Inspection of containers

Start of patent period: 1981 04 15 Longest possible duration:

© Published on: 1981 1125 © inventor:

© dependence:

© References which have been considered to delimit the prior art: DE-AS 1648640 DE-OS 1773635 US-PS 3521074

- 2 No. 364981

The invention relates to a device for detecting foreign bodies and / or cracks in translucent containers, comprising a transport device for successively transporting the containers through a test zone, means for rotating the containers about an axis during passage through the test zone, a light source for generating a the containers at

Traversing the inspection zone in the direction of the axis scanning point light beam, a light collecting means for receiving the light after passage through the container, a downstream of the light collecting means for generating a light transmittance of the respective container corresponding electrical test signal, a circuit arrangement for generating a comparison signal and a comparator circuit for comparison the test signal with the comparison signal and a controllable by the output of the comparator circuit means for sorting out of containers.

Such a device is described in DE-AS 1648640.

This known device for controlling a translucent or transparent container has a device with which a punctiform beam is generated, which in the vertical direction of

Wall of the container is guided along. The container itself is guided by a transport device through the inspection zone in the vicinity of the device, wherein the point light beam is carried in the horizontal direction with the passing container. During this transport, the container undergoes a rotational movement.

Passed through the container portion of the punctiform beam falls into one of its

Detection provided light collection device. The present in the known device light collecting device is designed so that it collects the light passed through the container regardless of the instantaneous location of the spot light beam on the wall of the container and supplies a downstream light detection device. This generates an electrical test signal corresponding to the light transmission capability of the respective container. This signal is compared with a threshold signal generated in the device and there is a sorting of the individual container according to the result of the comparison.

However, such a known device can not now decide whether a result of the comparison corresponding to a sorting out is actually due to an impurity or a crack in the wall of the container, or whether a change in intensity does not ³⁰ is based solely on a more or less abrupt change in shape of the container, which in itself would be no reason for sorting out the relevant container. Such abrupt changes in shape occur, for example, in bottles between the area of ​​the bottleneck and the actual volume of the same or at the transition from the wall of the container to the bottom. In order to exclude undesired rejections in a known device, one is in this ge35 zwungen, a segregation of the container only hei occurrence of a correspondingly large change in the inspection signal to occur. However, this corresponds to an undesirably lower sensitivity to contamination and cracks, which should necessarily lead to a separation.

Another document of the prior art, namely DE-OS 1773635 describes a 40 electronic circuit for a device for surface inspection of reflective bodies. This circuit and its associated device are designed so that an electronic compensation of the influence of different large gloss behavior of individual uniform body is achieved from each other. This prior art is based on the respective individual copy an average of the gloss behavior of the same to determine. This average value is then continuously compared with the inspection signal which is to be obtained during the course of the entire examination of the respective individual copy of a series.

U.S. Patent No. 3,521,074 describes a corrective action to improve on a tester. This improvement consists in compensating the aging phenomena of this test device, such as the aging of the photomultiplier, the⁵⁰ Amplifier and the like. to reach.

It is an object of the invention to achieve in a device of the type mentioned a significantly higher selection sensitivity with respect to the cracks of the wall and foreign bodies in containers, u.zw. without there being any impairment of this higher selection sensitivity due to the irregular shape of the containers to be examined.

This object is achieved in that the gradual change of

Size of the comparison signal as a function of the position of the scanning light beam along the axis of the container adjustable level sensor are provided, which are successively switchable from the instantaneous position of the Abtast5 light beam detecting sensors.

The invention is based on the idea to provide measures for changing the size of the comparison signal, u.zw. depending on the continuously changing instantaneous vertical position of the impinging punctual beam.

With a device according to the invention it is possible to detect the presence of foreign bodies 10 and / or cracks in containers, independently of where these foreign bodies and / or

Cracks are in the container.

The invention will now be described by way of example with reference to the drawings. FIG. 1 shows a side view, partly in section, of a device according to the invention; FIG. 2 shows a section along the axis II-II of Fig.l; 3 shows the side view of a part of Vor15 direction according to Fig.l; 4 shows the plan along the axis IV-IV of Fig.l; Figure 5 is a partially schematic oblique view of the light collecting device used; Fig. 6 is a circuit diagram of a part of the control circuit; 7 shows the circuit diagram of an auxiliary control device which is necessary for correcting the color density; 8 shows some typical test and control signals; Fig. 9 shows some test and other control signals generated in the control circuits; and Figure 10 shows test and other signals generated during the inspection of translucent containers.

The device shown in Fig.l to 4 contains a light source -1- and a converging lens -2- to produce a concentrated, narrow, band-shaped light beam -3-.

The light beam -3- is projected onto a rotating drum -4, on which are tangentially mounted twenty outwardly reflective silvered plane mirrors -5- each having an axial length of 2.54 cm and a width of 1.9 cm , The drum -4- is fixedly mounted on an axle -6- on which a pulley -7- sits. This is driven by a belt -8- (see Fig.3) coupled to a pulley -9- of an electric motor -10-. Drum -4-, electric motor -10- and the associated mechanism are housed in a housing -15- with an opening -16-, through which the light beam -3- through the housing -15- on the mirror-bearing drum -4- can fall.

When the light beam -3- strikes each of the rotating mirrors -5-, it is reflected down through an opening -17- in the bottom of the housing -15- onto three silvered plane mirrors -18, 19 and 20- and thereby supplies recurrent scanning light band -3a.

The reflected light band -3a is further reflected by the mirrors -18, 19 and 20- which are mounted at an angle in a vertical pillar -21- located under the housing. A rotatable table -22- (shown only partially), which is shown in phantom in Fig.l and 3, enclosing the column -21-. The pillar -21- has a vertical slit -23- through which the further reflected light band -3a can emerge from the pillar -21-.

It can be seen that, for example, due to the relative arrangement of the mirrors -18, 19 and 20-, a light band -3b - which falls on the higher side of the mirror -19- is reflected at a more extended angle to the horizontal than the beam. (Fig.4), when he leaves the deeper side of the mirror -18-.

The inclination of each of the mirrors -18, 19 and 20- against the vertical axis of the column -21- is adjusted so as to ensure that the subsequently generated light spot impinges on a translucent container with the most effective angle of incidence. It can be seen that internal light reflections in the container can cause erroneous readings. Furthermore, it must be ensured that critical portions of the container are sufficiently scanned and thus overlapping, as described above, may be desired during vertical scanning.

In Fig, 3 are photosensors -45- are provided, which are mounted in a conventional manner on spindles -46 and 47- and protrude into the column -21- through a slot. The photosensors -45- are operationally connected to a controller as will be shown later. The photosensors -45- are aligned so that they are the outer ends of the scanning

- 4 No. 364981

Light beam -3b-, which is reflected by the plan mirrors -18, 19 and 20-. It can be seen that the vertical positions of the photosensors -45- are adjusted on the column -21- with the spindles -46 and 47-.

The rotating table -22- carries a vertical cover with a vertical slot -26- (it is shown in Figure 4 in three positions). It serves to shield all but one part of the repeating light beam -3b- and thus concentrate this part on a narrow light spot -3c which is incident at an angle of 20 ° (see Fig. 3) when table -22 - and slot -26- turn. As a result, the concentrated light spot, hereinafter referred to as the scanning beam -3c, scans at an angle in the vertical plane and moves at an angle in the horizontal plane.

The bottle to be examined -27- is conveyed by fingers (not shown) around the edge of the rotating table -22-. They push the bottle against rotating rollers -28- which rotate the bottle as it passes through the scanning zone which is illuminated by the concentrated scanning beam -3c, which is substantially focused on the axis of the bottle -27-. The light falling through the bottle -27- is collected by a light collection device -30- which is connected to a photoelectron multiplier -34-. which passes the signal to a control circuit (see Fig.4). The light collecting device -30- and the controller will be described later in detail.

In Fig. 5 there is shown the light collecting means 30-30 having a plurality of optical fiber elements 31-31 mounted on the front surface of a panel 32 in a matrix arrangement. The optical fiber elements -31- are arranged in the form of a regular square matrix (partially shown in FIG. The other ends of the fibers are gathered together in a bundle, the end face of which is suitably shaped, optically polished and optically connected to a photosensor element -34-, namely a photoelectron multiplier tube. In front of the panel -32- is at a distance H a Lichtstreuschirm -35- mounted, which consists in this embodiment of a panel made of Überfangopalglas.

It can be seen that the scanning beam -3c is refracted and reflected as it passes through the bottle -27-, but for the sake of simplicity Fig. 5 shows the scanning beam -3c as a relatively narrow beam of light falling on the diffusing screen 35c he went through the bottle.

The scanning beam -3c- moves up and ah and passes through the surface of the scattering screen -35-. In order to compensate for a decrease in sensitivity as the beam approaches the edge of the screen, mirrors (not shown) are mounted on the edge of the panel whose reflective surfaces are perpendicular to the front surface of the panel and directed inwardly toward the array of optical fiber elements are.

The light collection device -30- described here consists of 247 plastic optical fiber elements each 60 cm long and 1.524 mm in diameter. They are arranged in a regular 13 19 matrix with 2 cm distance between the elements. The front parts of these elements are mounted on a black nylon board 28 by 40 cm and 6.35 mm thick with suitable fasteners.

The nylon matrix board is mounted in a narrow rectangular cabinet frame, parallel to and at a distance of 4 cm from a panel of 28 size coping glass. 40 cm. On the four inner surfaces of the support frame between the matrix board and the diffuser silver-plated mirrors are fastened at the front with which, as described above, a correction of the edge effect is achieved. The back ends of the optical fiber elements are tightly bundled, the end surface of the bundle is polished and located about 5 mm away from the photocathode surface of a 50 mm photomultiplier tube.

The power of the light-harvesting device can be adjusted by changing the parameters, eg number, cross-sectional area and distance of the individual optical fiber elements in the matrix, the distance between the front ends of these elements and the diffusing screen and the scattering properties of the screen itself.

For certain types of applications, it may be desirable for the device to have a graded or locally non-uniform sensitivity to the incident light. This can be done by a

The distance D becomes variable and depends on the location of the considered elements in the matrix, or in another case, the cross-sectional area of ​​the optical Graded fiber elements in the matrix,

In operation, the test device according to the invention is arranged next to or in a conveying path (not shown) of the bottles to be examined. The bottles to be tested are led to the rotating rollers -28- and while the table -22- turns against it and turned.

It can be seen that, during the sampling period, the scanning beam -3c passes through the

Cover slot -26- is continuously scanned in a vertical plane and simultaneously moved in a horizontal plane by the rotational movement of the cover member -25-. Furthermore, the bottle -27- is continuously rotated while passing through the scanning zone, wherein the sampling frequency is selected so that the total area of ​​the bottle is over-stroked to 25%. Sampling frequency and bottle rotation are synchronized so that between two consecutive samples

10% overlap occurs.

Should the amount of light falling on the photodetector fall below a predetermined threshold such that the light spot or beam is partially or totally obstructed or absorbed by a foreign body in the bottle, the output of the photomultiplier tube is reduced and thus generates an electronic reject signal indicating that a foreign object is in the bottle -27-. The precipitating signal generated thereby sets an excretion mechanism in operation as described for example in GB-PS No. 1,206,136. Through this, the bottle -27- from the return to the conveyor line (not shown) derived.

It will be appreciated that for successful operation of the test apparatus of the invention, it is necessary to synchronize and control the various operations. Some operations of the tester must be synchronized with the rotation of the rotating table -22-, ie with the carriage of the bottle -27- from the feed through the test zone to the outlet. Other operations must be synchronized with or controlled by the location of the scanning beam relative to the bottle being inspected. How these sync and control signals are received and processed will now be described.

Control signals related to the rotation of the table 22 are readily obtained in a conventional manner from any clock generator -69- (see Fig. 6), eg, electrical sensors located at the periphery of a rotating disk, which are just as many Has teeth or openings on its periphery, as are bottle stations on the rotating table -22-. For example, in the present embodiment, the disc has eight teeth and is driven by the rotary table -22- via a 1: 1 gear transmission. Thus, the necessary number of output pulses are generated as each bottle is transported around the table. These clock pulses are provided in the further description with the reference -TP- and corresponding reference numerals.

The clock generator generates signals for controlling the following functions:

1. Angular extent of the test zone during which a bottle is tested.

2. Clocking a sorting signal via a shift register memory before triggering the device for sorting.

3. Activation of the sorting device at exactly the right time to ensure accurate mechanical synchronization.

It is always necessary to recognize the beginning and end of the desired sampling of a bottle, so that the testing process is triggered during and only during this period. Furthermore, it is necessary to provide control signals for the actuation of adjustable encoders which provide different decay voltages as are necessary for the different heights of the scanning beam. This will be described later. Signals suitable for all these purposes are obtained from the photosensors -45-. They are referred to as sampling signals in the sequence and are provided with the reference symbols -SP- and reference numerals.

- 6 No. 364981

Before the operation of a device according to the invention is considered, will briefly discuss the problems that occur in the examination of translucent containers hei where the presence of foreign bodies or jumps to be detected. The signal output of the photomultiplier tube -34- varies in size as a function of

Amount of light falling through the bottle -27-.

The output signal of the photoelectron multiplier -34- shown in Fig. 8a can be decomposed into two components. One component consists of the voltage generated by the scanning beam -3c- which has penetrated the bottle -27-, the other component is caused by interspersed ambient light, which penetrates through the light shield and increases the amount of light striking the photoelectron multiplier, as will be shown. This signal component will vary depending on the intensity of such ambient light. Fig. 8 (f) shows the output of the photoelectron multiplier in which the proportion of the ambient light has been removed. How this portion of the ambient light is eliminated will be described later.

Fig. 9 (a) shows the output of the photoelectron multiplier -34- corrected for ambient light. This signal shows two effects of inclusions, namely the peaks -XI and X2-. In fact, these peaks are XI and X2 valleys because it is a negative signal.

In order to detect these peaks -XI and X2- in the test signal, the test signal must be compared with a predetermined comparison signal. How to obtain this comparison signal will be described later.

FIG. 9Cb) shows the test signal of FIG. 9 (a) but superimposed with a constant reference voltage. By this comparison voltage only one inclusion or jump is detected, namely that which corresponds to the peak -XI-, while the other, which corresponds to the

Peak -X2- corresponds, remains undetected.

From FIGS. 9 (c) and 9 (d 1) it can be seen that a variable comparison signal, as illustrated, for example, in FIG. 9 (d 1), is to be preferred over a constant comparison signal according to FIG. shows how this variable comparison signal is superimposed on the test signal. Both peaks -XI and X2- and thus the inclusions or 30 jumps that cause such peaks in the test signal are detected. The main reason for a variable comparison signal is that the amount of light, for example, by a bottle falls between the bottom of the bottle, bottle neck and bottle shoulder changes. A variable comparison signal is therefore not necessary if the translucency through a translucent container does not substantially change between one part of the container and another. 35 When examining stained glass containers, additional processing of the test signal may be necessary. In such containers, provision must be made to compensate for the variations in the total light transmittance that occur between one and the next container and are caused firstly by changes in the color density and secondly by changes in the wall thickness between the one and the next container.

Figure 8 (g) shows the test signal obtained from two similar bottles having slightly different color densities.

The test signal generated in Figure 6 at the load resistance -RI- of the photoelectron multiplier -34- is shown in Figure 8 (a). This signal -8a is supplied to the non-inverting input of a differential amplifier-Al. To simplify the description, 45 the various signals generated in the device are provided with the reference numbers of the corresponding drawings in which they are shown. From the sampling end signal SP2 generated by one of the

Photosensors -45- (Figures 8 (b) and 8 (c)), a control pulse SP4 is generated in a conventional manner and supplied to a switch -S1- to form a capacitor -CI- of a sample-and-hold circuit -60- (Fig. 8 (e)). The control pulse SP4 is passed through a delay stage -DL1- which closes the switch -S2- which connects the capacitor -CI- to the photoelectron multiplier -34-. After the test or sampling period, the switch -S2- is closed, the capacitor -CI- is now charged to a voltage corresponding to the transmission of ambient light. This voltage is again the inverting input of a

- 7 No. 364981

Differential amplifier -Al- fed until it is discharged at the end of the next sampling or test period. Accordingly, the signal 8 (f) is produced at the output of the differential amplifier -Al-. It should be noted that the signal 8 (fl represents the test signal for a "detachable" bottle.

The variable output signal adjustable encoders -61- include a plurality of variable resistors -VR3, VR4, VR5, VR6, VR7, and VR8-which are connected to the switches -S3, S4, S5, S6, S7, and S8-, respectively. This circuit is used to produce a variable acceptance signal 9 (d) for a particular type of container to be tested. Each of the switches -S3- to -S8- consists of a CMOS switch controlled by an ordinary RS flip-flop and an adjustable delay circuit of pulses originating from the photosensor -45. The signal 9 (d) is supplied to a capacitor -C3- and a resistor -R3-, which serve to suppress switching shocks. The signal 9 (d) is applied to the anode of a diode D1 and an adjustable amplifier -A3-.

If the voltage signal transmitted to point -62- is more negative, such as the voltage signal transmitted to point -63-, the voltage is clamped at -62- and at -63-. The voltage at the 1: 1 isolation amplifier -A2- is shown in 9 (d).

However, if a non-removable container is scanned, then the signal developed at the resistor -R2- and thus also the point -62- has the waveform 9 (a). By this waveform 9 (a), in two cases, the point -62- is not clamped at the potential of the point -63- during the test or scanning signal. As a result, the waveform 9 (a) is transmitted to the 1: 1 amplifier -A2- and from there to the non-inverting input of a differential amplifier -A3-. At the inverting input of the differential amplifier -A3- is the waveform 9 (d). Consequently, at the signal output of the differential amplifier -A3-, the waveform 25 9 (e) is applied to the resistor -R4-. Thus, during the actual sampling period, the output of this amplifier -A3- consists of only two error signal pulses relative to a zero volt level. This signal 9 (e) again is applied to the anode -70- of another diode -D2- and clipped in a conventional manner to the peak waveform 9 (f). This waveform 9 (f) is applied to the non-inverting input of a differential amplifier -A5- whose gain is adjusted with the resistors -R6 and R5- to provide an output signal of the waveform 9 (g) which is applied to the gate -Gl- , In this embodiment of the invention, the output is 15V and the smallest useful value is one This waveform 9 (f) is applied to the non-inverting input of a differential amplifier -A5- whose gain is adjusted with the resistors -R6 and R5- to provide an output signal of the waveform 9 (g) which is applied to the gate -Gl- , In this embodiment of the invention, the output is 15V and the smallest useful value is one This waveform 9 (f) is applied to the non-inverting input of a differential amplifier -A5- whose gain is adjusted with the resistors -R6 and R5- to provide an output signal of the waveform 9 (g) which is applied to the gate -Gl- , In this embodiment of the invention, the output is 15V and the smallest useful value is one

Error signal pulse about 0.25 V.

In Figs. 6 and 10, the sample start signal SPI and the check end signal SP2 are applied to a flip-flop EF1 to form an output pulse, namely, the test gate pulse SP3 supplied to the AND gate -G1.

In FIGS. 6 and 10, both the start pulse TP1 and the end pulse TP2 are supplied from the clock generator -69- to a flip-flop FF2-. The output of the flip-flop -FF2-, namely a gate pulse TP6, is applied to the AND gate -Gl-. It should be noted that the 40 start pulse -TP1- does not occur immediately, but one or more scans must be made on the bottle -27-. This is necessary, as will be described later, when color correction is required. The AND gate -Gl- feeds a conventional five-stage shift register -71-, to which again a conventional sorting control circuit -72- whose output is connected to a sorting device. The clock unit -69- also provides a control signal for 45 the five-stage shift register -71- and a signal -TF4- with which the Aussortiersteuerschaltung -72- is put into operation at the right time. Afterwards it sends another signal -TP5- which resets the sorting control circuit -72-. It should be noted that the

AND gate -Gl-as long as not working until all three signals are applied, namely 9 (g), SP3 and TP6. Then, a signal from the AND gate -Gl- is passed to the five-stage shift register, which again provides a signal for the Aussortiersteuerschaltung -72-, which is only then put into operation until it receives the clock pulse TP4. This gives you a precise synchronization.

If the color density and / or wall thickness of the translucent containers vary greatly from container to container, as shown for example in the Schwingungsform8 (g) must

- 8 No. 364981 certain corrections are made. In principle, it should be noted that either the test signal or the corresponding comparison signal must be changed to a suitable signal value in order to achieve a desired ratio between the test signal and the comparison signal. In other words, when the absolute value of the comparison or test signal doubles, it is necessary either to halve this absolute value of the test signal or to double the absolute value of the comparison signal.

FIG. 7 shows an auxiliary control circuit between the points -64 and 65-of the circuit shown in FIG. A control pulse SP6 coincident with the start pulse SP1 (see Fig. 8 (h)) is applied to a sample-and-hold circuit -73- consisting of two switches -S9 and S10- and a capacitor -C9-. Pulse SP6 closes switch -S9-, thereby discharging capacitor -C9- (see Fig.8 (kl).) Pulse SP6 is also supplied to a delay stage DL4. The output pulse SP7 of delay stage DL4 closes the switch -10-. The signal from the differential amplifier -Al- is now used from point -64- to charge the capacitor -C9-, as shown in Figure 8 (k) ), the solid line shows the absolute value of the signal voltage 8 (g) coming from the differential amplifier -Al-. In order to obtain a correct comparison with the comparison signal, the test signal must be amplified to the dashed value. Consequently, since the voltage applied to the capacitor -C9- at this particular instant is V2, this voltage must be boosted to the voltage V3. It should be noted that the delay of stage DL4 was chosen so that the scanning of the capacitor -C9- occurs only after a suitable period of time to set the sampling point at a selected vertical location on the container to be scanned. The control pulse SP6 is also supplied to a flip-flop FF3- which places the switches -SU and S12- from the position a to the position b, thereby disconnecting the auxiliary circuit from the main control circuit. The sample and hold circuit -73- passes the signal from the capacitor -C9- to a variable amplifier -A6-, this again to a conventional comparison circuit -CP1-, which is set to a reference voltage of, for example, 10V. The control pulse SP7 is fed via a further delay stage -DL5- a flip-flop -FF4-. The flip-flop -FF4- feeds a clock -CGI-, this again a divider chain, which is provided with the reference number -74- and constructed in a conventional manner. She rejects the which is provided with the reference number -74- and constructed in a conventional manner. She rejects the which is provided with the reference number -74- and constructed in a conventional manner. She rejects the

Flip-flops -FF5, FF6 and FF7- and a decoder counter -CT1- which operates in a predetermined sequence. The divider chain -74- controls a series of switches -S14, S15 and S16- which constitute part of the amplifier circuit of the variable gain amplifier -A6-.

In operation, the signal is supplied from the capacitor -C9- to the variable gain amplifier -A6-, and after an appropriate delay, the divider chain begins to operate and changes the gain of the variable gain amplifier -A6- in its value. Thereby, the voltage V2, that is, the output voltage of the capacitor -C9- changed until it is amplified to a value above the reference voltage of the comparison circuit -CF1- when the comparison circuit triggers. When the voltage V2 is boosted to the voltage V3, ie slightly above the reference voltage of the comparator circuit -CP1-, the comparator circuit -CP1- trips and stops the timer.

This trip action keeps the switches -S14, S15 and S16-in the desired position to keep the gain of the variable amplifier -A6- at the value at which the voltage V2 is sufficiently boosted to trigger the comparator circuit -CCP1-. Finally, the signal from the comparison circuit -CP1- is applied to the flip-flop FF3-, which places the 45 switches -Sil and S12- back in the original position a. In principle, therefore, between the points -64 and 65- an amplifier circuit.

It should be noted that the start pulse TP1 (see Fig.10) does not appear until some time after the first sampling start signal SP1. Consequently, the gate pulse TP6 is the AND gate -Gl- supplied some time after the start of the pulse SP6. This will lock out the device during this initial period.

It should be noted that the gain of the variable amplifier -A6- is now set so that the output signal, namely the test signal of the differential amplifier --A1- has been amplified, so that the voltage is above a preset value. This projection is repeated whenever a new container enters the test zone. This will be

It is ensured that the selected test or scanning signal voltage of each container is set for a predetermined value. Thereby, the same comparison signal can be used regardless of variations in light transmittance occurring in various containers due to differences in color density or wall thickness.

If, for some reason, the input signal coincides with an inclusion or jump in the container, then the counter -CT1- is set up to automatically reset the flip-flops -FF3, FF4, FF5, FF6 and FF7- after one full revolution, to set the amplification factor -1- on the variable amplifier -A6- and simultaneously to close the -Sil and S12- switches. Consequently, when this unamplified signal is supplied from the variable amplifier -Al-, the control circuit of Fig. 6 operates to reject the container.

In the embodiment shown in Fig. 7, the variable amplifier -A6- is designed to have a maximum amplification factor of 4 with 50 discrete steps. The comparison circuit -CP1- triggers at 10 V. Consequently, a signal between 2.5 and 10 V supplied by the variable amplifier -Al- can be processed by this control circuit.

Thus, this circuit can handle variations between 2.5 and 10 V in 50 stages, ie it can detect variations of 0.15 V in the test signal.

As a critical change in wall thickness or color density, that change may be defined as a result of an inherently acceptable container would fail only because the light transmission through the container would be too low due to this change, or undetected because of this change or crack in an unacceptable container stay because the main light transmission through the container would be too large.

Although, in the embodiment described above, a spotlight beam performs both vertical and horizontal scanning in the inspection zone, the invention may be applied to a device whose spotlight beam scans only in the vertical direction.

The embodiment described above is used for containers in which color density and / or

Wall thickness vary significantly from container to container, so the Abtastprüfsignal can be changed with a setting factor. However, it will be appreciated that in a similar manner, the corresponding pick-off signal may also be varied with a set-up factor to obtain the desired ratio between the test and compare signals. Although the ambient light correction has been made on the test signal, it can also be performed on the acceptance signal. This can be done by an electrical addition of the ambient light signal to the comparison signal, whereby the effect of the ambient light is switched off during the test.

PATENT CLAIMS:

1. A device for detecting foreign bodies and / or cracks in translucent containers, with a transport device for successive transport of the container by a

A test zone, means for rotating the containers about an axis while passing through the test zone, a light source for producing a spot beam scanning the containers as they pass through the test zone in the direction of the axis, a light collecting means for receiving the light after passing through the containers the light collecting device downstream circuit arrangement for generating an electrical test signal ent40 the Lichtdurchlaßvermögen of the respective container, a circuit arrangement for generating a comparison signal and a comparator circuit for comparing the test signal with the comparison signal and controllable by the output of the comparator circuit means for sorting out of containers, characterizedin that for the stepwise variation of the size of the comparison signal (9d) as a function of the position of the scanning light beam (3c) along the axis of the

Container (27) adjustable level sensor (VR3 to VR8) are provided, which are successively switchable from the instantaneous position of the scanning light beam detecting sensors (45).

2. Apparatus according to claim 1, characterized in that between the circuit arrangement (34) for generating the test signal and the comparator circuit arranged control amplifier circuit (A6, CP1, FF3 to FF7, CGI, CT1, DL4, DL5, S9 to S16) is provided Wall 10 No. 364981 compensates for changes in the light transmittance of successive containers (27) caused by light and / or color to a predeterminable value by automatic readjustment of the test or comparison signal (8g or 8d).

3. Apparatus according to claim 2, characterized in that the control amplifier circuit 5 has a sample and hold circuit (73) for storing an instantaneous value of the test signal (8g) that the sample and hold circuit (73) via an amplifier (A6) adjustable Amplification is connected to an input of a comparator (CP1) whose other input is connected to a reference signal source (VRef), that the gain control of the amplifier (A6) from the output of the comparator (CP1) is controllable (74; S14, 15, 16) , and that the input of the amplifier (A6) after completion of the gain adjustment to the output of the circuit arrangement (34) for generating the test signal is connectable.

4. Device according to the