basis for the interpolation of the position resolution

Fig. 12:

Representation of the dosing rate as a function of the mechanical stroke length and the working pressure

Fig. 13:

Presentation of the concept of cooling

Fig. 14:

Oscillogram of a dosing process with cavitation protection during suction

Fig. 15:

Oscillogram of a dosing process without cavitation protection

Fig. 16:

Oscillogram of a dosing process with electronically limited 0.9 mm stroke length

Fig. 17:

Oscillogram of a dosing process with braked start of the end stop

Fig. 18:

Oscillogram of dosing with slow dosing

Fig. 19:

Representation of the dosing movement and the associated magnetic current requirement during slow dosing with cavitation protection during priming

Fig. 1 shows a longitudinal section through a Magnetdosierpumpe (short MD called). A housing 1, which is provided in the region of the magnet (upper outer side) as contact protection against the hot surface with housing ribs 3, goes on the bottom in a bottom plate 4 with mounting holes. As is generally known, the magnetic jacket 17 of the drive magnet is arranged in the upper region of the housing 1. The one end face of the housing is enclosed by a Gehäusedekkel 5, which is placed on the housing 1 and connected thereto. In the center of the housing cover 5 and the same axis to the longitudinal axis 18 of the generally rotationally symmetrical magnet is a manually operable adjusting 7 for the adjustment of Hubverstellbolzens 8 integrated into the housing cover, which limits the axial movement of the pressure member 20 and thus the stroke of the diaphragm pump. The adjusting member 7 and other controls are protected by a cover 9. Below the cover 9 connections for the control lines 10 and for the power supply 11 are provided. On the opposite side of the cover, a dosing head 12 is arranged, in which a membrane, for example, made of plastic 13 is firmly clamped at its periphery. The dosing head 12 also carries an inlet valve 14 and an outlet valve 15 to the between the diaphragm 13 and dosing head 12 in the dosing Press 16 via the inlet valve 14 sucked dosing via the outlet valve 15 in the dosing. The Magnetdosierpumpe works on the volumetric principle, ie a predetermined volume to be sucked in each stroke on the one hand and on the other hand discharged via the exhaust valve 15. The membrane 13 is placed on the drive for this purpose in an oscillating motion. As the drive for the diaphragm 13, as the name "Magnetdosierpumpe" says, an electromagnet, formed by a rotationally symmetrical magnetic jacket 17, in which a rotationally symmetrical magnetic coil 2 is integrated. The magnet coil 2 is formed by a likewise rotationally symmetrical coil carrier 51 made of plastic, which is wound with a coil winding 29, consisting of a plurality of turns made of enameled copper wire. For example, the magnetic coil comprises 800 turns with a wire diameter of about 1 mm. The coil carrier and the winding are designed according to the requirements of the working voltage and can be additionally isolated by further insulation means such as foils. The magnetic shell 17, a rotationally symmetrical solid body encloses, together with the magnetic disk 25, which closes the magnetic circuit from the magnetic shell 17 to the pressure piece 20 back, the pressure piece 20 with the arranged in the center of the pressure piece push rod 19 which is axially displaceable together with the pressure piece. Towards the Hubverstellbolzens 8 towards the push rod 19 cooperates with the adjusting member 7 as manually adjustable Hubverstellvorrichtung. With the opposite end, the push rod 19 cooperates with the elastic membrane 13. On the Hubverstellbolzen 8 facing part of the push rod 19, the pressure piece 20 is fixedly connected to the push rod. On the part of the push rod 19 facing the dosing head 12, the core 30 of the diaphragm 13 is firmly connected to the push rod. The push rod 19 with pressure piece 20 is mounted axially displaceably in a sleeve 26 arranged in the center of the magnetic jacket 17. In the direction of the pressure piece 20 facing inner end face 24 of the magnetic shell, an O-ring 21 is arranged, which dampens the stop of the inner end face 22 of the pressure piece against the opposite inner end face 24 of the magnetic shell, if necessary. Within the inner end face 24 of the magnetic jacket, a compression spring 23, for example a spiral spring, is arranged in a hole facing the inner end face 22 of the pressure piece and holds the pressure piece away from the inner end face of the magnetic jacket 24 when the magnet is not activated an air gap is created between the two end faces. On the Hubverstellbolzen 8 side facing the magnetic jacket carries a magnetic disk 25 which is fixedly connected to the magnetic shell by, for example, screws or press-fitting and which closes the magnetic circuit from the magnetic jacket to the pressure piece out. The outer surface of the rotationally symmetrical pressure piece is mounted axially displaceably in the magnetic disk 25 in a further bushing 27. On the side of the adjusting device for supporting the Hubverstellbolzens 8 a Hubdeckel 28 is attached to the magnet jacket, which is shaped so that it keeps enough distance on the one hand to the magnet jacket and the pressure piece, not to hinder the movement of the pressure piece and on the other hand by the fan 43rd caused air flow directed specifically to the position sensor 36.

Adjustment, Hubverstellbolzen and push rod are arranged coaxially with the longitudinal axis 18. Now, if the solenoid coil 2 is energized, the pressure piece 20 is moved to the compression spring, wherein the air gap narrows, at the same time the membrane is pressed into the dosing, with the result that in the dosing an overpressure arises, the exhaust valve 15, eg a spring-loaded ball valve, opens and the dosing medium is pressed into the dosing. Now, when the magnet is deactivated, the pressure piece by the compressed compression spring 23, the e.g. may be formed as a spiral spring moves in the opposite direction to Hubverstellbolzen 8, with the result that the connected to the diaphragm 13 push rod 19 entrains the membrane in their movement, whereby in the dosing chamber 16, a negative pressure is created, which opens the inlet valve 14, so that once more dosing can be sucked into the dosing. Due to the alternating, oscillating movement of the diaphragm by means of the magnetic drive, the flow of the metering medium in the metering line is formed.

The position of the unit of push rod 19, pressure piece 20 and diaphragm 13 is scanned by the position sensor 36, the measurement signal is in a defined relationship to this position; this relationship may be considered as possible execution e.g. be strictly proportional. The measurement signal of the position sensor 36 always refers to the position of the part of the movable unit, where it acts. This point of attack is formed by the reference element, which in this context is to be understood in an abstract sense. Depending on the requirements of the position sensor, it may be designed as a concrete component to be mounted in addition, but also only from a characteristic embodiment, e.g. an edge or surface on any of the components required anyway, e.g. on the pressure piece 20, exist.

In the exemplary embodiment, a sensor carrier 31 is attached to the magnetic jacket 17 (see also the schematic representation in Fig. 6), on the one hand in longitudinal direction photosensitive CCD cells 32 (CCD = charge coupled device) and opposite a light source 33, e.g. a light emitting diode (LED), carries.

The sensor carrier 31 connected to the magnetic jacket has a central opening 34, which is penetrated by the push rod 19. On the sensor support 31 penetrating the push rod part is arranged as a reference element, a shadow body 35 fixed. In the oscillating movement of the push rod 19 so the shadow body 35 is taken and passes over the photosensitive cells 32 touchless. As now esp. Fig. 5, which shows a view in the axial direction, the light source must be arranged so that the light beam on its way is not covered by the push rod 19 to the photosensitive cells; ie, that the light source 33 is disposed above or below the push rod 19 and the line of photosensitive Cells are opposite to the height of the axis of the push rod. As now shown in particular in Fig. 7, is cast by the light source by means of the shadow body 35 on the photosensitive cells 32, a shadow, which divides the cells in principle in light (h) and not illuminated (d) cells. Since the series of light-sensitive cells arranged parallel to the longitudinal axis 18, for example 128 pixels covering a distance of approximately 8 mm, is only partially exposed or shaded in the border area, the transitional situation of the shadow profile SV is shown in FIG. The height of the rectangular areas shown in FIG. 8 represents the illuminance of the respective pixels. By means of a special method, which will be described later in detail and shown in FIG. 10, this limit situation is used to determine the respective position of the shadow body and thus To determine the position of the push rod and thus the diaphragm exactly. This measuring device, consisting of push rod-side shadow body and sensor carrier side light-sensitive CCD cells with opposite light source, serves to measure the actual position or the speed of the oscillating push rod and to use this information for the regulation of the movement.

The push rod, which causes the diaphragm to oscillate, travels a distance each stroke equaling the mechanical stroke length. To take into account mounting tolerances, the longitudinal extent of the photosensitive CCD cells must be slightly larger. In principle, this also applies to any other conceivable position sensor used.

x_s:

Sollwert der Position des Druckstücks

x_l:

Istwert der Position des Druckstücks

x_Sl:

Regelabweichung der Position des Druckstücks

v_s:

Sollwert der Geschwindigkeit des Druckstücks

v_l:

Istwert der Geschwindigkeit des Druckstücks

v_Sl:

Regelabweichung der Geschwindigkeit des Druckstücks

SG:

Stellgröße

KSG:

Korrigierte Stellgröße

I_M:

Magnetstrom

For the control loop formed by the sensor and the control device, as shown in particular in FIGS. 3 and 4, the following mechanical and electronic components are required. The short names contained in the two diagrams mean:

x _s:

Setpoint of the position of the pressure piece

x _l :

Actual value of the position of the pressure piece

_xsl :

Deviation of the position of the pressure piece

v _s :

Setpoint of the speed of the pressure piece

v _l :

Actual value of the speed of the pressure piece

v _S1 :

Control deviation of the speed of the pressure piece

SG:

manipulated variable

KSG:

Corrected manipulated variable

I _M :

magnet power

The fixed part of the magnetic drive consists of the magnetic jacket 17 with magnetic coil 2 and the magnetic disk 25 each with inserted plain bearing bushes 26 and 27 for the unit of pressure piece 20 and push rod 19. The movable part of the magnetic drive whose movement is to be regulated, consists of Push rod 19, with the pressure piece 20 as the drive element and the diaphragm core 30 are firmly connected. The return spring 23 retrieves the pressure piece after a successful stroke and thus causes the suction. The outer ring of the membrane 13 is fixedly mounted in the dosing head 12, the metal membrane core 30 injected in the membrane moves the central surface of the membrane as a displacement element in the dosing head. The inlet valve 14 closes on the suction side, the outlet valve 15 on the pressure side of the dosing and offers each a connection option for the outer casing. With the push rod 19 is e.g. connected at the end remote from the dosing head, a reference element whose position is scanned by a non-contact in the present case position sensor 36. In the exemplary embodiment, the reference element is a shadow body 35 in the form of a disc and the position sensor is a light barrier-like arrangement consisting of the previously described light source 33 in cooperation with the series of photosensitive cells 32, which detects the position of the disc optically and thus without contact by their shadowing.

The position sensor 36 outputs an actual signal x ₁ which is proportional to the position of the reference element 35. In the case of the speed controller this is derived in the embodiment by a differentiator 37 after the time (dx _l / dt) and thus in addition a speed-proportional actual signal v _l formed. For the control, of course, other methods are suitable, which provide a proportional to the membrane velocity signal. Depending on the type of regulation and requirements of the dosage, a time profile for the desired value 38 of the position x _S or the speed v _{S is} specified. By means of a desired-actual comparison 39, the control deviation is determined as position deviation x _Sl = (x _S -x _l ) or velocity deviation v _Sl = (v _{s -v l} ), and the result is applied to a PID controller 40 (FIG. PID controller = controller with proportional, integral and differential components). Its output, the manipulated variable SG, corresponds to a request value for the magnet current. Prior to further processing, a position correction 41 takes into account the fact that as the position advances, the magnet requires less and less current for a required force, thereby improving the stability of the regulator. The position correction 41 consists in subtracting a position-proportional component from the output signal of the PID controller 40 and results in a corrected manipulated variable KSG. An amplifier 42 incorporates the power switching stages and drives the solenoid 2 with the desired current. The amount of the position-dependent current correction, the conversion of the current setpoint into a specific magnet current and possibly the derivative constant for the formation of the velocity signal v _l are determined by the three proportionality factors k1, k2, k3. The factor for the position-dependent current correction k1 should be chosen so that the degree of current reduction is as close as possible to the course of the magnetic characteristic The two factors k2 for the power amplifier and k3 for the derivation of the speed signal can be chosen on the basis of practical considerations, such as working with the most easily manageable value ranges of the associated quantities.

In Fig. 3, the control loop for a position controller, in Fig. 4, the control loop is shown schematically when used as a speed controller. The described control circuit sets the predetermined time profile for the desired value of the position x _s or the speed v _s , of course within the scope of its possible control accuracy.

The specification of the specific profile for the position, the speed or the acceleration and the switching between these modes is done on the basis of the requirements arising from the functions described below, for example, taking into account the functional limits of the controller such as control speed, achievable accuracy, etc.

With such a control, it is possible with a magnetic metering pump to specify a desired speed of the diaphragm 13 and thus to control the effective flow rate of the metering medium.

Likewise, the membrane position can be controlled directly. This function makes it possible to selectively approach specific positions in selected dosing phases and, if necessary, to maintain them at standstill.

By regulating the sequence of movements by means of a position sensor, in contrast to unregulated operation, it is possible to react to changes in operating variables which occur over time or are caused by environmental conditions or specimen scatters, ie statistical deviations within the production series, and whose harmful influence is minimized , Examples which may be mentioned are the membrane rigidity or the viscosity of the metering medium. Both require a proportion of magnetic force which must be applied in addition to the force created by the action of the working pressure on the membrane surface. These disturbances can be compensated by detecting their effect and readjusting the magnetic current. With an unregulated dosing pump with a predetermined solenoid current, even if this is kept stable even by means of control, such disturbances remain unconsidered.

In addition, it is possible by controlling the movement by means of a position sensor, in contrast to the spontaneous dosing process in unregulated operation on to react to internal and external factors, which are described below, and to ensure operating conditions, with the help of which selected hydraulic properties of the dosage can be specifically caused or avoided. As an example, reference is made to the function described below of the protection against cavitation during suction.

For example, individual applications of a magnetic metering pump of the previously described type, which has a position sensor and influences the course of movement of the membrane by means of a regulation and change of the magnetic coil current, will be explained below.

In describing these possible applications, FIGS. 14 to 19 show, by way of example, oscillograms of the respective metering process. In the diagrams, the upper curve Pos in each case represents the membrane movement with a scaling of 0.5 mm / div. group; the end stop EPos is at the top of the diagram. The rising part of the curve Pos corresponds to the metering stroke, the falling part to the suction. The lower curve l _M shows the associated magnetic current with a scaling of 1A / div .; the zero line l _Mo is at the bottom of the diagram. The designations "Pos", "EPos", "l _M " and "l _Mo " are shown by way of example in FIG. 14 and apply mutatis mutandis to the following diagrams in FIGS. 15 to 19, without being expressly mentioned there again to be.

Avoidance of flow losses in highly viscous media

The function of controlling the velocity of the membrane 13 can be used, in particular, for high viscosity media (e.g., lecithin) to limit flow losses in valves and other bottlenecks. High flow rates have a negative influence on the dosing accuracy in such media by additional pressure losses due to flow resistance. In addition, it is advantageous here if the limited speed allows more time for the defined opening and closing of the valves. Overall, both effects improve the dosing accuracy of highly viscous media.

To achieve this, the membrane velocity is kept limited to a selectable maximum value during the entire dosing process. This maximum speed depends, inter alia, on the viscosity of the specific medium to be metered and is to be selected by the operator or specified directly in the form of several predefined values adapted to common applications.

cavitation

In the case of slightly outgassing media (such as, for example, sodium hypochlorite), it is possible, in particular during suction, but also in the metering stroke, if the flow velocity at the bottleneck is too low due to local undershooting of the vapor pressure, etc. Depending on the chemical composition of the dosing medium and its temperature depends, cavitation occur, which has increased wear. Cavitation can be avoided by limiting the speed by regulation to values well below a critical flow velocity even during the suction, ie the return of the membrane 13. For this purpose, a braking force of the magnet is counteracted by the regulation during the return movement of the force of the return spring 23, and the membrane velocity corresponding to the medium velocity is limited to, for example, 1 mm / 50 ms.

14 shows by way of example the oscillogram of a metering process with a stroke duration of 400 ms, a stroke length of 2 mm and nominal working pressure of 10 bar with activated cavitation protection during suction.

FIG. 15 shows the oscillogram of a dosing process with self-released suction, given otherwise identical settings.

During suction, in Fig. 14, the velocity is limited to a value of about 1 mm / 50 ms by appropriately driving the solenoid 2. the control device prevents the membrane driven by the return spring 23 can move back faster than the said speed; the diagram shows the magnetic flux flow during the suction phase, which ensures this. In Fig. 15, the driving of the magnet is omitted during the suction phase, here takes place during which no Magnetstromfluß instead. It results in phases a significantly higher speed, which may already result in cavitation.

Electronic stroke length adjustment

The invention makes it possible to save the mechanical device for adjusting the stroke length (adjusting member 7 and Hubverstellbolzen 8). For this purpose, the control device, the desired stroke length electronically, eg by an operator input, communicated. If the desired stroke length has been carried out, the reached position of the diaphragm 13 is held electronically and this is subsequently retracted for aspiration. The membrane can remain in the position corresponding to the desired stroke length for a short time in order to give the exhaust valve 15 sufficient time to close, or to retract immediately after the desired stroke length has been carried out.

16 shows by way of example the oscillogram of a metering process with a stroke duration of 400 ms and nominal working pressure of 10 bar with an electronically limited stroke length of 0.9 mm. As can be seen from the diagram, the membrane does not fully move to the end stop at the top of the diagram, but stops after 0.9mm of movement and then carries out the suction process.

Recording the position of the stroke length adjustment controller

Dosing pumps according to the prior art often provide a mode in which the dosing strokes carried out over the set volume of the displacement chamber (stroke length) are converted directly into a metered total volume and this. is displayed as volume flow in the unit l / h. For such functions, knowledge of the stroke length set by the operator is required since this depends on the volume metered per stroke. The position of the Hubverstelleinrichtung must be converted for this purpose in metering pumps of previous design by a separate sensor into an electrical signal and read into the controller. An example of a practical realization would be a rotary encoder on Hubverstellorgan.

A motion-controlled metering pump does not need an additional sensor, as it can detect the actually worn membrane path during the stroke using the integrated position sensor. By subtraction of the two position values in the end positions, which can be measured after reaching the mechanical stop, as soon as the movement has come to a halt, the stroke length can be calculated directly and is available for further processing.

Avoiding pressure peaks

In a magnetic metering pump of previous design, the metering medium is opened relatively abruptly when opening the exhaust valve, which has a high acceleration of the metering and a corresponding pressure peak result. Furthermore, the membrane and thus also the metering medium moves very fast in a magnetic metering pump of previous design, particularly in the last part of the lifting movement due to the decreasing air gap, which is associated with a hard abutment of the pressure piece and high instantaneous flow velocities or pressure peaks.

A motion controlled solenoid metering pump as described can avoid these negative effects by increasing the speed until the exhaust valve opens and just before it reaches the end stop deliberately lowered and the pressure piece dosed on the last stretch just before the stop is braked. In a modification, it is also possible not to start the attack, but to terminate the membrane movement shortly before reaching the stop deliberately. Thus, for example, the O-ring 21 omitted or be dimensioned much smaller. In addition, this significantly reduces the running noise.

17 shows by way of example the oscillogram of a dosing process with a stroke duration of 400 ms, a stroke length of 2 mm and nominal working pressure of 10 bar with a braked approach of the end stop. As the diagram shows, the membrane velocity is reduced to a value of approx. 0.6mm / 50ms before reaching the end stop at the top of the diagram.

Avoid over-promotion

In the case of a magnetic metering pump of the previous version, the so-called overfeeding occurs with very little back pressure. It arises from the fact that at the end of Dosierhubs the exhaust valve does not close immediately, but the metering medium in a kind of lifting effect through its high speed in conjunction with its inertia the dosing further flows through it opens the intake valve prematurely, so that an excess amount of dosing in the output line arrives. Due to the over-promotion unregulated pumps are only from a minimum working pressure of e.g. 2-3bar useful to use; To ensure this, usually a so-called. Pressure-holding valve is inserted into the outgoing dosing.

In a motion-controlled Magnetdosierpumpe can be almost completely avoided by electronically limiting the membrane velocity shortly before reaching the end stop or even during the entire Dosierhubs responsible for the over-promotion lifting effect. The working range of the metering pump expands so much to small working pressures out that a pressure relief valve can be omitted in many practically occurring Dosiersituationen.

The movement corresponds to that previously shown in Fig. 17 with the difference that it refers to a situation at a particularly low working pressure.

Slow dosing to avoid concentration fluctuations

For certain applications in which good mixing with a process medium flow is important, the most uniform possible introduction of the dosing medium in the process is required.

In the case of a motion-controlled magnetic metering pump, the available time, which results from the repetition frequency of the metering strokes, can be divided so that the portion remaining after deduction of the suction duration is utilized to a maximum for the forward movement, except for a short rest phase. The speed to be controlled is calculated from the distance to be traveled (set stroke length) and the available time. The degree of utilization of the available time depends on the one hand on the requirements of the dosing application, on the other hand on the possibilities of the cooling concept, which must dissipate the increased thermal power loss due to the almost uninterrupted magnetic control.

FIG. 18 shows, by way of example, the oscillogram of a metering operation with a stroke duration of 500 ms, a stroke length of 2 mm and nominal working pressure of 10 bar in the slow metering operating mode, here combined with slowed suction for protection against cavitation. As the diagram shows, the total stroke time is changed from 500ms to a compression stroke of about 250ms and a suction of about 180ms, which together account for 430ms or 86% of the total stroke time; the remaining 70ms are used to define the movement phases.

Certain applications require the possibility to dispense smallest subsets as evenly as possible over a very long time, thus achieving a quasi-continuous dosage. For these cases, motor pumps are used in the prior art, e.g. work with a stepper motor and a self-locking gearbox. A total stroke is performed reduced speed at these metering pumps or divided into several sub-steps with intermediate rest periods, at the end of the total stroke a complete (fast) suction phase is performed, and then continued the dosing in the manner described.

The invention makes it possible to meet these requirements with the simpler and thus more cost-effective design of a magnetic metering pump. For this purpose, the membrane 13 must be guided in controlled operation at very low speed along the stroke, wherein at the stroke end also a complete suction at normal speed is performed, so that the total stroke time can be almost completely used for the pressure stroke. The speed can be in a very wide range of e.g. 1 mm / min to 1 mm / s and beyond.

One possible embodiment may insert small rest periods between partial movements in which the membrane 13 is held in a constant position. This allows the exhaust valve 15 clearly defined states that are not at extremely slow, the stoppage approaching movement are given more, resulting in high demands on the exhaust valve 15. The thermal load differs in this embodiment compared to the linear moving version practically not, since in both cases the working pressure a quasi-static magnetic force must be opposed.

Another possible embodiment can reduce the thermal load by the lifting movement is divided into small part movements as described above and in the intermediate stoppage phases, the membrane 13 is additionally reduced by a small discharge path to a pressure relief by clearly closing the exhaust valve 15 and concomitantly to achieve a reduced magnetic power demand during the stoppage phases. The partial strokes are then each to supplement this discharge path, so that overall an unchanged stroke is covered. The unloading path must be smaller than the (pressure-dependent) deformation path of the diaphragm to avoid partial suction during retraction between the sub-strokes, thus degrading accuracy. This embodiment works advantageously in conjunction with the pressure compensation described below, since in this case the deformation of the membrane is measured during operation and the discharge path can be better adapted to the actual conditions.

pressure compensation

In controlled motion, at steady state (i.e., steady state), the controller adjusts a solenoid current at all times that just covers the (time-varying) external forces.

This magnetic power requirement results on the one hand from the currently applied force and on the other hand from the resulting residual membrane gap between each residual air gap between the inner end face of the pressure piece 22 and the inner end face of the magnetic shell 24. It results in a characteristic current flow l _M during Dosierhubs, as he in particular in Fig. 19 is shown. The oscillogram shown there shows an example of the current profile with a stroke distributed over approx. 2.0 s over a stroke length of 2 mm and nominal working pressure of 10 bar. At the end of the stroke, a slower aspiration to protect against cavitation takes place, but this is not important for the following consideration. The time scale of the diagram was adapted to the slower stroke.

• Der erste schnelle Stromanstieg (im Beispieldiagramm der Zeitbereich 0 bis 80ms) wird verursacht durch das induktive Verhalten der Magnetspule 2, welches keine Stromänderung in Nullzeit zulässt, sowie die Geschwindigkeit der Regeleinrichtung, die sich zunächst auf die geforderte Bewegung einstellen muss. Der ansteigende Strom lässt die Magnetkraft ansteigen, bis die äußeren Kräfte überwunden werden und sich das Druckstück 20 zusammen mit der Membran 13 in Bewegung setzt. In dieser Phase wird in erster Linie das Magnetfeld aufgebaut.
• Der quasi lineare Stromanstieg nach dem ersten Regeleinschwingvorgang bis zum eigentlichen Strommaximum (im Beispieldiagramm der Zeitbereich 80ms bis 400ms) lässt auf einen ansteigenden Kraftbedarf schließen, da der Magnetstrom bei konstanter Kraft und abnehmendem Luftspalt gleichfalls abnehmen müsste. In dieser Phase wird im Dosierraum 16 bei noch geschlossenem Auslaßventil 15 der Innendruck kontinuierlich erhöht, indem die Membran 13 eine zunehmende Kraft ausübt und sich dabei elastisch verformt. Bei diesem Vorgang bewegt sich der Membrankern 30 in den Dosierraum 16 hinein, baut Druck auf, und der elastische Walkbereich der Membran weicht in gleichem Maß dem Druck nachgebend gegenläufig zur Bewegung des Membrankerns zurück. Die Membran 13 verformt sich in sich selbst, in der Summe findet aber so gut wie keine Volumenänderung statt, was auf die Tatsache zurückzuführen ist, daß das Dosiermedium praktisch nicht komprimierbar ist und zu diesem Zeitpunkt beide Ventile geschlossen sind. Am Ende dieser Phase entspricht der Kammerdruck dem äußeren Arbeitsdruck. Der bis hierhin zurückgelegte Weg entspricht dem Betrag der Membranverformung, also dem Totbereich zu Beginn der Dosierung, und trägt praktisch nicht zur Dosierung bei. Die aktuelle Position wird gespeichert und als gemessene Verformung im weiteren Verlauf des Dosiervorgangs berücksichtigt (im Beispieldiagramm ist der Totbereich 0,3mm).
• Am Punkt des Druckgleichgewichts öffnet das druckseitige Auslaßventil 15. Nun ist der auf die Membran 13 wirkende Druck praktisch identisch mit dem äußeren Arbeitsdruck und erhöht sich nicht weiter, als Folge bildet der Magnetstrom eine konstant wirkende Kraft bei abnehmendem Restluftspalt ab und fällt bei weiter fortschreitender Bewegung kontinuierlich (im Beispieldiagramm der Zeitbereich ab 400ms). Da die Strömungsgeschwindigkeit des Dosiermediums durch Anwendung der beschriebenen Verfahren vernachlässigbar klein bleibt, entstehen keine nennenswerten Druckschwankungen, so dass weiterhin aus dem Stromverlauf auf die Magnetkraft geschlossen werden kann (siehe Fig. 19).
• Der Verlauf des Magnetstroms nach Erreichen des Druckgleichgewichts und Öffnen des Auslaßventils 15 ist für die hier beschriebene Messung der Membranverformung nicht weiter relevant. Bei der praktischen Durchführung der Messung der Membranverformung kann daher beispielsweise zunächst eine lineare Vorwärtsbewegung mit einem Geschwindigkeitssollwert geregelt werden, welcher für das Ausmessen des Strommaximums optimiert ist, und unmittelbar nach dem Erfassen und Abspeichern der Membranverformung auf einen abweichenden Bewegungsablauf umgeschaltet werden, der nach den Erfordernissen einer der übrigen beschriebenen Funktionen gestaltet ist. Beispielsweise kann so zu Beginn des Hubs innerhalb einer relativ kurzen Zeitspanne die Membranverformung gemessen und danach der eigentliche Dosierhub über die restliche zur Verfügung stehende Zeit als Langsamdosierung ausgeführt werden.

The lower curve shows a l _M initially relatively steep current increase until the membrane 13 starts to move. After a brief overshoot, the current continues to increase as the movement progresses, until a current maximum is reached. From this point, the current drops approximately linearly for the remainder of the way until the end stop EPos is reached. In the suction phase, a further current flow prevents the membrane from running back too fast as protection against cavitation. The following conclusions can be drawn from this characteristic behavior:

• The first fast current increase (in the example diagram of the time range 0 to 80ms) is caused by the inductive behavior of the solenoid 2, which does not allow zero current change, as well as the speed of the control device, which must first adjust to the required movement. The rising current causes the magnetic force to increase until the external forces are overcome and the pressure piece 20 moves together with the diaphragm 13 in motion. In this phase, the magnetic field is built up in the first place.
• The quasi-linear current increase after the first Regeleinschwingvorgang up to the actual maximum current (in the example diagram of the time range 80ms to 400ms) suggests an increasing power requirement, since the magnetic current at a constant force and decreasing air gap would also decrease. In this phase, the inner pressure is continuously increased in the metering chamber 16 with the exhaust valve 15 still closed, by the membrane 13 exerts an increasing force and thereby deforms elastically. In this process, the membrane core 30 moves into the metering chamber 16, builds up pressure, and the elastic Walk region of the membrane deviates to the same extent the pressure yielding in opposite directions to the movement of the membrane core back. The membrane 13 deforms in itself, but in total there is virtually no change in volume, which is due to the fact that the metering medium is virtually incompressible and at this time both valves are closed. At the end of this phase, the chamber pressure corresponds to the external working pressure. The distance traveled so far corresponds to the amount of membrane deformation, ie the dead zone at the beginning of the dosage, and practically does not contribute to the dosage. The current position is saved and taken into account as measured deformation in the course of the dosing process (in the example diagram the dead band is 0.3 mm).
• At the point of pressure equilibrium opens the pressure-side outlet valve 15. Now, the pressure acting on the diaphragm 13 is virtually identical to the external working pressure and does not increase further, as a result, the magnetic current forms a constant force with decreasing residual air gap and falls as the movement continues continuous (in the example diagram the time range from 400ms). Since the flow rate of the dosing medium remains negligibly small by using the described method, no appreciable pressure fluctuations occur, so that it is still possible to deduce the magnetic force from the course of the current (see FIG. 19).
• The course of the magnetic current after reaching the pressure equilibrium and opening the outlet valve 15 is not relevant for the measurement of membrane deformation described here. In the practical implementation of the measurement of the membrane deformation therefore, for example, first a linear forward movement can be controlled with a speed setpoint, which is optimized for measuring the maximum current, and are switched immediately after the detection and storage of the membrane deformation to a different course of motion, according to the requirements one of the other functions described is designed. For example, at the beginning of the stroke within a relatively short period of time, the membrane deformation can be measured and then the actual metering stroke can be carried out as a slow metering over the remaining available time.

The membrane deformation measured by observation of the magnetic current profile can now be used as the basis for a correction of the mechanical stroke length HL and included in the membrane path to be carried out. For this purpose, the point of the current maximum is set as the actual starting point of the dosage, from which the desired stroke length is executed and the stroke is then stopped, even before the mechanical end stop is reached by impingement of the pressure piece 20 on the inner end face of the magnetic jacket 24. At working pressures below the rated pressure, the membrane deformation will be less and the last part of the possible mechanical path of the pressure pad will remain unused, i. the air gap is not completely closed.

The deformation of the membrane is i.a. Depending on material properties and can therefore change due to aging or will be subject to specimen spreads. These two aspects are taken into account by no predefined, derived from component parameters value is used for the correction of the membrane deformation, but at each stroke the specific conditions are re-measured by measurement.

The magnetic current can be detected by measurement, but this is not absolutely necessary. Since the amplifier 42 converts the corrected manipulated variable KSG as a magnetic current input according to the factor k2 into a magnetic coil current I _M , the corrected manipulated variable KSG can be used directly as an image of the magnetic current, whereby this is used for further processing without additional measuring effort can be derived from already existing signals of the control device.

Improvement of dosing accuracy in partial lift operation

The above-described method for determining the deformation of the diaphragm due to its elasticity by observing the magnetic current waveform also enables an improvement in the accuracy of the partial lift operation.

In a magnetic metering pump according to the prior art without compensation of the membrane deformation, the metering is not only pressure-dependent, but also in Teilhubbetrieb not strictly proportional to the set mechanical stroke length. Rather, the effective metering on stroke only begins after an initial dead zone from the point of complete membrane deformation. If one plots a characteristic which shows the metering power as a function of the set mechanical stroke length, a linearly rising curve results, which only after a minimum stroke length corresponding to the dead zone of x _T1 , x _T2 , x _T3 ... X _{Tn is} a real metering performance has (see Fig. 12). Since this minimum stroke length corresponds to the membrane deformation, it is also dependent on the working pressure p ₁ , p ₂ , p ₃ ... P _n .

This characteristic shift x _T1 , x _T2 , x _T3 ... x _Tn requires prior art recalibration under real working conditions as soon as the previously set stroke length is significantly changed, since the new dosing with sufficient accuracy via a proportional conversion from the previous and the newly set stroke length can be determined.

If the membrane deformation is compensated, as previously described, then the proportional error in the partial stroke operation is also eliminated, so that the metering pump is practically over the full usable adjustment range of the stroke length of e.g. 20% -100% can be operated without having to perform the previously necessary recalibrations, which in an unregulated metering pump with an adjustment of the stroke length by more than. 10% are necessary to ensure the specified metering accuracy.

Estimation of the working pressure based on measured electrical quantities; electronic pressure limitation; Overpressure detection

The previously described measurement of the membrane deformation allows on the basis of empirical values for the material properties of the membrane sufficiently accurate conclusions about the working pressure in order to realize additional functions described below.

Unregulated magnetic metering pumps according to the prior art have the fundamental property that the force that develops the drive magnet, in the course of the lifting movement by the decreasing air gap increases sharply. The magnet current is sized so that the force at the point of approach, i. at the largest air gap, sufficient for the rated working pressure. At the end of the stroke a multiple of this force is applied. This has the consequence that the pump in case of faulty piping, e.g. in the case of accidentally closed shut-off devices, it is possible to develop a pressure which rises sharply above the maximum operating pressure, especially if it is operated in the partial stroke, that is to say with a reduced stroke length.

For magnetic pumps that use a position sensor and a controller to control the membrane movement, however, the position of the pressure piece and thus the length of the residual air gap can be determined at any time. Together with the known current-force-displacement characteristic of the drive magnet, it is possible to dynamically adjust the maximum current with which the controller can control the magnet in the course of Dosierhubs to the current membrane position, that the developed maximum force over the entire way remains approximately constant values limited. Thus, the developed maximum pressure can be much more accurately and independently of the set Teilhublänge be limited, so that the use of additional pressure-limiting resources can be unnecessary in many cases.

The application of the invention moreover makes it possible that the control of the metering pump can obtain information on the overpressure state by independent measurements, so that a reaction to this state is possible without the involvement of external operating means, such as e.g. the generation of a message alarm and / or stopping the metering pump.

The achievable accuracy of these functions depends on the reproducibility of the underlying material properties, especially of the membrane. This accuracy can be increased by a single calibration in the production phase or in the actual application by the metering pump is operated at a known pressure and then for the further calculation, the ratio between this known pressure and the membrane deformation found in this case is used.

The above-described applications of the position sensor together with the scheme show that the exact position of the membrane can be detected and monitored by the use of a position sensor, for example on the push rod or the pressure piece during the entire lifting and suction. The situation assessment and monitoring means that situational Control specifications that lead to the described advantages are accurately maintained by means of the actual value measurement.

Cooling of the magnet and other components

Compared to magnetic metering pumps of previous design is in certain modes such. the slow dosing applied to the solenoid 2 for much longer time to continuous operation with power, resulting in a significantly increased heat loss result. In particular, when installed in a plastic housing, the problem of heat dissipation. Magnetic metering pumps are often installed in a plastic housing in a splash-proof design in order to improve the insensitivity to aggressive chemicals in typical applications. In these cases, the task of controlled magnetic metering pumps is to ensure cooling solely by heat conduction through the housing wall without exchange of air.

In the case of magnetic metering pumps of the previous type, the magnet is usually installed in the housing 1 in such a way that the magnet jacket 17 has heat-conducting contact with the housing 1 in the largest possible area of its surface; This contact can be improved, for example by encapsulation of the magnet in the manufacture of the housing. The heat dissipation takes place in part through this interface from the magnetic sheath 17 to the inner wall of the housing 1. The other part of the heat loss of the magnet is discharged together with the heat loss of other components in the interior of the housing to the inner housing air, which heats up accordingly. This heat is also forwarded by convection to the inner wall of the housing, from where it is passed together with the directly coupled portion of the heat loss of the magnet through the wall of the housing 1 and ultimately discharged by convection from the outer wall of the housing 1 to the surrounding air , Due to the multiple interface transition exclusively by convection and the usually not very good thermal contact of the magnetic shell 17 to the housing 1, for example due to registration accuracy, Entformungsschräge the plastic housing, etc., the magnet is already very hot in unregulated operation according to the previous version; the temperature may be above 100 ° C, for example. The outer wall of the housing is also very hot especially in the region above the magnet, which is usually absorbed by the design by means of housing ribs 3, which, inter alia, act as shock protection, in that only a small portion of the total surface, namely the upper part of the comb of the ribs, can be touched. Since the housing ribs 3 deliver significantly less heat to the skin when touched against a smooth surface, the case temperature is thus perceived as less hot. However, the formation of relatively narrow air channels obstructs the convection and worsens the ribs thus the heat dissipation of the housing, which further increases both the surface temperature and the internal temperature.

In the case of a magnetic metering pump according to the invention, the conventional design of the heat removal due to the described problem, especially with slowly executed strokes, is not sufficient. There is a much more effective heat removal required by forced circulation of the internal air by means of a fan. In Fig. 13, the concept of cooling is shown in more detail. In the upper part of the housing 1, the magnet is centered by a plurality of, in the described embodiment, three holding webs 50 installed so that the magnetic jacket 17 in a maximum range of its circumference and its end faces a clearance distance, for. of at least 5-10 mm to the housing 1 has. In the lower part of the housing, the control electronics 44 and a fan 43 is arranged so that the fan generates a circulating air flow 47, which flows around the magnetic sheath 17 and also to be cooled electronic power components 45. The fan 43 may be part of the control electronics 44 or an independently mounted in the housing 1 component as in the embodiment described in detail. Of course, the fan can also be arranged elsewhere; It is important that the air circulation ensures that the amount of heat is dissipated by the heat as evenly as possible reaches all areas of the housing inner wall and thus uses them for heat dissipation. It is also conceivable that the fan is located outside the housing and is sealed to this.

The arrangement of the holding webs 50 and the space between the magnetic jacket 17 and the housing 1 form one or more flow channels, which deflect the air flow 47 as effectively and over a large area around the entire surface of the magnet and the air pass all parts of the inner wall of the housing 1. The heat loss of the magnet is given in the embodiment much more effective than by convection alone to the inside air and passed through the intense turbulence just as well into the wall of the housing 1. It is also important here that, in contrast to the previous design not only mainly the area of the housing is heated, which is in contact with the magnet, but practically the entire surface of the housing is uniformly heated by application of the invention and thus heat dissipation to the surrounding air contributes. The particularly hot areas of the housing surface which arise in the previous embodiment, in particular above the magnet, are thus avoided, so that, for example, the housing ribs 3 can also be omitted to minimize the contact surface that can be touched. This improves the heat dissipation of the housing again, since the associated with the housing ribs 3 obstruction of the convection is also eliminated.

In the embodiment described in more detail, the lifting cover 28 is designed so that it selectively passes a portion of the air flow 47 at the position sensor 36 and this part of the air flow via one or more outflow openings 46 the main air flow again. Due to the principle-based mounting of the position sensor 36 close to the (hot) magnet, the position sensor is exposed to particularly high temperatures. With passive cooling of the prior art, the magnet would be heated very high due to the poorer heat dissipation, and the position sensor 36 would in turn assume approximately the surface temperature of the magnet. When using the cooling by air circulation according to the invention, the temperature of the position sensor 36 is maintained approximately in the range of the inside air temperature, if care is taken in particular in the construction of the sensor carrier 31, that this sufficient thermal delimitation of the sensor elements (CCD receiver 32 and light source 33) ensures the metallic parts of the magnet. The above applies mutatis mutandis to any possibly in the housing cover (5) built-in electronics (6). This is also cooled by a deliberately guided over them away partial air flow 49.

position sensor

As already stated, as the position transmitter in the described embodiment, the shadow body 35 for scanning the position is attached to the extended push rod 19 whose shadow is imaged on the line of charge coupled device (CCD) cells 32. The active sensor elements which are described in more detail in this exemplary embodiment and which detect the position are arranged on the side of the pressure piece facing away from the dosing head. As the light source 33 is an LED, the optical receiver is an electronic module with a CCD line 32, which are mounted here together on an intermediate part, the sensor carrier 31. The mounting on the sensor carrier 31 makes it possible to treat the position sensor 36 in the production process as a separate assembly and e.g. pre-assembled and tested outside the final installation site. In addition, the described photocell-like arrangement is a touch and thus wear-free working sensor.

For the basic operation, the mounting location of the sensor is of no importance, the relevant determination can rather be made according to structural considerations such as space, assembly order, etc. Moreover, the parts described here as fixedly mounted (light source 33, receiver 32) and those moving together with the diaphragm (shadow body 35) can also change their function.

In the exemplary embodiment, the CCD module 32 is driven by an evaluation unit which contains a microprocessor and generates the required control signals. Instead of a microprocessor, the evaluation unit can also be realized by a DSP (Digital Signal Processor) or in discrete technology.

As a light source 33, in principle, each component is suitable, which has a sufficiently narrow confined spot. Together with the imaging geometry shown in more detail in FIG. 7, this determines the width of the shadow area SV, see FIG. also Fig. 8.

It is also possible to use a plurality of elements or a line emitter as the light source 33, with the aid of which the shadow profile SV can be specifically designed according to particular aspects. As an example, the achievement of a higher brightness, without affecting the image sharpness in the direction of movement.

The CCD line 32 is a linear array of M optical receivers (hereinafter called pixels) arranged in a regular raster R of several μm. In the example, this is 128 pixels in the grid of about 64μm on a total length of about 8mm, i. M = 128 and R = 64μm.

The control signals which are generated by the evaluation unit, determine the exposure time, during which the individual pixels of the CCD line 32 integrate the incident light in each case in a separate amplifier within the CCD chip and cache for later evaluation. This integration occurs not only over the exposure time, but also across the photosensitive area of each pixel. After the exposure, the brightness values belonging to the pixels are successively read out of the CCD module by further control signals as analog values and detected by the evaluation unit.

Exposure and readout of the brightness values take place alternately in the simple case. Depending on the design, some commercially available CCD line components also offer options for the simultaneous execution of both processes, by buffering the integrated measured values after the exposure and immediately releasing the integrators again for a subsequent measurement. By simultaneously reading out the results of a measurement run during the exposure phase for the subsequent passage, the measurement speed can be increased.

In the diagram shown in FIG. 8, the integrated brightness values H are shown corresponding to the actual shadow profile in the area of the addressed pixels in the specific embodiment. The shadow area SV extends over the pixels # 60 to # 63 in this example.

As a simple evaluation method, a decision threshold H _v (shown in Figure 8 as the dashed line.) Arbitrarily set at, for example half the maximum brightness and searched for the one pixel whose brightness value H, first, the threshold below the shadow transition H _v; in the example this would be Pixel # 62.

In other embodiments, the brightness profile may be in opposite directions from unlit to illuminated CCD cells with increasing pixel number; On the one hand, this depends on the arrangement of the elements light source 33, CCD module 32 and shadow body 35 and on the other hand on the internal organization of the CCD 32 used. In this case, the pixel whose brightness value at the shadow transition first exceeds the threshold is searched for.

After the three phases of exposure, readout and processing, a position value is available. The total time requirement of the three phases determines the repetition rate at which position values are obtained. The measurement resolution is equal to the pixel pitch R of the CCD line, corrected by the imaging ratio A, which results from the mounting distance with the individual components.

For the imaging ratio A (see Fig. 9):
A = s' / s = x ₃ / x ₂

Where s = actual movement of the shadow edge
s' = Projected movement of the shadow edge in the plane of the CCD
x ₂ = distance between optically effective shadow edge and light source
x ₃ = distance between CCD plane and light source

This method determines the position by counting pixels, so it is to be regarded as a digital method. Deviations and displacements of linear parameters such as component sensitivities have practically no effect on the result compared to analogous methods. If one measures the imaging ratio A for practical values, assembly tolerances also have only a small influence. In a practical embodiment with x ₃ = 21 mm and x ₂ = 20 mm results in a nominal value for the imaging ratio A of 1.05; ie a movement of the shadow body 35 by a certain distance results in a 1.05-fold shift of the shadow area SV in the plane of the CCD cells 32. Let us now assume a mounting tolerance for x ₃ , ie a possible variation of the distance of the CCD cells 32 from the light source 33 by ± 0.3 mm, and a concrete mounting case at the upper end of this tolerance range with x ₃ = 21.3mm and x ₂ = 20mm. In this case, the imaging ratio A is calculated to be 1.065. The imaging ratio changes by 1.065 / 1.05 = 1.014 or + 1.4% in this example. This deviation can easily be eliminated by a one-time calibration, eg during production. The linearity is determined almost exclusively by the accuracy of the pixel grid within the chip geometry, deviations are thus negligible.

Although the previously described method for determining the position of the shadow body 35 and thus the position of the diaphragm 13 already yields very precise and linear position values, an even more accurate position resolution can be achieved by interpolation. In this advanced embodiment, by evaluating the pixel brightnesses H, a positional resolution is achieved, eg between pixels 61 and 62 (see Fig. 10), which is finer than the pixel raster R by interpolating the brightness values of the pixels in the region of the decision threshold. The goal is to determine the point at which the brightness curve intersects the decision threshold H _V and to assign this point of intersection a value on a virtual position scale whose x-values in the middle of the pixel correspond exactly to the pixel number.

• Δx_l/(Δx_l + Δx_r) = (ΔH_l + ΔH_r)
• Δx_l = ΔH_l/(Δx_l + Δx_r)

For this purpose, the two pixels to the left and to the right of the decision threshold H _{V are} searched for and the distances ΔH of the associated brightness values are evaluated from this threshold. As shown in FIG. 10 or in FIG. 11, the following applies:

• Δx _l / (Δx _l + Δx _r ) = (ΔH _l + ΔH _r )
• Δx _l = ΔH _l / (Δx _l + Δx _r )

The distances .DELTA.x calculated from the respective center axis of each of the two adjacent pixels, in this example pixels # 61 and # 62, in multiples of the pixel width to the intersection form the following relationship with the brightness distances .DELTA.H relative to the pixel # 61 to the left of the searched intersection (left-side neighboring pixel): $${\mathrm{Ax}}_{\mathrm{I}}\mathrm{/}\mathrm{(}\mathrm{\Delta }{\mathrm{x}}_{\mathrm{I}}\mathrm{+}{\mathrm{Ax}}_{\mathrm{r}}\mathrm{)}={\mathrm{AH}}_{\mathrm{I}}/\mathrm{(}\mathrm{\Delta }{\mathrm{H}}_{\mathrm{I}}\mathrm{+}{\mathrm{AH}}_{\mathrm{r}}\mathrm{)}$$

With (Δx _l + Δx _r ) = 1 (1 pixel width) the result is: $${\mathrm{AH}}_{\mathrm{I}}={\mathrm{AH}}_{\mathrm{I}}/\mathrm{(}\mathrm{\Delta }{\mathrm{H}}_{\mathrm{I}}\mathrm{+}{\mathrm{AH}}_{\mathrm{r}}\mathrm{)}$$

Relative to the pixel # 62 (right-hand neighboring pixel) on the right of the desired intersection, the relationship applies: $${\mathrm{Ax}}_{\mathrm{r}}\mathrm{/}\mathrm{(}\mathrm{\Delta }{\mathrm{x}}_{\mathrm{I}}\mathrm{+}{\mathrm{Ax}}_{\mathrm{r}}\mathrm{)}\mathrm{=}{\mathrm{AH}}_{\mathrm{r}}/\mathrm{(}\mathrm{\Delta }{\mathrm{H}}_{\mathrm{I}}\mathrm{+}{\mathrm{AH}}_{\mathrm{r}}\mathrm{)}$$

With (Δx _l + Δx _r ) = 1 (1 pixel width) the result is: $${\mathrm{Ax}}_{\mathrm{r}}\mathrm{=}{\mathrm{AH}}_{\mathrm{r}}/\mathrm{(}\mathrm{\Delta }{\mathrm{H}}_{\mathrm{I}}\mathrm{+}{\mathrm{AH}}_{\mathrm{r}}\mathrm{)}$$

In this example, the intersection is 61.7. If the brightness progression in the interpolation range follows an ideal straight line, then both calculation paths lead to the same result, so it is basically sufficient to carry out one of the two calculations. However, with this property, error contributions can be minimized by a not exactly straight brightness curve in the considered transition range or by always expected measurement inaccuracies, for example by performing both calculations and averaging their results.

In other embodiments, depending on the brightness curve, the conditions on both sides of the point of intersection with respect to unlit and illuminated CCD cells can be reversed; In this case, the direction information on the left and right, if necessary, change their function and the interpolation equations must be adapted accordingly.

In addition, other embodiments are possible in which the brightness values of more than two pixels are used for the calculation. The position can then be determined by redundant multiple calculation and e.g. Averaging several results are formed. As a further possibility, a different linear interpolation than shown here or an interpolation with the data other than the direct neighbor pixels can be used.

Deviations and displacements of linear parameters, e.g. Component sensitivities affect the result only within the interpolation range. The steepness of the brightness gradient in the shadow transition, resulting from the sharpness of the image of the shadow edge on the CCD plane, is of subordinate importance since it does not affect the interpolation within wide limits; only the linearity of the brightness curve is decisive for the accuracy of the interpolation.

• Verbesserung der Störimmunität durch Filterung
  Die Störimmunität des Sensors kann durch Filtermaßnahmen verbessert werden. Eine Filterung kann sowohl auf Ebene der Helligkeitswerte der Pixels als auch auf das Ergebnis der Positionsermittlung selbst angewandt werden. Im ersten Fall arbeitet das Verfahren mit Helligkeitswerten, die über mehrere Pixels oder über mehrere Durchgänge gemittelt wurden, im zweiten Fall werden mehrere zunächst ermittelte Positionsergebnisse zu einem abgeleiteten Positionswert zusammengefasst, mit dem dann die weitere Bearbeitung vorgenommen wird.
• Kompensation von Montageabweichungen
  In einer definierten Phase, z.B. in der Ruhephase vor Ablauf des eigentlichen Dosierhubs, kann der Positionswert für diese Phase ermittelt und in einem Referenzspeicher abgelegt werden. Während der aktiven Bewegungsphase werden dann die Positionswerte relativ zu dem zuvor ermittelten Referenzwert verarbeitet. Durch dieses Verfahren ist es möglich, fertigungsbedingte Montageabweichungen der Ruhelage sowie Verschiebungen während des Betriebs z.B. durch Wärmeausdehnung automatisch zu kompensieren und damit die Genauigkeit zu verbessern.
• Kompensation von Skalierungsfehlern
  Bei einer erweiterten Alternative kann durch Anfahren zweier oder mehrerer bekannter Positionen, hier Referenzpositionen genannt, die Skalierung des Positionssensors abgeglichen werden. Dies kann einmalig im Zuge des Produktions- bzw. Prüfverfahrens oder auch wiederkehrend im Betrieb geschehen.
  Im ersten Fall können die Referenzpositionen durch externe Einrichtungen, z.B. Raststellungen oder externe Messeinrichtungen, vorgegeben werden. Aus den in diesen Referenzpositionen gemessenen Positionswerten kann zusammen mit der Kenntnis über die wirkliche Lage der Referenzpositionen ein Korrekturwert für die Skalierung des Positionssensors abgeleitet und für die weitere Verarbeitung gespeichert werden.
  Im zweiten Fall des wiederkehrenden Skalierungsabgleichs sind bekannte Positionen, z.B. mechanische Anschläge oder Referenzsignale von weiteren vorhandenen Einrichtungen zur Positionserfassung notwendig. Befindet sich die Membran während des Betriebs an einer solchen bekannten Position, kann aus dem an dieser Stelle gemessenen Positionswert ebenfalls ein Korrekturwert für die Skalierung des Positionssensors abgeleitet und für die weitere Verarbeitung gespeichert werden.
• Kompensation der optischen Empfindlichkeitsparameter
  In einer erweiterten Ausführung können die Helligkeitswerte der voll beleuchteten Pixels dazu herangezogen werden, um einen repräsentativen Wert für die Beleuchtungsstärke zu ermitteln. Hierzu kann beispielsweise aus einer geeigneten Gruppe von Pixels der Mittelwert der Helligkeit gebildet werden. Anhand der ermittelten Beleuchtungsstärke kann die Belichtung so gesteuert werden, dass die zur Verfügung stehenden Wertebereiche optimal ausgenutzt werden; beispielsweise kann die Lichtquelle in ihrer Helligkeit oder ihrer Einschaltdauer so gesteuert werden, daß die Beleuchtungsstärke der voll beleuchteten Pixels wenig unterhalb der Übersteuerungsgrenze des CCD-Bausteins liegt. Einer Steuerung der Einschaltdauer der Lichtquelle ist eine Taktung mit veränderlichem Ein-/Ausschaltverhältnis gleichzusetzen. Bei jedem Messdurchgang wird dann die Beleuchtungsstärke anhand der Verhältnisse des vorangegangenen Durchgangs so korrigiert, dass sich eine gleitende Anpassung der Belichtungsparameter an eventuelle Veränderungen von Bauteileeigenschaften, z.B. aufgrund von Alterung, ergibt.
• Kompensation von Verschmutzungen und Pixelabweichungen
  In einer erweiterten Ausführung kann der mechanische Aufbau des Sensors so gestaltet werden, dass in einer definierten Phase, z.B. in der Ruhephase vor Ablauf des eigentlichen Dosierhubs, der komplette für den Arbeitsweg genutzte Pixelbereich oder ein interessierender Teilbereich belichtet werden kann. Eine mögliche Ausführung ist z.B., die dem Magneten zugewandte Kante des Schattenkörpers für die Auswertung heranzuziehen, wodurch der Schattenkörper im Verlauf der Hubbewegung den Sensor überstreicht und einen Bereich der CCD-Zellen abdunkelt, der im vorherigen Ruhezustand beleuchtet war. In dieser Phase können die Helligkeitswerte aller relevanten Pixels ermittelt und in einem Referenzspeicher einzeln abgelegt werden. Abweichungen der Messwerte einzelner Pixels vom Idealwert können z.B. in Form von Korrekturwerten hinterlegt werden. Während der aktiven Bewegungsphase werden dann die Helligkeitswerte jedes Pixels mit Hilfe der zuvor ermittelten Referenzwerte bei jeder Messung zunächst korrigiert und erst dann weiterverarbeitet. Durch dieses Verfahren ist es möglich, fertigungsbedingte Empfindlichkeitsabweichungen einzelner Pixels sowie Verschmutzungen in gewissem Rahmen zu kompensieren und damit die Genauigkeit zu verbessern bzw. die Betriebssicherheit zu erhöhen.
  Natürlich sind für die CCD-Empfängerzeile auch zwei- oder mehrreihige Anordnungen möglich, um durch Redundanz eine erhöhte Sicherheit gegen Ausfälle, z.B. durch Verschmutzung, zu erreichen bzw. durch Mittelung die Meßgenauigkeit zu erhöhen. Für besonders große Hublängen können zwei oder mehr CCD-Zeilen kombiniert werden, um den Meßbereich über die Funktionsgrenzen einer einzelnen Zeile hinaus zu erweitern.

Independently of the previously described interpolation method, further methods for improving the sensor properties can be used based on the described basic principle. These methods are described below:

• Improvement of the interference immunity by filtering
  The interference immunity of the sensor can be improved by filtering measures. Filtering can be applied to both the brightness values of the pixels and the result of the position determination itself. In the first case, the method works with brightness values that have been averaged over several pixels or over several passes, in the second case several initially determined position results are combined to form a derived position value with which the further processing is then carried out.
• Compensation of mounting deviations
  In a defined phase, for example in the resting phase before the end of the actual metering stroke, the position value for this phase can be determined and stored in a reference memory. During the active movement phase, the position values are then processed relative to the previously determined reference value. By this method, it is possible to automatically compensate for manufacturing-related assembly deviations of the rest position and shifts during operation, for example by thermal expansion and thus to improve the accuracy.
• Compensation of scaling errors
  In an extended alternative, the scaling of the position sensor can be adjusted by approaching two or more known positions, here called reference positions. This can happen once in the course of the production or test procedure or also recurrently in operation.
  In the first case, the reference positions can be specified by external devices, such as detent positions or external measuring devices. From the position values measured in these reference positions, together with the knowledge of the actual position of the reference positions, a correction value for the scaling of the position sensor can be derived and stored for further processing.
  In the second case of recurrent scaling adjustment, known positions, eg mechanical stops or reference signals of further existing devices for position detection are necessary. If the diaphragm is in such a known position during operation, a correction value for the scaling of the position sensor can also be derived from the position value measured at this point and stored for further processing.
• Compensation of the optical sensitivity parameters
  In an extended embodiment, the brightness values of the fully illuminated pixels can be used to determine a representative value for the illuminance. For this purpose, for example, the mean value of the brightness can be formed from a suitable group of pixels. Based on the determined illuminance, the exposure can be controlled so that the available value ranges are optimally utilized; For example, the light source in its brightness or its duty cycle can be controlled so that the illuminance of the fully illuminated pixel is slightly below the overdrive limit of the CCD. A control of the duty cycle of the light source is a clocking with variable on / off ratio equate. During each measuring run, the illuminance is then corrected on the basis of the conditions of the preceding pass, so that a sliding adaptation of the exposure parameters to possible changes in component properties, eg due to aging, results.
• Compensation of dirt and pixel deviations
  In an expanded embodiment, the mechanical structure of the sensor can be designed so that in a defined phase, for example in the idle phase before the actual dosing stroke, the entire pixel area used for the working path or a partial area of interest can be exposed. One possible embodiment is, for example, to use the edge of the shadow body facing the magnet for the evaluation, whereby the shadow body sweeps over the sensor in the course of the lifting movement and darkens a region of the CCD cells which was illuminated in the previous resting state. In this phase, the brightness values of all relevant pixels can be determined and stored individually in a reference memory. Deviations of the measured values of individual pixels from the ideal value can, for example, be stored in the form of correction values. During the active movement phase, the brightness values of each pixel are then first corrected with the aid of the previously determined reference values for each measurement and only then further processed. By this method, it is possible to compensate for manufacturing-related sensitivity deviations of individual pixels and contamination in a certain scope and thus to improve the accuracy and to increase the reliability.
  Of course, two- or multi-row arrangements are possible for the CCD receiver line to achieve through redundancy increased security against failures, eg due to contamination, or to increase the accuracy of measurement by averaging. For particularly long stroke lengths two can or more CCD lines can be combined to extend the measurement range beyond the functional limits of a single line.

Adaptation of the magnet design and the thermal design

In order to be able to exploit the advantages of a magnetic dosing pump, which has been described in detail, particularly with slowed down movement to a standstill, constructive adjustments are required, in particular in the magnetic design and active cooling by means of internal fans in the interior of the device, which has already been described in detail above (see chapter "Cooling the magnet and other components").

Magnetic design according to the criteria normally used in magnetic dosing pumps is only of limited use for motion-controlled operation without modifications. In order to enable controllability in a wide range at all, it is essential to be able to react to the natural movements of the mechanical components even in the worst case with at least as fast magnetic current changes.

This is contrary to previously customary interpretation of the excessive inductance of the solenoid coil 2, through which the magnetic current I _M only after a time of normally about 20-50ms reaches its nominal value. This customary design is chosen such that the desired current is set by the interaction of the applied voltage and the impedance of the winding 29 (ohmic resistance, inductance). In the simple case, this current results in the supply voltage provided for the device, possibly less a tolerance margin; In current-controlled versions, the dimensioning is chosen so that the current flow at the lowest expected supply voltage is still guaranteed and is limited at higher voltages by a control loop to the predetermined value.

If the magnet is to be suitable for controlling the course of motion, a significantly lower number of turns must be selected so that the magnetic current can be influenced in a shorter time. With the same utilization of the available winding space, the reduction factor of the number of turns (N) has an approximately quadratic effect on resistance and inductance, as a result of which the current slew rate increases with unchanged voltage in the ratio N ² . However, the power requirement for a given magnetic force also increases in the ratio N, so that overall an effective reduction of the time until reaching the working current by a factor of N is achieved.

This will be explained in more detail in the following example. For this purpose, for the sake of simplicity, an approximately linear current increase, i. a purely inductive behavior of the solenoid 2 assumed. A solenoid coil according to the previous design can increase the magnetic current by 0.1 A / ms due to their impedance. A required working current of 2A is thus achieved within 20 ms after the voltage has been applied. If the working current is reached in half the time (10 ms), the previous number of turns must be halved and the wire cross section correspondingly doubled, i. the wire diameter is to be increased by the factor root of 2; Inductance and winding resistance decrease by a factor of 4, thus the current increase rate increases with unchanged voltage to 0.4 A / ms. The working current doubles to 4A and is achieved in half the original time (10ms).

The ability to control can in particular be exploited to slow down the natural movement. This results in a temporal extension of the magnetic current flow I _M up to almost continuous operation, thereby increasing the resulting loss energy per stroke. Depending on which of the described functions are implemented, this can cause a considerable increase in the dissipated heat loss. Depending on the extent of this increase, a thermal design according to extended criteria with changes in the mechanical structure necessary, which allow increased heat dissipation. The increased and longer-flowing working current I _{M of} the magnet must also be taken into account by more dimensioned components 45 in the control electronics 44.

The use of a position sensor in conjunction with a control of the membrane movement to improve the hydraulic properties and to expand the field of application of the magnetic metering pump must not cause economic reasons that all individual components of the magnetic metering pump must be redesigned. So it is important to ensure that the known and common components can be largely reused. This also applies to the installation dimensions, which is why it must be ensured in the structural design that as far as possible already used in unregulated Magnetdosierpumpen parts can also be used in regulated Magnetdosierpumpen.

1

Gehäuse

2

Magnetspule

3

Gehäuserippen

4

Bodenplatte

5

Gehäusedeckel

6

Elektronik im Gehäusedeckel

7

Verstellorgan

8

Hubverstellbolzen

9

Abdeckhaube

10

Steuerleitungen

11

Stromversorgung

12

Dosierkopf

13

Membran

14

Einlassventil

15

Auslassventil

16

Dosierraum

17

Magnetmantel

18

Längsachse

19

Schubstange

20

Druckstück

21

O-Ring

22

Innere Stirnfläche des Druckstücks

23

Druckfeder (Rückholfeder)

24

Innere Stirnfläche des Magnetmantels

25

Magnetscheibe

26

Druckkopfseitige Buchse

27

Druckstückseitige Buchse

28

Hubdeckel

29

Spulenwicklung

30

Membrankern

31

Sensorträger

32

Empfänger, CCD-Baustein

33

Lichtquelle

34

Öffnung

35

Bezugselement, z.B. Schattenkörper

36

Positionssensor

37

Differenzierer

38

Sollwertvorgabe

39

Soll-Ist-Vergleich

40

PID-Regler

41

Lage-Korrektur

42

Verstärker

43

Ventilator

44

Ansteuerelektronik

45

Elektronische Leistungsbauteile

46

Ausströmöffnung

47

Luftstrom des Ventilators

48

Teilluftstrom für Positionssensor

49

Teilluftstrom für Elektronik im Gehäusedeckel

50

Haltestege

51

Spulenträger

SV

Schattenverlauf

h

heller Bereich

d

dunkler Bereich

#58...#65

Zellen (Pixels) des CCD

H

Helligkeitswerte der Pixels

H_v

Helligkeitswert der Vergleichsschwelle (VS)

H_l

Helligkeitswert des Pixels links vom Schnittpunkt mit der VS (linksseitiges Nachbarpixel)

ΔH_l

Helligkeitsabstand des linksseitigen Nachbarpixels zum Helligkeitswert der Vergleichsschwelle

H_r

Helligkeitswert des Pixels rechts vom Schnittpunkt mit der VS (rechtsseitiges Nachbarpixel)

ΔH_r

Helligkeitsabstand des rechtsseitigen Nachbarpixels zum Helligkeitswert der Vergleichsschwelle

ΔX_l

Positionsabstand der Mittellinie des linksseitigen Nachbarpixels zum Schnittpunkt mit der VS

Δx_r

Positionsabstand der Mittellinie des rechtsseitigen Nachbarpixels zum Schnittpunkt mit der VS

x₁

Abstand zwischen Schattenkante und CCD-Ebene

x₂

Abstand zwischen Schattenkante und Lichtquelle

x₃

Abstand zwischen CCD-Ebene und Lichtquelle

p₁

Arbeitsdruck p₁

p₂

Arbeitsdruck p₂

p₃

Arbeitsdruck p₃

p₄

Arbeitsdruck p₄

x_T1

Totbereich bei Arbeitsdruck p₁

x_T2

Totbereich bei Arbeitsdruck p₂

x_T3

Totbereich bei Arbeitsdruck p₃

x_T4

Totbereich bei Arbeitsdruck p₄

s

Tatsächliche Bewegung der Schattenkante

s'

Projizierte Bewegung der Schattenkante

D

Dosierleistung

HL

Mechanische Hublänge

SG

Stellgröße

KSG

Korrigierte Stellgröße

k1

Faktor für die positionsabhängige Stromkorrektur

k2

Faktor für den Leistungsverstärker

k3

Faktor für die Ableitung des Geschwindigkeitssignals

x_S

Sollwert der Druckstückposition

x_I

Istwert der Druckstückposition

x_Sl

Regelabweichung der Druckstückposition

v_S

Sollwert der Druckstückgeschwindigkeit

v_l

Istwert der Druckstückgeschwindigkeit

v_Sl

Regelabweichung der Druckstückgeschwindigkeit

Pos

Positionssignal in Diagrammen

EPos

Endanschlag des Positionssignals in Diagrammen

I_M

Magnetstrom

I_Mo

Nullage des Magnetstromsignals in Diagrammen

LIST OF REFERENCE NUMBERS

1

casing

2

solenoid

3

housing ribs

4

baseplate

5

housing cover

6

Electronics in the housing cover

7

adjusting

8th

Hubverstellbolzen

9

cover

10

control lines

11

power supply

12

dosing

13

membrane

14

intake valve

15

outlet valve

16

metering

17

magnetic shell

18

longitudinal axis

19

pushrod

20

Pressure piece

21

O-ring

22

Inner face of the pressure piece

23

Compression spring (return spring)

24

Inner face of the magnetic jacket

25

magnetic disk

26

Printhead side jack

27

Pressure piece-side socket

28

pedal lid

29

coil winding

30

diaphragm core

31

sensor support

32

Receiver, CCD module

33

light source

34

opening

35

Reference element, eg shadow body

36

position sensor

37

differentiator

38

Setpoint

39

Target-performance comparison

40

PID controller

41

Position correction

42

amplifier

43

fan

44

control electronics

45

Electronic power components

46

outflow

47

Air flow of the fan

48

Partial air flow for position sensor

49

Partial air flow for electronics in the housing cover

50

retaining webs

51

coil carrier

SV

shadow History

H

bright area

d

dark area

# 58 ... # 65

Cells (pixels) of the CCD

H

Brightness values of the pixels

H _v

Brightness value of the comparison threshold (VS)

H _l

Brightness value of the pixel to the left of the intersection with the VS (left-sided neighbor pixel)

ΔH _l

Brightness distance of the left-side neighboring pixel to the brightness value of the comparison threshold

H _r

Brightness value of the pixel to the right of the intersection with the VS (right-hand neighbor pixel)

ΔH _r

Brightness distance of the right-hand neighboring pixel to the brightness value of the comparison threshold

ΔX _l

Position distance of the center line of the left-side neighboring pixel to the intersection with the VS

Δx _r

Position distance of the center line of the right-hand neighbor pixel to the intersection with the VS

x ₁

Distance between shadow edge and CCD plane

x ₂

Distance between shadow edge and light source

x ₃

Distance between CCD plane and light source

p ₁

Working pressure p ₁

p ₂

Working pressure p ₂

p ₃

Working pressure p ₃

p ₄

Working pressure p ₄

x _T1

Dead zone at working pressure p ₁

x _T2

Dead zone at working pressure p ₂

x _T3

Dead zone at working pressure p ₃

x _T4

Dead zone at working pressure p ₄

s

Actual movement of the shadow edge

s'

Projected movement of the shadow edge

D

Dosing capacity

HL

Mechanical stroke length

SG

manipulated variable

KSG

Corrected manipulated variable

k1

Factor for the position-dependent current correction

k2

Factor for the power amplifier

k3

Factor for the derivation of the speed signal

x _S

Setpoint of the pressure piece position

x _I

Actual value of the pressure piece position

_xsl

Control deviation of the pressure piece position

v _p

Setpoint of the pressure piece speed

v _l

Actual value of the pressure piece speed

v _Sl

Control deviation of the pressure piece speed

Pos

Position signal in diagrams

Epos

End stop of the position signal in diagrams

I _M

magnet power

I _mo

Zero position of the magnetic current signal in diagrams

Claims (34)

Magnetic metering, in which a movable pressure piece with a fixedly connected to this push rod is mounted axially movable in a fixedly mounted in the pump housing magnetic jacket in the longitudinal axis, so that the pressure piece with push rod in the electrical control (activation) of the magnetic coil in the magnetic sheath against the effect a compression spring is drawn with reduction of the air gap in a bore of the magnetic shell and the pressure piece returns after deactivation of the magnet by the compression spring in the starting position, so that the pressure piece and an actuated by this elastic displacement member with continued activation and deactivation of the magnetic coil performs an oscillating movement in the arranged in the longitudinal axis of the dosing head in cooperation with an outlet and inlet valve to a pumping stroke (pressure stroke) and a suction stroke, characterized in that with the unit of pressure piece (20) and Schubsta (19) a reference element (35) is connected, the position of which is scanned by a position sensor (36), wherein the position sensor outputs an actual signal (x _l ), which is in a fixed relationship to the position of the reference element and that via a Regulator within its control accuracy, the movement of the unit of pressure piece and push rod so influenced that it follows a predetermined setpoint profile (38). Magnetic metering pump according to claim 1, characterized in that the position sensor (36) scans the position of the reference element (35) according to a non-contact principle. Magnetic metering pump according to claim 1, characterized in that the reference element (35) connected to the push rod (19) and the position sensor (36) are arranged at the end remote from the metering head (12) and outside the metering head. Magnetic metering pump according to one or more of the preceding claims, characterized in that the reference element (35) influences the beam path of a light source (33) and the cooperating with it position sensor (36), which is arranged on the magnetic jacket (17), operates according to a photosensitive receiver principle , Magnetic metering pump according to one or more of the preceding claims, characterized in that the reference element (35) is a shadow body or a shadowing contour and the cooperating with it position sensor (36) which is arranged on the magnetic jacket (17), from an optical receiver ( 32) in the form of a series of photosensitive Charge-coupled receiver cells consists (charged coupled device, called CCD for short). Magnetic metering pump according to one or more of the preceding claims, characterized in that the position sensor (36) is arranged on a separate sensor carrier, which is firmly connected to the magnetic jacket (17). Magnetic metering pump according to claim 4, characterized in that the light source (33), the shadow body or the shadowing contour (35) and the receiver (32) represent a light barrier-like arrangement and the measured values are continuously or cyclically fed to the control loop. Magnetic metering pump according to one or more of the preceding claims, characterized in that the optical receiver (32) of the position sensor (36) consists of a number of linearly arranged receivers (pixels), preferably 128 pixels. Magnetic metering pump according to one or more of the preceding claims, characterized in that the light source (33) is a light-emitting diode (LED), which is arranged opposite the optical receiver (32) of the position sensor (36), that their light beams on the direct path to Receiver by the push rod (19) is not obstructed. Magnetic metering pump according to one or more of the preceding claims, characterized in that the output value of the position sensor (36) is formed by interpolation of the brightness values of a plurality of pixels lying in the shadow transition region. Magnetic metering pump according to one or more of the preceding claims, characterized in that filter measures are used in the processing of the signals of the position sensor (36). Magnetic metering pump according to one or more of the preceding claims, characterized in that zero position errors of the position sensor (36) are eliminated by means of a reference memory. Solenoid according to one or more of the preceding claims, characterized in that scaling errors of the position sensor (36) by moving to be eliminated of one or more reference positions. Magnetic metering pump according to one or more of the preceding claims, characterized in that exposure fluctuations of the position sensor (36) are compensated for by control or regulation of the light source (33) on the basis of the obtained brightness values of the pixels. Magnetic metering pump according to one or more of the preceding claims, characterized in that brightness fluctuations between individual pixels of the optical receiver (32) are compensated by incorporating a reference memory for the sensitivity of each pixel. Magnetic metering pump according to one or more of the preceding claims, characterized in that the signal (x ₁ ) read out of the position sensor (36) is further processed in a control device and compared with a desired value specification (38), wherein the control device controls the current flow (I _M ) Magnetic coil (2) influenced and so causes a correction of the motion sequence. Magnetic metering pump according to claim 16, characterized in that the control device alternatively by the position (hereinafter x _l called), the speed (hereinafter v _l called) or the acceleration of the pressure piece (19) or the membrane (13) via a control device Change in the coil current (hereinafter called I _M ) influenced. Magnetic metering pump according to claim 16, characterized in that the control device v _l of the pressure piece (20) in the suction phase and / or in the pressure phase can reduce targeted to counteract pressure losses caused by flow resistance, or the occurrence of cavitation. Magnetic metering pump according to claim 16, characterized in that the desired stroke length communicated by an operator specification of the control device and by the control device, the movement of the pressure piece (20) by appropriate driving of the magnetic coil (2) is electronically limited to the stroke length to be executed. Solenoid takes place according to claim 16, characterized in that the recognition on which the value Hubverstellorgan (7) is adjusted by measurement during metering directly via the positional sensor (36). Magnetic metering pump according to claim 16, characterized in that the control device v _I of the pressure piece (20) at the beginning and / or at the end of the printing phase, so for example in the first or in the last third of the stroke, so limited by driving the magnetic coil (2), that pressure spikes caused by rapid changes in the speed of Dosiermediumsstroms or arise by hard starting of the mechanical stop, be avoided. Magnetic metering pump according to claim 16, characterized in that the control device v _I of the pressure piece (20) at the end of the printing phase by driving the magnetic coil (2) so limited that the effect of over-promotion is avoided at low back pressure. Magnetic metering pump according to claim 16, characterized in that the control device distributes the forward movement of the pressure piece (20) during the printing phase by driving the magnetic coil (2) to the time predetermined by the repetition frequency of the metering strokes so that the application of the metering medium takes place as evenly as possible until to very slowly running Dosierhüben of eg a few minutes. Magnetic metering pump according to claim 23, characterized in that the control device in operation at quasi-continuous dosing, ie without substantial rest between suction and the following Dosierhub, converts the lifting movement to a reduced stroke length and increased stroke frequency with approximately maintaining the membrane velocity in Dosierhub in the time average means the desired metering results, and the suction by driving the solenoid (2) finished before the pressure piece (20) pushed by the return spring (23) all the way to the front mechanical (rest) stop or the Hubverstellbolzen (8) was, so that the movement of the pressure piece takes place only in the region of the stroke, in which the air gap and thus the magnetic current requirement (I _M ) is small. Magnetic metering pump according to claim 16, characterized in that during the initial phase of the controlled forward movement of the pressure piece (20) either the control device itself or a further control unit observes the magnetic current (I _M ), closes it on the force curve and thus opening the exhaust valve (15) recognizes and with the help of this observation, the dead zone, which arises due to the elastic deformation of the membrane (13), measures and actually executed stroke by targeted termination of the lifting movement depending on the determined membrane deformation so influenced that caused by the membrane deformation error contribution (based on the Stroke or the metered volume) is eliminated and thus the dependence of the dosing amount is significantly reduced by the back pressure. Magnetic metering pump according to claim 16, characterized in that the control device in operation at reduced stroke length actually executed stroke during the forward movement of the pressure piece (20) depending on the measured dead zone by the elastic Deformation of the membrane (13) influenced by targeted termination of the lifting movement so that the error caused by the membrane deformation error contribution is eliminated and thus the linear dependence of the dosing amount of the percentage value of the set stroke length is significantly improved. Magnetic metering pump according to one or more of the preceding claims, characterized in that the control device during the forward movement of the pressure piece (20) measures the dead zone, which results from the elastic deformation of the membrane (13), and can make an estimate of the working pressure based on this measured dead zone and adjusts the dosage at a predetermined maximum value of the pressure to avoid a further increase in pressure. Magnetic metering pump according to one or more of the preceding claims, characterized in that the housing interior of the Magnetdosierpumpe including the magnet and the electronics (6, 44) are cooled. Magnetic metering pump according to claim 28, characterized in that for cooling the interior and the interior components arranged in the interior, a fan is arranged whose air flow forcibly guided the wall of the magnetic jacket (17) and / or the coil winding (29) and the inner wall of the housing ( 1) of the magnetic metering pump and other components washed around. Magnetic metering pump according to claim 28, characterized in that for cooling the position sensor (36), a part of the air flow (48) is passed over this. Magnetic metering pump according to claim 30, characterized in that with the lifting cover (28) guide surfaces and / or channels are connected, which direct a portion of the air flow (48) on the position sensor (36). Magnetic metering pump according to claim 29, characterized in that a further part of the air flow (49) on the in the housing cover (5) built-in electronics (6) is passed. Magnetic metering pump according to claim 28, characterized in that in the housing (1) of the magnetic jacket (17) is arranged free-standing in the interior so that it can be flowed around for cooling at its periphery by an air flow. Magnetic metering pump according to one or more of the preceding claims, characterized in that the coil winding (29) has a reduced number of turns with increased wire cross-section.

Images