SE429163C - DEVICE AND PROCEDURE FOR CONTINUOUS COMPENSATION OF ERROR SIGNALS BY ELECTROTOPIC SATURATION OF A TWO DIMENSIONS EXISTING - Google Patents

Description

'YSOZSÉS-O 2 l0 15 20 25 30 35 40 Some of the work data, which was used in computer-controlled rolling mill and referred to here are: desired bar iron diameter or setpoint, full and half intended tolerance for commodity and bar iron quality or carbon content in the bar iron. Some of the most important measurements during work are: current bar diameter or bar dimension, current bar side profile or bar profile and a histogram of bar size measurements nings. Furthermore, during rolling, the bar temperature is measured, a parameter, which was used to compensate for the thermal expansion ' both for bar measurement and for the rolling control's data control.

To program the rolling mill computer to meet the strict requirements for rolling speed, bar dimensions and tolerance halving, it is desirable that all measurements The thunder roll has the following properties. Rod thickness measurement It must be possible for the bar to vibrate laterally web while moving in its longitudinal direction during rolling; it must be possible to do so under reconstructable conditions at about 300 Hz; they shall have a resolution of 0,0l27 m; they must have an absolute accuracy equivalent to a quarter of ordinary commercial tolerance) they must show a high degree of reliability; all measurements must be able to be performed under the difficult working conditions, which normally prevail in one bar rolling mill. The bar iron temperature measurements shall have similar characteristic properties. A histogram of the bar measurements also obtained through the system.

There are several different measures for measuring bar dimensions types of electro-optical measuring systems available. One earlier known bar iron measuring system works according to the self-illumination principle, whereby chopped infrared radiation from the hot rod iron formed by a lens on an infrared detector. Elemental edging detection circuit technology was used in an attempt to determine the more detector pulses relative to the rod edges.

Three newer electro-optical systems, which can be applied to measurement of bar iron dimensions, working according to the system that from behind illuminate the test object to be measured and imaged a shadow of the object through a lens on an electronic camera.

In one of these measuring systems, a sensing laser beam illuminates the test object and the lens system focus on the object's shadow a phototransistor. In another such measuring system, I illuminates a stationary light source with constant intensity test object 10 15 20 25 gzo 35 40 ß 7802955-0 and the lens system focuses the object shadow on an electronic sensed image pickup tube, which has two sensing shafts same direction. In the third mentioned system is replaced the image pickup tube of a self-sensing photodiode array.

The light-sensitive device in each of the three The backlit metering systems generate a primary camera impulse with a width that approximates the dimensions of the object between the shadow edges. The primary camera pulses are processed in edge detector circuits, which have either simple or pulse differentiation bodies, which further try to more precisely define the width of the camera pulses in relation to the object dimension.

Two additional electro-optical measuring systems are available which combine the above solutions to measure rod profiles. One of said systems combines two self-illuminated cameras, which are fixed at a right angle to each other arranged perpendicular to the working direction of the bar rolling mill.

This system makes only two bar diameter measurements, twisted 909 relative to each other, but do not provide bar iron nings. The second of said electro-optical bar profile measuring systems include two rear-facing cameras, which are in the right angle relative to each other mounted on a sensing device, wherein two rod diameter measurements and a sensing position measurement is indicated separately and / or stored in a multi-channel register during peripheral sensing of the rod.

All of the above known electro-optical sizes and profile measurement systems have had varying successes at special installation types. However, none of these measurement systems completely satisfactory when used as a measurement system for bar dimensions and profiles at current high speed rolling mill for hot rod. Such measuring systems can handle one or several of the measurement requirements not mentioned above.

The difficulties with known measuring systems persist first in that the measuring object must be placed in a certain position in the camera field of view. Secondly, they have cameras with too low a response speed and / or resolution. Third, can they do not combine high-speed system precision, since significant switching disturbances occur at these measuring velocities and differentiation disturbances also prepare special concerns. In addition, there are various environmental factors 7802955-0 _ 4 10 15 20 25 30 35 electrical disturbances exist to varying degrees, which further complicates the problem of performing definitive bar measurements at high speeds and with high reliability similarity. Fourth, incapacity or lack thereof is added ability to correct sources of error such as optical and electronic nonlinearities, which degrade the measurement system precision. Fifthly, there is instability, which ï causes a shift in the system calibration. Sixth there may be an inability to achieve a meaningful indication and graphical representation of bar diameter and profile information in the cold state regarding different peripherals modes for either the rolling mill driver or a control computer for the rolling mill. Seventh, there may be no possibility of provide a bar measurement histogram. For the eighth finally, there may be an inability to compensate or correct the magnitude of distortion due to high frequency vibrations in the transverse direction of the rod.

The main object of this invention is to propose one improved electro-optical measurement method and system. This should further have a high reaction rate according to the invention, high measurement repetition ratio, high accuracy, large stability eeh / or high reliability in a work environment with both high speed and high temperature as in modern bar rolling mill. IN Another object of this invention is to provide an improved electro-optical measuring method and apparatus, which allows accurate measurements of objects, which are placed in arbitrarily position in a camera's field of view, even if the subject vibrates transversely the direction of movement of the object.

Another object of this invention is to provide an improved electro-optical measuring method and system, which determines both the size of the object and the varying object position in the camera's field of view.

Another object of the invention is to achieve an improved electro-optical measurement method and system, which processes a camera signal to remove interference, combined with an object size pulse in the camera signal, thereby allowing exact definitions of the object size pulse and / or the object position in the camera's field of view. 10 l5 20 30 40 5 7802955-0 Another object of the invention is to propose a improved electro-optical measurement method and system, which corrects the camera's object size signals for optical and electronic nonlinearities and / or other sources of error.

Another object of the invention is to propose a improved electro-optical measurement method and system, as graphical reproduces and displays and / or stores one or two relative to each other perpendicular dimensions of an object and / or the profile of the object in one or more peripheral positions of the object.

A further object of the invention is to propose one improved electro-optical measurement method and system, which draws the profile of an object and displays and / or stores the drawing superimposed with one or more of the object's commercial tolerance data.

Another object of this invention is to propose an improved electro-optical measurement method and system, which draws and displays and / or stores one or more histograms from the measuring system.

A final object of this invention is to propose one improved electro-optical measurement method and system, which draws a profile of an object and / or a measuring histogram, as appropriate for use in a computer-controlled process.

The above objects are advantageously achieved for the in a bar hot rolling mill e.g. through a computerized, electronic optical system for measuring either one or two at right angles superimposed dimensions of a fed and vibrating hot rod profile either in stationary position or in different peri- fera locations. One or more rear-facing electronic camera heads were used, and these are mounted 900 separately on a separate sensing device for measuring two dimensions. Each camera X) and head is equipped with electronics that include camera AGC a common digital two-way sweep generator for simultaneous single-axis sensing each camera. Additional electronics process one bar shadow pulse in pulse flow fi detection circuit technology, which has a autocorrelator to eliminate interference. Other electronics units include a digital accumulator which provides one digital bar size signal and signals indicating the bar position in the field of view.

Each camera's bar size and bar position signals, a sensing position signal, bar temperature signals and others x) AGC = automatic gain controll vsozsss-o f 6 10 15 20 25 30 35 signals are converted by a digital computer, which is programmed to perform the following functions either "off-line" or "on-line": First, the correction of each bar size signal through digital compensation for field of view errors, others optical and electronic nonlinearities, rod temperature and other sources of error, with very accurate rod diameter measurements anywhere within the field of view is implemented. Second "on-line" calibration of the measurement "off-line" and automatic recalibration of the measurement to remove the calibration deviations and eliminate errors. Third, allow manual driving or automatic control of the sensing device and incremental digital storage of corrected bar diameter measurements for each camera during sensing. For the fourth to facilitate interaction with cathode ray tubes and printer terminals to indicate and / or store: (a) the bar diameter of each camera; measurement was anywhere in the sensing field; (b) use of stored bar diameter data and running of the data head, plot the bar profile deviation from the setpoint, where the drawing is superimposed with commercial full or half tolerance data; and (c) a histogram for each measurement and a measurement difference cleaning histogram. The computer is arranged to send profile and histogram data to a rolling mill's control system if requested by control system.

The invention will now be explained in more detail in connection to the accompanying drawings. On the same shows Fig. 1 is a block diagram of a fully computerized electro-optical 'system for measuring a dimension in an embodiment form of the invention Fig. 1A is a block diagram of the fully computerized electrode the optical system for measuring two dimensions, which includes dual cameras on a sensing device for determining the cross section and display a second embodiment of this invention. This scanning device can also be used in embodiments but according to Fig. 1. Fig. 2 the cross section of a rod with maximum and minimum tolerance limits with dashed circles and include a four-plane overlay relative to the bar profile orientation 10 15 20 25 30 35 40 fig. fig. fig. fig. fig. fig. fig. fig. fig. 3 \! C \ 10 12 7 7802955-0 the bar profile deviation printed by the computer over the angular positions of the sensing device in relation to the four-plane overlay of Fig. 2, provided in the embodiment of Fig. 1A and inclusive a work data start. A similar print of the bar profile can be obtained with the device according to Fig. 1 when using the sensing device according to Fig. 1A a block diagram of the camera electronics for each camera head in the two camera systems shown at the embodiments of Figures 1 and 1A. a section through a dewormed photocathode, which was used in an image analyzer tube such as that used the camera electronics of Fig. 4 a cross section through the photocathode shown in Fig. 5 a block diagram of a two-way sweep generator which was used in the camera electronics shown in Fig. 4 a time diagram of pulses generated by two-way sweep generator, master clock, window pulse the generator, and those of the camera electronics according to Fig. 4 showed the AGC extinguishing circuits a block diagram of the one shown in Fig. 4 camera electronics use the camera pulse processor. a block diagram of one at the one shown in Fig. 9 given the camera pulse processor using the autocorrelator a time diagram of various primary camera signals, differentiation units, autocorrelators and bar pulses, which occur in the pulse processor according to Fig. 9 a circuit diagram of a photomultiplier tube (PM) and an automatic gain circuit (AGC) reproduced in a camera self-balancing measuring circuit, which is included in the camera electronics of Fig. 4 a block diagram of a bar size and position accumulator, as used in it the camera electronics shown in Fig. 4 a block diagram of it in the embodiment the one-dimensional measurement system computer shown in FIG and includes references to it computer program 7802955-0 10 15 20 25 30 35 Fig. 14A Fig. 15 Fig. 16 Fig. 16A and 16B Fig. 17 Fig. ia Fig. 19 a block diagram of the computer displayed at two the dimensional measurement system which has a wide the device according to Fig. 1A and includes references to it computer program. The profile and mode programs can also be used for the sensing device for the embodiment the shape of Fig. 1 a computer disk card for the embodiments according to both Fig. 1 as 1A I a computer core card for the embodiment of Fig. 1 computer core card for the embodiment of Fig. 1A a typical histogram table printout for application 7 of the present invention in the embodiment either Fig. 1 or 1A a typical profile table, which was used for printing of the Fig. 3 profile of Fig. 3 in the embodiment according to Fig. 1A of the present invention a typical flow chart showing the computer of Fig. 1 and 1A during communication with a control system, which uses one or two histogram tables according to the present invention, as needed in the embodiment according to either Fig. 1 or 1A, and further it includes a profile table for use in the embodiment of Fig. 1A Description of preferred embodiments One-dimensional measurement system In particular, Fig. 1 shows a computerized electro-optical system for measuring a bar dimension, which has a rear view illuminated camera, which is arranged in a bar rolling mill. The measuring system measures the diameter of a bar iron 10, for example in a transverse position tion, mounted behind the outlet side of a pair of rollers ll. As explained below, the bar diameter signal is applied to a computer, which curves the cross-sectional dimension of the rod 10. Finally 10 15 20 25 30 35 40 9 7802955-0 bar diameter data is indicated, stored and transmitted to a rolling mill control system, which uses this data to set insert the gap between the rollers in the roller pair ll and create the brush size of the rod 10.

More specifically, a lamp 30 is arranged opposite the electrical tronic camera head 3l so that when the rod l0 breaks the light from the lamp 30, a rod shadow with one against the rod diameter proportional width in a lateral position is depicted on it electronic camera head 31. A typical arrangement of rear-illuminated camera head is shown in Fig. 4 and described below.

The lamp 30 is arranged to provide a light source in right angle to the bar 10, larger than the largest bar size to be measured in the field of view of the camera. For example, the above camera's field of view should be 7.62 cm and the light source used l0, l6 cm. In addition, the wavelength and brightness of the lamp must be compatible with the sensitivity of the electronic camera head 31 characteristic. Thus, for example, blue light from one DC-supplied fluorescent light source is suitably used for electronic camera head described above.

The shadow of the bar 10 together with excess light outside the edges of the rod 10, coming from the lamp 30, causes the electronics camera head 31 to generate a camera signal. This signal consists of a primary camera pulse, mixed with interference, which is fed across line 34 to the camera niken 35. As described below in connection with Fig. 4, is processed the camera signal to remove the interference and cause digital bar size and bar position signals, such as via cable 36 is supplied to the computer 27. Measurement and other signals fed via the cable 37 from the computer 27 to the camera electronics 35.

The computer 27 in the present electro-optical bar measurement system also receives digital brush size signals for the rod 10 via thumbwheel selector 42 over cable 43. Brush size signals, according to the example 4.445 cm, was used to determine the bar size. deviation of the game and for other purposes described below.

In addition, the computer 27 also receives a digital composition signal from the thumbwheel selector 44 via the cable 45. the signal, which here, for example, represents 0.230% carbon in steel, was used as a factor in correcting the bar iron 'zsozsssc-o 1 ,, 10 15 20 25 30 35 40 the size of thermal expansion and others described below purpose. Furthermore, the computer 27 is also supplied with suitable order data. signals, including date, time and size tolerances for the bar iron 10 from the source 46 via the cable 47. Alternatively can one or all brush size signals, composition signals and other data signals are fed by a control system, which is directly connected with the rolling of the rod, depending on the wishes one has.

To achieve temperature corrections at diameter measurement of moving hot rods was used one Land Co. optical pyrometer head 48 adjacent the sensing device l2 and directed towards the hot, moving iron bar The optical pyrometer head 48 is arranged to generate a highly sensitive primary temperature signal, as via cable 49 is supplied to the pyrometer electronics 50. The primary the temperature signal is corrected by the gradation and linearization circuits in pyrometer electronics 50. The corrected temperature the signal, for example 910 ° C is fed via the cable 51 to the digital input indicator 52. In addition, the corrected temperature signal is supplied via the cable 53 to the computer 27, where it was used for compensation of the temperature expansion of the hot iron bar 10. 5 Installation problems can prevent a Land Co. optically pyrometer head 48 and pyrometer electronics 50 from that providing a corrected temperature signal to the computer 27 and indicator 52 with ömäæd accuracy and sensitivity. If so case, an alternative to Land Co. the pyrometer arrangement be to replace them with an optical field sensing meter system according to US PS 4,015,476. The optical field sensing the pyrometer system briefly consists of a fast oscillating mirror, which is arranged in a pyrometer head and intended for a field of view through which the hot iron bar l0 passes.

The hot iron bar is imaged through a slit on a highly sensitive, infrared detector in the pyrometer head. The infrared detector feeds a peak value detector as well as collection and holding circuits to measure and store a non-linear signal for the bar temperature.

The stored, non-linear signal is fed via the cable 53 to computer 27, where it must be graded and / or linearized.

The stored temperature signal is updated for each sensing of the oscillating mirror, for example, every 20 ms through a busy / ready signal via the dashed cable 54. 10 15 20 25 30 40 11 7802955-0 The stored temperature is also displayed on the correct scale and linearized with less frequent updating and added bar temperature indicator 52. Steps to adjust field scan frequency and field of view width to achieve an adaptation to different installations.

Another feature of the present bar measuring system is one automatic recalibration system. As described below, this function is initiated each time it is detected that the end of the advanced hot iron bar 10 leaves the pair of rollers ll. For this reason, hot metal detectors 55 detect the presence respectively the absence of hot rod material 10 and feeding one corresponding signal via line 56 to the hot metal detector electronics 57. An attendance / absence signal is supplied via the cable 58 to the computer 27 where it starts the above-mentioned automatic recalibration system.

All camera signals, setpoint signal, composition signal, other signals, the temperature signal and the the metal presence / absence signal, which is supplied to each cable 36, 43, 45, 47 53 and 58 are compiled by the computer 27 for to form a number of functions under the control of connected or stand-alone computer programs, as referenced below to. One of these functions is to feed bar diameter data, bar deviation data superimposed with commercial tolerance references from the computer 27 via the line 59 to the cathode ray tube (CRT terminal 60 and to accept interaction between a standard control panel on terminal 60 and computer 27 via a : abel 61.

Another function of the computer 27 is to feed rod diameter-- data and roll start data from the computer 27 via the cable 62 to the printer terminal 63 and to accept the interaction between a standard control panel on terminal 63 and the computer 27 via a cable 64. The printer terminal 63 produces a printing 65 as a data log. Another feature of computer 27 is to feed bar diameter data and a measurement system history grams via cable 66 to the control system 67 in response to a corresponding interrogation signal to the computer 27 via the cable 68.

Pig. 2 shows a sectional diagram illustrating the bar 10 cross profile. Dashed circular lines 69 and 70 illustrate the maximum and minimum standard trading tolerance for the set diameter. The brush size is, for example, 4,450 cm. Other characteristic features of Fig. 2 are described below with reference 'zsozsss-o 12 10 l5 2o_ 25 30 35 40 to the embodiment of Fig. 1A.

It should be noted that the display on the CRT terminal 60 is essentially the same as printout 65. Thus, the CRT the terminal 60 bar diameter information in a shape that is unique and very useful for the bar measurement system operator as for the machine operator for a rolling mill, where the bar measurement Applied.

Electronic camera head A typical, rear-illuminated, electronic camera head, used in the electro-optical bar measurement system according to Fig. 1 is shown in Fig. 4 as a camera head 31, located along an optical axis on the side of the rod 10 facing away from the lamp 30.

This arrangement illuminates the field of view 80 and provides one bar shadow 81, which varies vertically in proportion to? the width between the edges 82 and 83 of the glowing rod iron.

An end view of the iron bar 10 makes it appear stationary but in reality the iron bar 10 vibrates in a path 84, while it is fed forward in its longitudinal direction with speeds up to 20 m / sec. For this reason, the bar shade 81 varies not only vertically proportional to the rod size but also shifted horizontally and vertically within limits of about 76.2 mm in the bar diameter web 84. This phenomenon requires a larger field of view 80 than a stationary bar, which increases the difficulty of accurately measuring the bar.

Since the bar shadow 81 varies vertically and its position varies both horizontally and vertically, is the camera head 31 provided with a telecentric lens system 85, which is performed to allow only parallel light beams with one focal plane, extending at least from the bar web 84 nearest horizontal edge to its furthest point horizontal edge. This is achieved by a seven-element lens 86, which has 10.16 cm eye field 80 within which 76.2 mm bar web 84 is vertically centered. Other properties of the lens 86 includes a 2: 1 image reduction and a telecentric one aperture 87 with a very narrow horizontal optical aperture 88, through which the rod shadow 81 is projected. The bar shadow 8l transmission is limited by an optical filter 89 to let through only blue light from the lamp 30, whereby undesirable effects from other light sources with other wavelengths in the field of view are eliminated. 10 15 20 25 30 35 40 13 7802955 -0 The telecentric lens system 85 thus provides a horizontal tentatively oriented rod shadow 81, which varies vertically between edges 82 and 83 and remains sharp at the focal point, while the rod 10 vibrates within the web 84. The rod shadow 81 has the same size along the optical axis but because it displaced vertically from the optical axis in different directions, it becomes larger according to a non-linear function. This phenomenon caused by a combination of electronics, the core and nonlinearities of the lens and is referred to as a field of view errors corrected by the computer 27, as described below.

It transmitted through the telecentric lens system 85 the bar shadow 81 is imaged on a photosensitive device 90, which can be sensed at 300 Hz, has a resolution of at least l at 10,000 and has a high sensitivity to blue light. Preferred For example, the device 90 is an image analyzer having a photocathode 91 with a central image transfer surface, which receives the imaged the bar shade 81. The photocathode 91 is arranged behind a light permeable layer at the working section of the image analyzer tube 90.

From the photocathode 91 emitted photoelectrons are focused through external means and passes through an electron aperture 92 so that they can enter the photomultiplier section of the image analyzer.

Preferably, the device 90 is an ITT Co., USA, high resolution image dissector tube no. F4052RP.

The camera head 31 also includes a cylindrical deflection device. and focusing coil device 93, which surrounds the image analyzer 90 cylindrical body. The coil device 93 comprises separate deflection coils for the Y-axis and the X-axis and a focusing coil, all of which are fed from separate, external power sources.

Standard mymetal metal shielding surrounds the coil device 93 cylindrical outer wall and thereby provides an effective shielding against radial magnetic fields. A preferred coil device 93, designed for use with the above image analyzer tube 90 is Washborn Laboratory, Inc., U S A, no YF2308-CC3C.

Sometimes the standard mymium shielding in Washborn means Laboratory Inc. flushing device 93 none satisfactory shielding against both radial and axial magnetic field sources.

For example, if the image analyzer tube 90 operates at high sensitivity, level and an electrical device, which generates a powerful magnetic field, displaced near the measurement, can the output signal from the image analyzer tube 90 changes. About this relationship 7802955-0 - '14 lO 15 20 25 30 35 409 meet in practice, there is an alternative solution, which requires one a modification of the Washburn Standard Myme Metal Shield for to improve the attenuation of axial magnetic fields. This means essentially an extension of the cylindrical Standard Washburn the mymetallic shield axially to the lens system 85 and one of ~ closing the end face at the filter 89 except for one optical aperture to image the steam shadow 81 on the photocathode 91 in the image analyzer 90. Further attenuation of axial magnets field can be achieved by a second cylindrical mymetallic mantle around the extended standard shield. Furthermore, can the standard coil shield is used without extension but a axial field damping is achieved by adding a second and possibly a third cylindrical mymetal shield with the same axial length as the first. The present electro-optical bar measurement systems may undergo other calibration offset and changes in the optical the image analyzer tube and other electronic nonlinearities which is part of the bar measurement system. These changes and varying measurement conditions can be identified by performing counter dependency calibration checks and consequent corrections of the calibrated bar signals as described below. These calibration checks are made possible by modifying the image the analyzer tube 90, which is provided with a dewormed photocathode 91 as shown in Fig. 5. As shown in Fig. 5, the dewormed photocathode includes 91 patterned, the image non-evaluative surfaces adjacent to the image values the surfaces. More specifically, the calibration masks 94, 95 made by selective placement of the usual photocathode 91 photosensitive materials on an image-permeable glass sheet 96 using a precision mask to shape the calibration the reference patterns. For example, the calibration mask 94 may pass of a simple, 6.35 mm large mask, which is centered on the photocathode 91 right side. The calibration mask 94 is referred to as the "right mask". and can be used for direct control of the bar measuring system calibration offsets during the RTMASK described below computer program. The calibration mask 95 is assumed to consist of five 2.54 mm large masks at 2.54 mm intervals on the photocathode 91 left side. The calibration mask 95 is designated therein following as "left mask" and can be used for direct control of variations in the optical and electronic bar measurement system nonlinearities under the computer program described below 10 15 20 25 30 35 7802955-0 15 LFTMSK. Figure 6 is an enlarged cross-section through Figure 5 for to display the 94 mask recess in the masked photocathode 91. The recess extends to the 90 glass surface of the image analyzer tube 96.

Whenever the bar measuring system works, one is added single-axis bi-directional sweep signal coil for deflection along The y-axis and a fixed current are applied to the focusing coil as determined. written below. During normal bar measurements, the coil for X- X-axis deflection no current. This causes the Y-axis sensing to pass the "C" sense, i.e. the central of the photocathode 91 image interpretation area as shown in Fig. 5. Whenever the detector 55 notes that there is no rod 10 within the camera field of view, the computer 27 selects either the right one or the other left calibration mask 94 and 95, respectively, by applying a positive or negative bias voltage, which is applied to the deflection the coil for the X-axis. This X-axis bias voltage shifts the photocathode 91 Y-axis sensing to corresponding "R" sensing and "L" sense on either side of the "C" sense, as shown in Fig. 5.

The X-axis bias has the effect of shifting the right calibration mask 94 or the left calibration mask mask 95, over the electron aperture 92 in the image analyzer tube 90. When the simple X-axis sensing voltage is applied the deflection coil for the Y-axis, the image of the right is moved or left calibration mask 94, 95 effectively up and down over the electron aperture 92 in the same manner as the current one the rod shadow 81 moves in the "C" sensing position.

It can be noted that the original camera pulse in line 34 has the same pulse width regardless of whether the right one or the left calibration mask 94 or 95 is selected by computer 27, which occurs when the bar shadow B1 with corresponding size and position are imaged on the center surface of the photocathode 91. Follow-up the masked photocathode 91 can efficiently provide an on-line check of the bar measurement system shootings and also changes in optical and electronic nonlinearities.

Camera electronics A typical camera electronics used in the electronics the optical bar measuring system according to Fig. 1, is shown in Fig. 4 as camera electronics 35. Its details are best understood by 78Û2955'Û M 10 15 20 30 35 40 refer to Figures 4 and 7-13. All electronic components in it are conventional fixed devices and include TTL (transistor-transistor-logic) logic element, where logic sym- boler indicates or presupposes their use.

In general, Fig. 4 shows a two-way sweep generator 97 which is shown in Fig. 7 and is referred to in the time diagram of Fig. 8.

The sweep generator 97 includes a 12 MHz crystal oscillator 124, which provides a square wave train of clock pulses 8A for the entire electro-optical bar measurement system. Except for current measurements of emitted bar pulses, are all digital operations synchronized by the clock pulse 8A, in addition to the two-way sweep signal 8E and the sweep reset pulse 8D. The latter two pulses are generated in sweep circuits with about 300 Hz. The clock pulse 8A and the bidirectional sweep signal 8E is synchronized by the scan reset pulse 8D at each sweep cycle, so that the sweep signal 8E can be divided for each arbitrary using an appropriate submultiple of the clock pulse 8A. The clock pulse 8A was used for current measurements while pulses for other bar measurement system requirements are obtained by divide the clock pulse SA all the way down to the bidirectional one the 8E frequency of the sweep signal. It should be noted that the clock pulse BA and the absolute frequency of the bidirectional sweep signal SE value is not critical because the bar measurement system is calibrated by placing standard bars in each camera's field of view.

However, sweep stability and sweep linearity are high degree critical, as they directly affect the bar measurement system precision. IN The main clock shown in Fig. 4 receives a clock pulse 'train 8A with 12 MHz and sweep reset pulses BD with 300 Hz from the sweep generator 97. The master clock 98 includes the buffer memories, digital counters, division devices and logic circuits to supply all synchronized pulses, such as used in camera electronics 35 for timing and measurement purposes. These also include intermediate stored 12 MHz clock pulses 8A, intermediate stored 300 Hz scan recovery pulses 8D. Ytterli- more pulses generated thereby is a 300 Hz fast sensing pulse SH with short duration and a computer-ready pulse, similar pulse 8H but with a longer duration. The computer-ready pulse is output on line 99 and the other pulses transmit theirs identity to other circuits shown in Fig. 4. 10 15 20 25 30 35 40 17 '7802955-0 A window generator l00 receives the clock pulse 8A of 12 MHz from the main clock 98 and generates by means of gates and logic circuits window pulses 8F once on each half of each reciprocating sweep cycle, which is shown in the time diagram in Fig. 8. An inverted window pulse 5? is also generated. The both window pulses BF, ÉF are applied to the others described below the circuits. The width and duration of the window pulses 8F, ÉF are determined through a control pulse via line 101, which is supplied from the computer '27. In short, the BF, ÉF width of the window pulses is related to it time required for the sweep signal 8E to sweep over only the photocathode 91. It is only a larger part of each upper or lower half of a full 300 Hz sweep cycle. For example. if the camera's field of view is 76.2 mm and the lens is 10.16 cm, as they is here, then 76.2 mm field of view is depicted downwards centrally for to cover the entire electrode of the photocathode 91. Over-sensing of the photocathode 91 is obtained in each of the upper and lower halves thereof bi-directional sweep cycle 8E. This over-sensing is even divided into two intervals at the beginning and end of each upper and lower half of the bidirectional sweep cycle 8E. Thus is the sum of the window pulse 8F (approx. 75%) and the over-sensing (approx. 25%) duration equal to the length of each upper and lower half in a bidirectional sweep cycle 8E. As an alternative arrangement, the width of the window pulse can be set manually by selective gate means, not shown, to replace the computer 27 control signal in line 10l.

Under the computer 27 program RTMASK described below, LFTMSK, GAGRCL, and CALIBR program the window generator 100 via line l0l to modify the BF of the window pulses, ÉF normal size and duration. Under RTMASK and GAGRCL window pulse size and duration of the right calibration the size and location of the mask 94 in Fig. 5. Under LFTMSK generated five for each size and location of the left the 95 elements of the calibration mask dimensioned and timed window pulses one by one to selectively cover the entire it left calibration mask 95. Under CALIBR, the window the size and time of the pulses selectively for the right calibration ~ the size and location of the mask 94 and for each of them five left calibration masks 95. The normal window pulses SF, Éï size is set through the subprograms described below GAGEIN.

The bidirectional sweep signal 8E is applied from the sweep generator. '78 () 2 $ ”55" Ü lg l0 15 20 25 30 35 40 the rator 97 a Y-coil deflection driver 102 and the vertical one or the Y deflection coil in the coil device 93. Constant current from the focus coil power source 103 is supplied to the coil device 93 focus coils. The size of the focus stream is adjusted so that all electrons emitted from each point on the surface 9l of the photocathode focused on a single corresponding point in plan with the electron aperture 92.

The V X coil driver 104 is connected to the horizontal or the X-deflection coil in the coil device 93. Below normal bar measurement work, no efficient current is supplied to the X- the deflection coil. Therefore, the vertical single axis of the Y-axis sensing appear as "C" sensing centrally on the photocathode 91 image interpretation surface, as shown in Fig. 5. During calibration controlled by the computer 27 during the programs described below RTMASK and LFTMSK; Positive and negative bias voltages are applied 'alternately via control lines 105 and 106 from the computer 27 to The X-Coil Driver 104. This causes easy detection of the Y-coil shaft to shift to either "R" sense ~ or "L" sensing position corresponding to the right mask 94 or left mask 95, depending on the bias control line 105, 106, which is live. As an alternative arrangements can the positive and negative bias currents manually selected from an unreported source, instead of that the computer 27 supplies them. 0 If one sums the sensing of the image analyzer tube 9, achieved by the coil arrangement 93, it can be stated that single-sensing only Y-axis or vertical, two-way sensing available at a time. This happens continuously as a and sweep without resetting. In normal bar measurement there are no X-axis sweeps. Only positive or negative bias is applied to check the calibration of the measuring system when no bar shadow 8l is measured.

Since the bar shadow 81 is detected over the camera's field of view, if the output current from the image analyzer 90 drops sharply then soon the bar shadow 81 appears. It then rises again when the shadow has passed. This current change, along with electrical disturbances from the bar rolling mill, are converted to voltage, is amplified in an amplifier not shown in Fig. 4 and is the primary camera signal output from the camera head it 31 and appears on the lead 34. Ie the original 10 15 20 25 30 40 19 7802955-0 the camera signal at this point consists of a not too well-defined bar pulse, mixed with disturbances.

The image analyzer tube 90 in the camera head 31 operates in one self-balancing measuring loop 107 together with the camera pulse the processor 108, the AGC circuit 109 of the photomultiplier, which provides a variable control voltage in line 110 and a voltage controlled high voltage source lll for the image analyzer 90 photomultiplier section. Image analyzer 90 deviation section also supplied from a separate but stable high voltage source 112. 1 The camera pulse processor 108 is shown in Figs. 9 and 10 while Fig. 11 illustrates the time pulses of the processor. Furthermore, one is included buffer memory, dual differentiators, level detectors, zero-pass detectors and autocorrelators to remove noise from the original camera signal and from the differentiating organs. Signals processed in this way are combined with inverted window pulse ÉF in the processor logic to ensure that only measuring pulses with suitable amplitude and behavior at the right time, passing out for measurement purposes. This prevents also the passage of measurement pulses when the window is not open.

The camera pulse process 108 produces an intermediate stored camera signal 11A and precision square wave measurement pulses llP, llí, which are generated through an internal rocker. The width of the bar pulses varies proportionally. with the bar shadow 81 and therefore proportionally with the bar dimension between the bar edges 82 and 83.

The AGC circuit 109 of the photomultiplier shown in Fig. 12 and described below, receives the cached signal 11A and includes a comparator, a switching integrator and an amplifier to provide a switched variable control voltage in line 110. This control voltage is applied the high voltage source lll of the photomultiplier section to vary the gain in the image analyzer tube 90. The comparator establishes a reference gain level and an internal logic circuit generates an AGC quench pulse 8G by combining window pulse 8? with the inverted bar pulse llf. AGC- the extinguishing pulse effectively defines the time intervals when the camera signal must be read.

Now the self-balancing measuring loop 107 will work described. When there is no rod 10 in the measuring device, only light is emitted from the lamp 30 on the photocathode 91. This leads to the photomultiplier section in the image analyzer 07802955-o N l0 15 20 25 30 35 40 tube 90 generates a current flowing through line 34 and is proportional to the light intensity of the lamp 30. Photomultiples the reinforcement of the catheter section in the tube 90 is adjusted to a high level, initially through the AGC control provided in circuit 109. the effective level of voltage. When the brightness decreases or the image As the analyzer tube 90 ages, the AGC circuit 109 is compensated automatically for this by the photomultiplier section high voltage level from the voltage source lll is adjusted to change the gain in the 90 photomultiplier section of the tube thereby maintaining a constant camera signal amplitude.

When the rod 10 is hit by the light beam of the lamp 30, it works AGC circuit 109 also to maintain a constant output amplitude from the image analyzer tube 90. The self-balancing The measuring analyzer tube 90 then allows the image analyzer tube 90 to work with a high level of sensitivity, while a reasonably high signal / noise ratio is maintained, which is desirable for efficient processing of the primary camera pulse.

Precision bar pulses llP, clock pulses 8A, clock reset sensing pulses 8D and fixed sensing pulses 8H are applied the indication adjustment ll3. It has logical circuits, arranged to count clock pulses 8A for the duration of each of two bar pulses llP, which arise during a bidirectional sweep cycle, after which it is divided by two. The count is synchronous initiated by the clock reset pulse 8D, which occurs in bottom of each individual bidirectional sweep signal. Logical circuits are sensed by fixed sensing pulses 8G under conditions of reason to emit a binary rod size signal on line ll4 for indication purposes. To avoid indicator flicker, smooth the binary bar size signal over one in advance certain number of bidirectional sweeps, such as 4, 32, 512 sweeps, by unreported means.

Binary rod size signals are fed via line ll4 to a digital indicator ll5. This device includes integrated count-decoding indication modules, which are calibrated to indicate the uncorrected size of the rod 10 in decimal units, obtained somewhere in the camera's field of view.

The term uncorrected rod size was used for rod dimensions at this part of the bar measuring system, since no correction for optical and / or electronic non-linearities, rod temperature and rod composition have been undertaken. 10 15 20 25 30 40 7802955-0 21 The computer 27 performs corrections of the uncorrected ones the bar size signals and feeds a corrected binary bar size play signal via line 116 to the corrected digital bar size indicator 117. This digital indicator is structured in the same way as the digital indicator 115. Then both the bar size indicators 115, 117 have optical indication, intended to be synchronized and updated every 51l sweep under control of clock reset pulses 8D and fixed sensing pulses 8H. It should be noted, that the difference between readings on the bar size indicators 115, 117 indicate for a bar measuring operator and for rolling mill staff that (a) the bar measuring system correctors work as they and (b) the rolling mill rolls a product with the intended size.

Computer correction of bar pulses 1lP is based on exact determination of not only the rod size but also the rod center the position of the line in the field of view of the camera in relation to the optical axis of the camera head 31. To do this, store bar pulses llP, clock pulses SA, clock reset pulses 8D and fast sensing pulses Bra to bar size and position accumulator 118, as illustrated in the block diagram in Fig. 13, and the tuning of the pulses is shown in Fig. 8. Two separate counting and locking circuits, each controlled by one standard control gate, provides binary bar size output signals in line 119 and binary bar center position outputs in line 120. The binary bar size signals in line 119 have been developed similar to the uncorrected rod size the signals associated with the time indication circuit described above circuits 113. The binary bar position signals allow that perform corrections of the bar size signal to an accuracy at 1/256 of the camera's field of view. transfer of data between the computer 27 and other parts of the bar measurement system takes place through a logical measurement computer data transmission circuit 121. The logic circuit 121 receives an order signal islands line 122, which shows that the computer 27 is in such a position that this transfer can take place. The order signal 122 is logical combined with the "data ready" pulse in line 99 generated through the main clock 98 as described above. Their combined presence causes the logic circuit 121 to generate one transmitter prompt signal in line 123 and to synchronize the reconciliation of the measuring system with the computer 27. 7802955-0 22 10 15 20 25 30 Two-way swing generator The two-way sweep generator 97 is shown in the block diagram of Figure 7 and the time diagram of the figure 8. In order to perform the rod size measurements with a system precision of a quarter of it otherwise customary commercial tolerance in a 7.62 cm field of view, the two-way sweep of the Y-axis in the image analyzer tube 90 must be extremely linear and repeatable. Conventional analogues sweepers are generally difficult to perform for and hold this required linearity level. But if a waiver with with regard to system accuracy is acceptable for something measuring systems, analogous circuits may be considered. In order to however, meet the high standards of accuracy in the present measuring system, the two-way sweep on the Y-axis is generated by digital means with a crystal oscillator as time base, digital counters and a 14-bit digital / analog converter, which develops it current two-way sweep waveform 8E. Digital measures are taken to modify the sweep waveform 8E, described below.

The applied time base is a highly stable 12 MHz crystal clock oscillator 124, having square wave output} One buffer memory, 125, prevents irregular charging of the time base 124 during sweeps and feeds a clock pulse train 8A to the differential line driver 126. The output signal of the driver 126 is supplied as clock pulse 8A to the main clock 98 in the camera niken 35. The output of the buffer memory 125 also supplies clock pulses SA to a digital division unit 127, which has counters and logic devices, which generate waveforms 8B and 8C. The waveform SB is supplied with an up-down counter 128, a 14w or binary reverse counter. Waveform 8B has 5/12 of the base clock frequency or SMHZ. The waveform SC is a time pulse which is applied to the logic counter reversing circuit 129 and occurs twice at 12 clock cycles. Waveform 8B uses five pulse modes during one period of 12 clock cycles and waveform 8C uses two modes. This leaves five unused heart rate modes among 12 clock cycles in the two-way sweep period.

When the logic counter reversing circuit 129 senses that the up-down counter 128 has reached the full count value of all ones, it returns a countdown activation signal to the counter 128. The adjustment of the countdown activation signal occurs at the first tuning pulse 8C after being full count value reached. When the counter 128 senses the countdown 10 15 20 25 30 40 23 7802955-0 signal, it starts counting down to the next clock pulse SB.

When the logic counter reversing circuit 129 perceives all zeros in the counter 128, it generates an enumeration activation signal the next time a time pulse 8C arrives. The counter 128 starts count up with the next clock pulse 8B.

The up / down counter 128 has a 14-bit binary output, which fed over terminal 130 to the binary 14 ~ lt digital / analog converter 131. The D / A converter 131 adjusts the counter 128 and emits an extremely linear, analog, two-way sweep signal 8E. This signal is temporarily stored in the sweep circuit intermediate memory 132 to prevent overcharging of the D / A converter l3l. Then it is fed as sweep signal 8E to the Y coil driver 102 in the camera electronics 35.

When the up / down counter 128 reaches the last down »Xten, generates the reset pulse BD, which resets the logic circuit 129 and D / Converter 131. Differential line driver 133 supplies the reset signal to the main clock 98 i camera electronics 35.

As mentioned above, there are five unused heart rate modes in one period of 12 clock cycles. These can be used to take care of an exact, non-linear modification of the extremely linear the scan signal 8E by arranging a digital multiplier 134 in series between the digital divider 127 and the up / down counter 128, shown in Fig. 7 in broken line. The digital multiplier 134 receives waveform 8B instead of the up / down counter 128 and produces by means of a suitable multiplier a modified wave form BB '. The up / down counter 128 receives the modified one waveform 8B 'and changes together with the setting pulse SC on the order signal the total enumeration or total the countdown, depending on the specific value of the multiplier.

This multiplication still gives a triangle sweep with soft curved sides as indicated in the form of the modified line sweep signal 8E '.

The digital multiplier multiplier 134 fed via line 135 and can start with the computer 27. Alternatively, the digital multiplier can be set by not reproduced manual bodies. Regardless of its power source can the multiplier is used to perform sweep corrections for to eliminate optical and / or electronic errors, for which no other corrective action has been taken here. 'zsozass-o 24 10 l5 20 25 30 35 40 Camera pulse processor Camera pulse process 108 is shown in Figures 9 and 10 block diagram and in the time diagram of Figure 11. The camera pulse processor 108 converts the primary camera pulse in line 34 to an_exact bar output pulse in line 1lP with a hredd with well-defined flanks, which accurately represent the dimensional the relationship between the rod edges 82 and 83. Due to the differentiator, the autocorrelator and others below described execution details, the camera pulse processor is suitable 108 very well for processing the primary camera pulses at camera sensing speeds of up to 300 Hz, and eliminates still the effects of the camera signal and the differentiation organ noise and disturbances.

Fig. 9 shows the camera pulse processor 108 in the form of a block diagram, where alphabetical names refer to the waveforms of Fig. 11. The primary camera signal in the line 34 is temporarily stored and amplified by the buffer memory 136 to output signal 11A. Signal 11A is differentiated through a first differentiator 137 with an output signal 11B.

The first differentiated signal 11B is applied to the upper and lower threshold detectors 138, 139, which output signals 11C and 11D. Threshold detectors 138 and 139 provide output signals. nals when their positive input (+) has a lower voltage than theirs minus input (~). The first differentiated signal 11B is differentiated again in a second differentiator 140, which emits output signal l1E. The second differentiated signal 11E is applied zero crossing detectors 141 and 142 for start and stop.

These detectors are intended for triggering at positive and negative zero-pass transitions greater than 1 millivolt, wherein bar pulse start and stop zero signals l1F and llG u delivered. Bar pulse start and stop zero signals 11F and llG is applied together with the upper and lower threshold signals l1C and l1D fixed delay autocorrelator 143. Bar pulse the start and stop zero signals 11F and 11G are processed internally in the respective autocorrelator circuits as described below. The the lower and upper threshold signals 11C and 11D define narrow windows below which the bar pulse start and stop signals llm and ll "0" are triggered, whereby an exact tuning for the leading and trailing edges of the bar output pulse 11P is established. 10 l5 20 25 30 35 40 25 '7802955-0 As mentioned above, the electronic camera contains 31 signal in line 34 possibly also electrical disturbances.

This can be high frequency interference with low amplitude, which are often magnetically connected to the electronic camera signal from high voltage, ballast rectifier, rolling mill motor actuators located near the electronic camera 31 etc. Without autocorrelator 143 with fixed delay these disturbances can jeopardize the triggering of the bar output pulse llP. For example, if a transition of the camera signal 11A produces a first differentiating voltage 11B below a -3 volt threshold for the detector 138, a lower one is enabled threshold signal 11C, which allows the zero crossing detector 141 to be supplied starts trigger signal for a bar output pulse. Since the gain of the differentiating means 137 and 140 increases with input frequency, a high-frequency interference peak with low amplitude to produce a first output signal 11B from the differential which is lower than the detector 138 -3 volts threshold. This is exactly what will happen in the rolling mill environment without raising the bar pulse generating circuit.

For this reason, the autocorrelator 143 includes fixed delay, which is included in the processor 108 for the primary the camera pulse actually separate autocorrelator bar pulse start- and stop circuits 144 and 145, shown in Fig. 10. the start and stop circuits 144 and 145 are arranged to distinguish between other-differentiated signals 11E, which come through high-frequency disturbances and those that have arisen through valid bar pulse signals. While the camera signal 11A falls, the second differentiated signal 1EE rises to a positive voltage for about 10 microseconds before turns to a negative voltage. Of illustration technology For this reason, this detail has not been shown on a scale with the signal l1E waveform in Fig. 11. The zero-pass detection of the other differentiated the signal 11E by the detectors 141 and 142 is the trigger point for the start and stop bar pulses in the signal l1 fl and 11 "0", the bar output pulse l1P leading and trailing edges are established.

Autocorrelator bar start and stop circuits 144 and 145 takes advantage of the relevant ascent and fall period 10 microseconds in the second-differentiated signal 11E.

This is done by generating the autocorrelator activating start and stop signals 11L and 11l, as described below. '7802955-0 26 10 15 20 25 30 35 40 The autocorrelator start signal III is generated when the second differential the retired signal llE is continuously positive below at least half of this period of 10 microseconds before it turns into negative. Similarly generated the autocorrelator stop signal 11l when the other differentiated the signal 11E is continuously negative during at least half of the period of 10 microseconds, before the reversal to positive.

The autocorrelator start and stop signals llL and llN are logically "AND-associated" in circuits 144 and 145 with associated lower threshold signal 11C and 11D and bar pulse start- and the stop zero pass signals llF and llG to generate the bar pulse start and stop signals llH and ll "0".

These signals cause the exact generation of bar output pulses llP. It is now understood that high frequency disturbances such as causes positive resp. negative deviations from the other differentiated signal llE with a duration shorter than 5 microseconds, does not cause any signals llL and llN, enabling the start and stop of the autocorrelator, whereby a tripping of the rod output pulse lLP is prevented.

Still referring to Figure 10, the function should now of the autocorrelator bar pulse start circuit 144 is described.

The function of the bar pulse stop circuit 145 is identical to that of the circuit 144 function with the exception that it responds to a second differentiated signal llE, which is continuously negative for 10 microseconds, before turning positive.

Both circuits 144 and 145 have conventional logic devices.

The lower threshold signal 11C is inverted in the amplifier 146 and is supplied to one of the three inputs of a NAND gate 147.

Gate 147 provides the bar pulse start signal llm below special logical conditions.

The bar pulse start zero pass signal 11F is applied to one Schmitt trigger 148 and inverted in an amplifier 149, whereby the trigger signal 11H is emitted, which is applied to the NAND gate 147 and a monostable delay means 150. A negative continuous transition of the signal lLH triggers the monostable the delay means 150, which emits a 5 microsecond long logic "lv pulse lII at output Q and a 5 microsecond long logical "O" -ouls llJ at the output Ö. The pulse llI is applied to one of the two inputs of the AND gate 151. Schmitt triggers 148 10 15 20 25 30 b) LP! 2., 7802955-0 output is also connected to the second output of AND gate 151 and also to the rocker input 152 reset input. Pulsen llJ is added to the flip-flop 153 clock input. The upper threshold signal 11D is applied to the data input of the flip-flop 152 to release the autocorrection the latch starter circuit 144 during the falling of the camera signal 11A edge and to block this circuit below the signal 11A rising flank.

If the signal ll fl goes negative, the input of the inverter 149 goes positively. This positive process cancels the return of the rocker 152. position condition and puts a logical "l" on the AND gate 151 one entrance. AND gate 151 now passes through pulse lll to the bell input of the rocker 152 and thus forces one logic "l" pulse llK at the output Q. After 5 microseconds delay, the monostable delay means 150 time out and thereby causes the expiration § to change the permit and switch to a logic "l" pulse llJ. This also sets the input of the flip-flop 153, the data input of which is applied to the signal 11K from the flip-flop 152 Q output. If the signal llK is a logic "1", the flip-flop 153 is set output Q and outputs a start enable signal lLL. The signal 11L, generated by the signal 11H, is logically interconnected with the signals llH and llö, the inverted lower threshold signal nal, in the NAND gate 147 to output the bar pulse start signal llM. Thus, it will be appreciated that a bar pulse signal is delayed, after which it is combined with itself to carry out an autocorrelation function with fixed delay.

If below it through the monostable delay means 150 controlled period of 5 microseconds Schmitt ~ trigger 148 output is at a low level, which indicates that the second the signal llE is too narrow to be valid bar signal, the reset of the flip-flop 152 is at a low level and forces the signal llK to a logic "0". When after 5 microseconds of the monostable delay means 150 delay time expires, the signal 11J sets the flip-flop 153 with its data input in a low state. This forces the rocker 153 Q output to a logic "0" and prevents further processing of the bar signal.

The monostable delay means 150 can be released again, so that it adjusts consecutive trigger pulses lLH. About multiple trigger pulses have a short duration of less than 5 microseconds, the monostable delay means 150, the Q output signal, is released 7802955-0: 28 10 15 20 25 30 35 40 lll will stay at a high level for all pulses and stops finally 5 microseconds after the last trigger pulse.

AND gate 151 allows the rocker 152 to reset itself at each pulse. Because the monostable delay means 150 output stays continuously at a high level below these multiple trigger pulses guarantee the combination of signal llI.with the Schmitt trigger pulse in AND gate 151, to the clock line to flip-flop 152 undergoes a logical transition from "0" to beer "for each trigger pulse.

As mentioned above, the pulse stop circuit 145 is identical to circuit 144 except that stop circuit 145 is triggered through a continuous negative second-differentiated signal 11E before the turnaround to positive. For this reason, it is obvious to those skilled in the art that the inverter 154, the NAND gate 155, the Schmitt trigger 156, the inverters 157, the monostable the delay means 158, the AND gate 159, the flip-flops 160 and 161 have the same construction and function as their equivalent in circuit 144. It was therefore deemed unnecessary to explain .these devices for showing how the NAND gate 155 emits bar pulse stop signal ll "0". 7 When both the electrical disturbances in the primary the camera bar pulse signal and those through the differentiators 137 and 140 produced disturbances eliminated, define the bar pulse start and stop signals l1M and ll "0", which generated in respective circuits 144 and 145, now exactly the setting of the front and rear flanks of the rod zeros in relation to the rod edges 82 and 83. Therefore, the signals 11M and l1 "0" to each setting and reset input on flip-flop 162. An inverted window pulse ÉF, shown in fig. 8 and emitted by the window generator 100, is supplied to the flip-flop 162 7 clock entrance. The data input of the flip-flop 162 is connected to 0 volts.

This enables the flip-flop 162 to emit a rod output pulse only when a window pulse 8F is present. The width of the window pulse and duration varies for bar measurement operations and for calibration controls explained above.

During the measuring operation, the Q output of the flip-flop 162 provides one exact bar output pulse llP, the leading and trailing edges of which are free from interference and precisely defines the cross-section of the bar 10 mensioner. During the calibration check, the computer 27 selects RTMASK or LFTMSK program, defines the bar pulse llP exact right and left mask 94 and 95 dimensions. 10 15 20 30 35 40 29 7802955-0 Photomultiplier AGC circuit AGC circuit 109 (automatic gain control) for image analyzer the photomultiplier section of the tube 90 is shown in Fig. 12. The AGC circuit 109, which is an important part of the self-balancing measuring loop 107, comprises a comparator 163, a coupled one integrator 164 and an amplifier 165. The amplifier 165 feeds the high voltage source 111 of the photomultiplier section 111 a coupled variable control voltage via a line 110. The coupled variable control voltage acts as an automatic gain control for the tube 90. This is done by vary the high voltage source of the photomultiplier section lll so that it keeps the anode current of the tube 90 at a constant reference value.

The intermediate stored camera signal 11A is applied to the comparator 163 one input through the summing resistor 166 and a summing resistor connection 167. The summing connection 167 is through the diode 168 limited to positive input signals. A comparator reference voltage from the voltage source 169 is set by one potentiometer socket 170 for the purpose of displacing the bar pulse and establish a nominal value for the coupled control signal which finally, the high voltage source 111 sets to a nominal reinforcement-producing value.

The intermediate stored and offset bar pulse at the summing point 167 also goes to an electronic switch 171 in it connected integrator 164. The window pulse 8F and the inverted rod pulses 11? are logically merged in AND gate 172 for to produce the AGC reset pulse 8G shown in FIG. 8. When a window pulse is present and a bar pulse is missing, the AGC reset pulse 8G causes the electronic switch 171 to connect the current to the integrator amplifier 173 and to charge the integration capacitor 174. Here both a window pulse SF and rod pulse llP are present, it opens electronic switch 171 and allows the integrator output at connection point 175 maintains a nominal input value to the amplifier 165.

Amplifier 165 consists of a summing resistor 176 which at one end is connected to the output of the integrator connection point 175, while its other end is connected to the input of the operational amplifier 177. The feedback resistance 178 controls the gain of the amplifier 165. A zener diode 179 limits the gain of the amplifier 175, so that none too high coupled control voltage arises in line 110, which ..._...-.__._......_._.__._ i vsozess-o 3 ,, 10 15 20 25 30 40 would overload the high voltage source lll. In summary can be said that when the AGC reset pulse 8G is missing, the the intermediate storage camera signal 11A is passed through the AGC circuit 109 and affects the high voltage supply of the photomultiplier section lll. In the presence of an AGC reset pulse, the camera the signal 11A and the output of the AGC circuit of the photomultiplier 109 is kept at a constant reference value, which is determined by charging the capacitor 174 in the integrator 164.

Bar size and position accumulator The size and position accumulator 118 is shown in Fig. 13, reference also being made to the time diagrams in Figs. 8 and 11.

In the present bar measurement system, the computer 27 is supplied, uncorrected, digital bar size and bar position data in a similar way but separate from and independent of uncorrected digital bar size data, which is displayed on the indicator 115.

The accumulator 118 is provided with a control gate 180, which sets bar pulse llP, clock pulse 8A, clock reset pulse 8D and rapid sensing pulse SH in the rod size accumulator circuit 181 and the rod position accumulator circuit 182. The circuit 182 determines the bar centerline anywhere in the camera's field of view.

Both circuits 181 and 182 are synchronized by clock feedback. position pulse 8D and both are sensed by a fast sensing pulse 8H in each complete sweep cycle.

The control gate 180 detects the front and rear flanks of each bar pulse 11P and halves the number of clock pulses 8A, which occurs during the two bar pulses present during the upper and lower halves of the sweep cycle. The control gate 180 directs these clock pulses to the input of the binary 14- bit the counter 183 in the bar size circuit 181, where counting of two bar pulses divided by two are recorded. At the end of one first sweep cycle, the said count value of the counter 183 is transmitted to the data input of a binary l4 ~ bit latch 184, wherein 'it is assumed that a previous application of the solid the sensing pulse 8H has been applied to the clock input of the latch. IN At the beginning of the second cycle, counter 183 is reset by the clock reset pulse 8D and is ready to receive a new one pulse sequence.

Digital 14-bit data, which represents uncorrected bar size between bar edges 82 and 83 from first the sweep cycle, is stored in latch 184 for a second sweep cycle.

During the second scan cycle, this data is transmitted via the cable 10 15 20 25 30 35 40 31 vsozøss-o 119 to the computer 27 for correction under the computer described below program CMPNST. At the end of the second sweep cycle, it is detected the data of the counter 183 and is transmitted to the latch 184 through pulse 8H, thus repeating the cycle. Bar size pulses count is always a sweep cycle before the blocked bar size data in the rod size accumulator circuit 181.

The control gate 180 also detects the first 11P rod pulse the flank at 185 resp. 186 under the upper resp. lower half of a sweep cycle, shown as waveform 8G in Fig. 8. The control gate 108 determines the sweep time between the leading edges of the pulse llP 185 and 186 and this time divides by two, whereby what is then referred to as the sweep time for the bar center the position of the line. The guide gate 180 further includes a rod position. time base, which is developed by dividing the 12 MHz clock pulse train 8A by a factor of 160 in the divisor 187, whereby 8A / 160- clock pulses are generated. 8A / onion clock pulses are directed to it binary 8-bit counter 188 clock input in the bar position accumulator 182 as long as the sweep time for the position of the bar centerline lasts. The the count value registered in the counter 188 represents the position of the center line 10 of the rod, which can be located anywhere in the camera's field. This position of the bar centerline was determined completely independent of the rod size measurement, which took place in the size accumulator 181 or elsewhere. 1 I At the end of the first sweep cycle, the count value is transferred for the centerline position in the counter 188 to the data input thereof binary 8-bit latch 189, assuming a previous one application of a fixed sensing pulse 8H has been applied the locking input of the latch. At the beginning of the second cycle is reset the counter 188 through the clock pulse 8D and is ready to receive one new pulse sequence, regarding the position of the bar centerline for counting. 8-bit data, which represents the position of the center line of the bar in the camera's field of view, stored in latch 189 for a second scooter. During the second sweep cycle, this data is transmitted via the cable 120 to the computer 27 for correction of optical errors regarding rod size data in the accumulator 181 below described computer programs CMPNST. At the end of the second sweep cycle the data of the counter 188 is sensed by the latch 189 through the pulse 8H, thus repeating this cycle. Pulse count regarding the centerline position of the bar is always a sweep before the barred ones data in the bar position accumulator 182.

The bar position accumulator 182 divides one half of 7802955-0 32 a sweep cycle in 256 parts of 0.046 mm. Camera head 31 optical center line is at the l28th part. Total represents the parts 10.404 cm of the Y-axis sweep, applied to the Y-axis deflection coil with a normal field of view of approx. 7.62 cm. The useless field of view is 2.784 cm, the distance that Y- the deflection coil of the shaft sweeps away at the top of the photocathode 91 and bottom.

Computer A block diagram of the electro-optical shown in Fig. 1 the computer 27 of the bar measuring system is shown in Fig. 14. The computer is one digital systems, programmed to implement the various written functions. A commercially available Fortran- programmable or em fixed minicomputer can be used 15 or, if desired, the computer 27 may be included in the whole of the rolling mill control computer installation. The computer 27 is in the present example Ven Westinghouse Electric Co., USA, model W-2500 with one operating systems to adapt different levels of data such as listed below. The computer 27 is provided with conventional main components including an input buffer memory 190, output buffer memory 191, disk memory l92, disk switch 193, core memory 194, all through various channels connected to the data processing unit 195. The work of the computer 27 is controlled in turn according to the Line and on-line computer programs 196. These include: in FIG. 15 and 16 show computer cards 197, service program 198, bar computer program 199, compensation program 200, calibration program 201, recalibration program 202 and histogram program 207, all described below. All communication with the bar measuring system computer 27 from external sources occur via the input buffer memory l90, which includes means for converting analogue and digital input signals to digital form. These include signals, supplied via wires or cables to the computer as follows: Camera Electronics 35 via cable 36; the hot metal detector 57 via line 58; rod temperature 50 via cable 53, 54; setpoint size play 42 via line 43; rod assembly 44 via conduit, 45, other data 46 via cable 47, control system 67 via cable 68; CRT terminal 60 via cable 61 and write terminal 63 via 40 cable 64. 7 10 15 20 33 7802955-0 All communication with the bar measuring system computer 27 to external sources occur through the output buffer memory l9l, which also includes means for converting output signals to digital and analog form. These include signals such as supplied via wires or cables from the computer according to the following: control system 67 on cable 66 and camera electronics 35 on cable 37.

Individual wires in signal cables have been used in the drawings and these have been drawn according to their source and function, as described above.

The CR0 terminal G0 includes an operator control panel interaction with the computer 27.

The write terminal 63 includes a control panel for the operator's control of the computer 27. The terminal's 63 computer printers 65 includes a rod diameter deviation printer as well as a table data according to the list below.

In general, it is allowed for both terminals 60 and 63 to print the same data. All collaboration between either the control panel is listed by program abbreviations, for example as follows: OFFLTNE MEASUREMENT SYSTEM ABBREVIATED NAMES AS FOLLOWS: IJISTOGRAM FOR EACH HEAD MP BUILDS COMPENSATION CARD FOR EYE FIELDS cL CARRIES OUT CALIBRATION CHECK ON THE RIGHT AND LEFT MASKS TY PRESSES SHORT, FURNISH AND SHIFT FACTORS AND MASK VALUES OF ALLOWS FEED OFF FIRE AND SHIFT CORRECTION- FACTORS ZE RESETS ALL CARDS AND CORRECTION FACTORS lll WARNING il! LF LEFT MASK DEVIATION TEST RT RIGHT MASK DEVIATION TEST (ALSO ALLOWS WINDOW ENTRY) TR DISC TRANSMISSION OF MEASUREMENT VALUES COMMON FOR CONTROL FUNCTIONS IN XT OUTLET TO MONITOR AND TRY TO WRITE COMMON AREA CONTAINING CARDS, CORRECTION FACTORS FOR FIRE AND SHIFT MASK VALUES AND WINDOW VALUES TO DISK MEMORY.

DISK MEMORY DATA IS UPDATED ONLY IF THE DISC SWITCH 12 IS UP. THIS DATA IS READ FROM THE DISK MEMORY WHEN THIS PROGRAM (20) REQUESTED BY MONITOR. veozess-aoi Mi Disc switches l93 include switches called "switches" 10 "and" switch 12 "in the program below. These switches must turn on "write enable" to update software or data on the disk memory.

Computer program The following table lists individual programs and groups associated with the computer program 196 used.

COMPUTER PROGRAM IDENTIFICATION USE V, OFF-LINE ON-LINE SHORT (197) DISC MEMORY CARD X CORE MEMORY CARD 'X X SERVICE PROGRAM (198) , IDL HANDLER M: IDL CD: IDL EB: IDL GAGTSK SUBCLL GAGTRN BAR MEASUREMENT DATA PROGRAM (199) GAGEIN COMPENSATION PROGRAM (200) GAGMAP CORDAT ZERO MAPRNT GAGTPC CMPNST CALIBRATION PROGRAM (201) CALIBR RE-CALIBRATION PROGRAM (202) - RTMASK X GAGRCL X LFTMSK X HISTOGRAM PROGRAM (204) GAGHST X HISTOGRAM SECTION WITH CONTROL SYSTEM X N #! N! N> C? $ N N N> 4 N> N% åN {N> 4> MEMORY CARD DISC MEMORY CARD, see Fig. 15. Program address in disk memory 192.

CORE MEMORY CARD, see fig. 16: Program address in hexadecimal core memory l94 10 15 20 25 30 35 ss 7802955-0 smzvïcnmosmi (198) IDL control program, M: The IDL program processes all data transmission between IDL hardware (channels 30 and 32) and measurement data input sub-program GAGEIN. It communicates with the IDL hardware via the IDL channel driver CD: IDL. A dual buffer memory scheme was used to reduce the total transmission time through to initiate an extra IDL transmission on both channels into one other data buffer or cache just before data exits the control program. In this way, data can be transferred to this other intermediate memory by means of the IDL hardware by using interruptions between service request interrupts SRI, which takes place outside the sequence area while the measurement software is engaged in processing data from the first intermediate memory. When this processing is complete, the control program is switched on again. About the data transfer in the other the cache is not completed, the work is interrupted until IDL's external MACRO routine indicates two interruptions by overfilling cache devices. The program is resumed through IDL's external MACRO routine EB: IDL when two overflows of intermediate memories have nats. If the data transfer in the second intermediate memory is completed or after the program has been resumed by EB: IDL, read the middle memories. A data transfer using intermediate the memory l is started and an output is made from the control program. Measuring the software now processes data in the cache 2 and repeats the above standing sequence.

A monitoring timer with a half-second timeout is set before each IDL transfer begins. If two cache overflows not returned within this time period, the clock routine will to no longer inhibit the program and set it variable ISTAT on 1 to indicate a time error on the IDL transmission.

The variable IBUF is set through this program to set indicate which cache - 1 or 2 - contains data from the last IDL transfer. The variable IRSTRT must be from the beginning reset by the calling program so that this program know when entry has taken place for the first time. When IRSTRT = 0, initializes the dual cache mechanism. This program then sets IRSTRT = 1 to indicate the dual cache operation in progress. If input to the trader has been made with IRSTRT = -1, an IDL breaker order is sent to both IDL channels to stop all ongoing transfers. This order is usually prompted by 7802955-0 36 10 15 20 25 30 35 40 the calling program before a call output signal is output, so that all IDL transmissions are stopped.

This program calls the IDL channel driver CD: IDL and uses IDL's external MACRO program EB: IDL. Therefore, these programs must connected to the ID: M: IDL control program.

The IDL Merchant, CD: The IDL program was used to transfer data from the trader control blocks (HCB), which are defined in the IDL trader M: IDL to the IDL hardware (channels 30 and 32).

The control is transferred to this program by input The HCB address in the B-register and by skipping to CD: IDL (CD: IDL must be declared externally). Control program control block HCB is a Qåordstabell, which has the following structure: Word no Explanation Example for channel 30 0 Forced cache ~ input IDL code DAT X'B30 ' 1 Break IDL code DAT X'F30 ' 2 Return address - l ADL RTRl-l 3, Available DAT O _ 4 Cache IDL code DAT Z'530 ' 5 Mode in kernel memory containing address of data DAT X'llFB ' 6 Number of words for transmission DAT 20 Address of the data cache SIZE l DAT 354 SRI address Vector (10 + SRI x 2) This program performs three functions, which use the HCB table.

First, a break code (HCB - word l) was sent on I / U (input / output) subsystem. The bottom seven pieces of this words define the number of the channel to be broken. For it others were sent a forced intermediate memory input (HCB - word 0) on the I / O subsystem. This command starts the IDL hardware on it selected channel. Third, the intermediate storage input the transmission code of the I / O subsystem to initiate data transmission the lead. The data is transferred to the core memories from the selected IDL channel during service interruption (SRI). Those used by the SRIs the addresses and counters are set by this program, which uses data obtained in the HCBs.

'IDL control program, EB: The IDL program is called when the POS / l- the intermediate memory is overfilled and a service requirement crime. The call is made through a program outside the regular 10 15 20 25 ' 30 35 40 of 7802955-0 sequence instruction area as a result of interruption due to of overfilled intermediate memory, which occurs when an intermediate stored data transmission on either of the IDL channels 30 and 32 is completed.

Each entry in this program causes the arithmetic word (ECB7) for memory overload in the external MACRO control block to step forward.

When counting to 2, it becomes through the IDL control program M: IDL suspended program reactivated. If the count does not come 2, return to the output program M: BOX for POS / l intermediate memory overload and the state of the suspended program is unchanged. Thus, if the IDL control program M: IDL requisitions data from all four IDL channels, it resets the cache the overfill bill and postpones the program. It is activated again when the IDL external MACRO program counts two interrupts by the intermediate memory being overfilled due to completion data transfer.

GAGTSK, a program belonging to a disk memory (program 20), is an off-line program, intended for reading into disk memory stored, off-line measurement subprograms' overlays in core memory and to incorporate them into governance. GAGTSK calls one especially subroutines of the core memory when shortening parameters led to the same by the operator-interactive subscribers described below program call overlay SUBCLL. All programs and their benefits abbreviated designations, are described in the list of sub-programs SUBCLL.

GAGTSK also transfers a common area belonging to the disk memory to the core memory, and, if the disk memory sector switch 12 is in the position to write, the updated common area is written back to the disc when it exits the program.

An off-line recorded signal IGAGOF is set at the input of this program and is reset at the end.

SUBCLL, a disk memory subform in the form of overlay, run in off-line operation by means, with which one operator can cooperate with the off-line measurement system to run the any of the available off-line bar diameter measurements programs. It is transferred from disk memory to kernel memory and run through the off-line measurement program GAGTSK (program 20) by means of a system monitor disk read-and-transfer control program. By operating abbreviated designations introduced by the disk memory sectors, which were retransmitted as subprogram parameters to GAGTSK, which in turn transfers and runs the desired the program overlay. The subprogram function is described in this program list and is available to the operator when required assistance. 7802955-o N l0 15 20 25 30 40 The GAGTRN program is run in the off-line measurement system. The transmits a measurement data block of 572 words from a predetermined one. disk memory area to another area, intended for the control system 67. It performs a disk-core-disk memory transfer during use of the common memory meter area for intermediate storage. Disc Cover The file l0 must be ready for printing.

BAR MEASUREMENT DATA PROGRAM (199) in GAGEIN, a utility subprogram, is always added to each subprogram, which requires bar measurement data. They call it also added IDL-managed programs (M: IDL, CD: IDL, EB: IDL), for the actual acquisition of bar position ~ and diameter data as well as added compensation subprogram (CMPSNT), if compensation is required. It evens out the approved ones returned readings, both bar position and diameter, calculates deviations, and stores the result in standard tables.

Validity tests are performed and error signals are emitted as required extent.

COMPENSATION PROGRAM (200) _ GAGMAP, a subprogram stored on the disk memory as one overlay f, runs in off-line operation, leading to a compensation table used for on-line bar diameter measurements tasks and sub-programs and off-line measurement programs, such as requires compensated size data. The tables are on one common area and was used to compensate for the picture tube non-linearity across its image field. The tables are set up and printed by printer 63. This program must be run before any rod diameter data can be considered valid. The is called in through the SUBCLL subroutine and requires the operator intervenes.

The Compensation Map consists of 256 inputs, which belong to the 256 possible bar positions. Element 1 represents bottom of the total 10.404 cm wide field and element 256 represents the top of this field. Each element contains correction data, to be subtracted from the measured bar size, based on the position of the bar top and failed. The current correction is implemented through subpro- gram CMPNST. The use of the edges' 82 and 83 positions instead of the center position allows to use the card for all sizes of the rod 10. 10 15 20 25 35 40 39 7802955 '-0 During the card build-up procedure, a calibration rod 10 of 12.7 mm back and forth É 47.1 mm in a plane in right angle to the optical axis. While the rod 10 is being moved back and forth, the GAGMAP off-line calibration system is performed.

This program processes 10,000 measurements and calculates it average deviation for each part of the bar position.

This intermediate data is stored in a table of 256 elements, called ISUM.

The final compensation card, which is based on the 82 and 83 positions of the edges, are developed under the the table through the following work steps: 1. The compensation card is reset 2. A computer simulation is performed, whereby an imaginary 12.7 mm thick rod 10 is positioned 0.406 mm above the center of the field of view (slit 129). The positions of the upper edge of the rod 83 and lower edge 82, respectively, are calculated metrically as follows. eye field center + 0.406 + rod size / 2) upper edge 83 0.406 (eq. 1) Bottom edge 82 b1ickfältsmiät4É60,406 - rod size / 2 (eq_ 2) f Example: upper edge 83 = (52.0l8 + 0.406 + 12.7 / 2): 0.406 = 144 (eq. 3) Failed 82 = (52.0l8 + 0.406 - 12.7 / 2): 0.406 = 113 (eq. 4) 3. The value stored on the card for the position of the upper edge 83 (144) is the sum of the deviation stored in the ISUM table in relation to the center position (129) of the bar 10 and the bottom edge (82) on the card stored position value (113).

IMAP (upper position) = ISUM (bar center position) + IMAP underk. mode) (eq. 5) IMAP (144) = ISUM (129) + IMAP (113) (eq. 6) 4. steps 2 and 3 are repeated by dividing the middle of the rod 10 position to 0.812 mm above the center of the field of view, then 1.218 mm, 1,624 mm etc. This is repeated until the upper edge 83 of the rod 10 passes over + 47.1 mm above the center of the field of view. 'zsozsss-o 4., 10 ' 15 20 25 30 35 40 IMAP (145) =: sun (130) + IMAP (114) : mm UAM: Iam nan + 1mw uim IMAP (147) =: sun (132) +: mir (116) IMAP (220) = ISUM (205) + IMAP (189) 1mAPa (z21) =: sun (zos) + IMAP (190) The upper half of the card is now complete. 5. The lower half of the card is filled in in the same way. Based on the same calibration bar 10 of 12,7 mm in the field of view (128) middle, the position of the upper and lower edge 83, 82 is calculated metrically as follows. överkant es = (b1ickfä1tsm1tt + §ÉÉ맧É9ÉíÉë-) = o, 4oe (eq. 1) Bottom edge BZ = (center of view - šššågâšgëšgë): 0.406 (eq. 8) upper edge 83 = (52.0l8 + 12.7 / 2): 0.406 = 143 (eq. 9) Bottom 82 = (52.0l8 - 12.7 / 2): 0.406 = 112 (eq. 10) 6. The short value (ll2) of the lower edge 82 of the bar is the sum of the deviation, stored in ISUM, corresponding to the center position of the bar (128) and the value of the rod 10 stored on the card (143) for edges 83.

IHAP (sub.1 mode) = ISUM (bar center mode) + IMAP IMAP (ll2) = ISUM (128) + IMAP (143) (Eq. 11) (Eq. 12) 7. Steps 5 and 6 are repeated by successively dividing the position of the bar 10 in steps of 0.406 mm from the center of the field of view until the lower edge 82 of the rod 10 goes below -47.1 mm below the gaze the middle of the field.

ISUM (127) + IMAP (142) ISUM (126) + IMAP (141) ISUM (125) + IMAP (140) : amp (111) IMAP (110) 11-111? (109) 10 15 20 25 30 40 41 78Û2955'Û ISUM (§2) + IMAP (67) ISUM (53) + IMAP (68) IMAP (36) IMAP (35) The card under half is now complete. Card positions above 221 and below 35 were not used.

These positions relate to the unused part of the field of view in the shadow of the photocathode tube, shown in Fig. 5. 9. The card elements lll-143 are zero. This refers to an area ï 6.35 mm from the center of the bike. 10. The card belonging to the camera head 31 is stored in a common data memory FCOMPI.

CORDAT is a program, which is run during the off-line measurement procedure.

Its purpose is to enable the operator to enter the steep the offset and offset correction factors for the camera vudet 3l. The two variables are: IMULTl - Slightness Correction Factor for Camera Head 31 IOFSTl - Offset Correction Factor for Camera Head 3l.

The steepness correction is added to all bars through the field of view. the CMPNST compensation subroutine according to the following formula: size = (12.7 mm - size) * IMULT1 The displacement correction is added to all rod sizes by means of the field of view compensation sub - program CMPNST according to the following formula: Size = Size - IOFSTl ZERO is a program that runs in the off-line measurement system.

Its purpose is to reset the compensation card, all the steepness and displacement correction factors and the right mask calibration constant.

MAPRNT is another program, which runs on the off-line measurement system.

It does not require the operator to intervene. Its purpose is to print the field of view compensation card, steepness and displacement correction factors and the values for left and right masking.

GAGTPC calculates the hot brush size, based on an internal stored compensation equation. Three variables are required for this equation. First, the percentage of carbon, which obtained from IGRADE in the BDCCOM common area. For it second, the bar temperature obtained from ITMP22 on it common area SYSCOM. Third, the cold brush size, obtained from ICDAIM in the BDCCOM common area. The calculated heat size is stored in a preselected area in the BDCCOM memory. vaozessa-o p 42 10 15 20 25 30 35 40 CMPNST, a utility subprogram, is added to each subprograms, which require compensation of measurement diameter data.

This subprogram linearizes bar measurement data for the bar position in the measuring field, corrects measurement data for steepness and offset data through the CORDAT subprogram and executes a automatic calibration based on the data of the right mask, generated by the GAGRCC subprogram.

The size data of the rod 10 from the camera head 31 is linear. through the CMPNST subprogram, which uses the the pension card FCOMP1, provided by the off-line pro- gram GAGMAP. The compensation is performed through the following steps I l. Bar size and position data from accumulator 118 was used to determine the top edge 83 and the bottom edge 82 metric position on the compensation card as follows: I (bar center position + bar size / 2): 0.406 / 2): 0.406 top edge position = Bottom "= (" - " If the center of a 25.4 mm thick rod is located l9.05 mm above the center of the field of view, the position of the closing center is 52,018 mm + 19.05 mm = 71.07 mm. The position of the top and bottom edges is determined as described above. This means upper edge az position = (71.07 + ÉÉ-å- “ín 0.406 = 203 (equation) unaerkantens sz "= (71.07 - 2554): 0.406 = 140 (eq.14) _ 2. Compensation for the upper and lower edge position values are obtained from the card and associated values ICOR1 and ICOR2.

ICOR1 - IMAP (upper edge 83 position) (eq.l5) ICOR2 = IMAP (position of the bottom edge 82) (eq.16) 3. If both the lower and upper edge are above the gaze field center, applies to: Corr.bar size = uncorr. size -ICOR1 + ICOR2 (eq. 17) 4. If both the upper and lower edge 33, 82 are below eye center, applies to: Corr. size = uncorrected. size + ICOR1 - ICOR2 (eq. 18) 5. If the top edge 83 is above and the bottom edge 82 is below the center of the field of view is: Corr. rod size = uncorrected. size - ICOR1 - ICOR2 (eq. 19) CALIBRATION PROGRAM (201) CALIBR is a program, which is run in the off-line measurement system.

It does not require the operator's participation. Its purpose is to get a work log for the measuring printer 63. It performs the following features: 10 l5 20 25 30 35 40 43 '7802955-0 l. Steer to each left and right mask 95, 94 and: a. measure and write the size of each mask; b. Calculates and writes deviations from stored worm value; c. Measures and writes (+) steepness value d. Measures and writes (~) steepness value; e. Writes the window value used for each mask 2. measures and writes analog test size, + and - slope values. 3. Measures and writes digital tests. 4. Writes calibration change values, which were used for recalibration.

Recalibration program 202 RTMASK, a disc sub-program as one overlay, run in off-line operation by means, each of which of the following bar diameter measuring functions can be applied: l. An electronic window gate for deflection to the right can be modified to adapt to changes in image analysis tower 90 parameters. 2. To the right, deflected diameter reference values, which are race in common tables, can be updated to compensate for drift, component aging, etc. If no changes are desired, the program can be run cyclically with a deviation printout on printer 63 for observation electronics and temperature-related deviation. Returning from this subprogram, the image the analyzer 90 sweeps to the center an all-electronic window gate is reset and the current through the backlight lamps reversed to extend the life of the lamp. This program is primarily intended as a control tool for long-term deviation. It has the additional ability to update the window gate and the reference table value. It is activated through the SUBCLL subprogram and requires operator intervention.

GAGRCL is a program, which runs on the on-line system. The requires no intervention on the part of the operator. Its purpose is to automatically recalibrate the bar diameter measurement periodically by updating the deviation correction term described above ITMPl. It deflects the camera sweep for sensing the right mask 94 and equalizes the deviation term with each deviation from an initial calibration reference value. Before the end 7802955--0 44 10 l5 20 25 30 35 the sweep is reversed towards the middle by means of a normal window and the backlight source is reversed.

The automatic recalibration system ensures that measurement accuracy is maintained by checking the calorific whenever the rod 10 is not in the measuring field. This recalibration system is performed when a rod 10 leaves the measurement and before the next rod enters, as indicated by a signal from the hot metal detector electronics 57. This is accompanied by inversion of software to calculate a scale factor, based on the differences between an on-line measurement of a known I internal reference, such as the right mask 94, and an off-line measurement of the same internal reference, performed during system calibration.

The measurement of the next rod 10 in the measuring field, as follows after a recalibration, is corrected using it scale factor.

The key to the recalibration measurement is the masked one the photocathode 91 on the front of the image analyzer tube 90.

The deworming pattern is shown in Fig. 5. The photocathode 91 has five 2.54 mm wide masks with 2.54 mm spaces on the left side and on its right side it has a single 6.35 mm wide centered mask. The image analyzer tube 90 and the photocathode 91l construction and working methods have been described above in connection with Figures 4, 5 and 6. There is "C" sensing, "R" sensing and "L" sensing different positions, provided by X- shaft preload. There is no difference between the camera signals from the right mask and from the bar. If no adjustments are made regarding electronics, the measurement with the right mask would at time Tl differences are assumed to be due to electronic deviation. become the same as the measurement at time T2. Any The recalibration system uses only the right mask 94 to calculate the correction factors. The five left masks 95 was used only in the off-line calibration system for linearity checks. The right mask for the camera head 31 measured and preserved on the disk memory by performing the right one the mesh program (RT) in the off-line calibration system. The variable stored in kernel memory on the common data area MSKCOM under the name IMASKl. The data is transferred from the disk memory to it common area MSKCOM in the core memory when the control system is activated. 10 l5 20 25 30 35 40 45 j vsozsss-o The on-line measurement with the right mask 94 is performed through GAGRCL program. married. After the hot metal detector 55 detected that the end of the rod 10 has been rolled and that the measurement is reset, the GAGRCL deflects the image analyzer image to the right and measures the mask 94. The difference between the measured the value from camera l and IMASK1 is stored in the variable ITMPI on the common data area TMPOFF. This value represents changes in the bar position measurement from the original the calibration to the on-line calibration.

The on-line correction function is performed in the subprogram CMPNST, which uses the variable ITMPI. A steepness correction tion is applied to each measurement, based on the following metric formula: For the camera head 31: Corr bar size = bar size - âššggëëgšlåë-Ä-lay; '12.7 (eq.20) As an example for ITMPI = 0,0524 mm: The corrected size (12.7 X 0.01524) for a 12.7 mm rod = 12.7 - = 12,684? IN (eq.2l) The corrected size for a 25.4 mm rod = 25.4 -ÅÉÉLÉ-§59¿9l§3íl = 2s, 3694 IN (eq.22) The corrected size = 38.045l (Eq. 23) for a 38.10 mm rod = 38.10 - (åšeâåïš-QLQÄÉÉÉÄ The correction value for a 12.7 mm thick rod is equal with the value ITMPl} Similarly, the correction 2 X ITMPI is for a twice as thick ie 25.4 mm rod and 3 X ITMPI for one 38.10 mm thick rod. This is because the reduction of the lens is 1: 2.

Thus, a 12.7 mm rod such as 6.35 mm shadow is projected on photocathode 91, which is the approximate width of it right mask 94. 7802955-0 46 l0 15 20 25 30 35 40 LFTMSK, a disc sub-program such as storage, run off-line by means, each of which of the following bar diameter measuring functions can be applied: l. Left-deflected electronic window gates, use to select each of the five left-deflected rods the references on the left mask 95, can be changed for adaptation to changes regarding the 90 parameters of the image analyzer tube. 2. Left-deflected diameter reference values, stored in a common table, can be updated to compensate for yield, component aging, etc. 3. If no changes are desired, the program can be run cyclically with a deviation printout on printer 63 for each of the five left-deflected, etched bar references, to observe electronics and temperature-related deviation. Maximum cycling time is 32,000 seconds.

Upon return from this subprogram, the image the analyzer tube 90 sweep to the center, a fully electronic window gate stored again and the current through the backlight reversed to extend the life of the lamp. This program is designed as a control tools for field of view and electronic operation, with the_additional the ability to update the window gates and reference tables the value. It is called through the SUBCLL subprogram and requires törens intervention.

HISTOGRAM PROGRAM (2o4) GAGHST is a program that runs on the off-line measurement system.

It requires the participation of the operator. Its purpose is to achieve a number of readings from the camera head 31, storage in memory 194, table IBDGI2 and printing a histogram table for the camera head 31, sorted into parts of 0.005l mm for a range between + 0.127 to -0.127 mm, shown in Fig. 17.

In addition to this, it calculates and writes the mean and standard deviation for all readings. The operator must enter the number desired readings, the desired rod size, and request use of the histogram table with the control system 67, as partially shown in Fig. 19.

TWO DIMENSION AND PROFILE MEASUREMENT SYSTEM Referring to the drawings, especially Fig. 1A, is shown where a computerized electro-optical system for measuring two 10 15 20 25 30 35 40 47 vsozsss-o bar dimensions and the profile with two rear illuminated cameras, mounted on a sensing device in a iron rolling mill. This device according to the invention is similar the one shown in Fig. 1 but includes the further improvement that two measuring camera systems and a sensing device are used, which shown in Fig. 1A. The camera's 2 electronics 39 are the same as the electronics 35 for the camera 1 shown therein in Figs. 4-13 except that all reference numerals for devices, circuits, waveforms, time charts, computer programs, etc. have one prime character to denote the electronics of the other camera.

The measuring system measures two at right angles to each other diameters and the profile of the rod 10, e.g. behind the exit page of the roller seat 11, while the sensing device is either stationary or senses the outside of the rod 10 a prescribed one angular displacement. As explained below, they are both added the diameter signals and a sensing position signal one computer, which graphically represents the side profile of the rod 10. Finally bar profile data is imaged, recorded and transferred to a rolling mill control system, which uses this data to control the size and shape of the rod by (a) adjusting it lateral distance between the rollers in the roller chair ll, (b) setting the vertical position of the rollers in the roller chair ll and (c) setting the lateral rolling distance between the rollers in the roller chair immediately in front of the roller chair ll.

More specifically, the double sensing head 12 consists of reversible sensing mechanisms 13, driven by a motor 14, which is fed via a line 15 through an adjustable speed controller 16. A two-way switch 17 provides either manual or automatic sensing operation, such as signaled via line 18 to controller 16. This is because on whether a measurement system operator or computer should perform the desired manual or automatic control of the sensing device 12. For manual steering, manual speed, start-stop and sensing device 12 is provided with directional control in the control device 19, and these signals are applied via the line The controller 16. In the case of automatic control, they are manual the control signal sources deactivated and of the sensing device controller 16 receives corresponding signals from the computer so as explained below.

The sensor position sensor 21 is connected to the mechanism 13 and generates an analog signal representing sensing devices. 43 10 l5 20 25 30 35 40 absolute rotation position of the 12. The sensor signal is applied via line 22 to the sensing mode electronics 23, where it converted to both analog and digital sensing position signals. The analog sensing position signals are applied via line 24 to the sensing position indicator 25, which shall be observed by the measuring operator when the sensing is performed 'with manual control. The digital sensing position signals are supplied via the line 26 to the computer 27 ', where they are connected to computer command signals for automatic control of sensing the device 12. The computer 27 'then generates start-stop signals and speed control signals, as described below. These signals supplied over lines 28 resp. 29 of the sensing device control device 16. In automatic control, they are used digital sensing device signals to determine the bar profile as also described below. 3 The two-headed sensing device 12 mechanism 13 is arranged to support first and second, rear-illuminated electronic camera heads at right angles relative to each other and so that they are both at right angles to each other the bar iron 10 during the peripheral sensing of the bar 10 by a device with a given angular displacement. The sensing for imaging the bar profile is shown in Fig. 1A and Fig. 2 .as a 900 rotation of the sensing device 12. This gives enough camera signals to allow later graphics representation of the cross section of the bar 10 1800. An 1800 profile return giving is very useful for a rolling mill operator and a rolling mill control computer, as described below. In other agreements sensing requirements for measuring the rod size, sensing devices the angular displacement of the ring be other than 900. For example the lamp 30 and the camera head 31 at the one-dimensional measuring system, shown in Fig. 1, be mounted on the sensing device and rotated to provide another type of bar profile return. giving. IN The first lamp 30 is located opposite the first electronic camera head 31, so that when the rod 10 breaks the light from the lamp 30, a rod shadow with one against the rod diameter proportional width in a first transverse position is depicted in it first electronic camera head 31. Similarly, a second lamp 32 located opposite the second electronic the camera head 33, so that when the rod 10 refracts the light lamp 32 depicts a rod shadow with one against the rod diameter 10 15 20 25 30 35 40 49 7802955-0 proportional width in a 900 relative to the first offset transverse position in the head of the other electronic camera 33. many with the first rear-illuminated camera head, which shown in Fig. 4 and described below, is typical of both the camera heads.

Each lamp 30 and 32 is intended to emit a beam of light at right angles to the rod 10, which is wider than the rod 10 largest dimension, to be measured in the camera's field of view.

For example, the field of view of the camera referred to below may be 07.62 cm and the light beam used in this case be 1.06 cm wide. In addition, the wavelength and light intensity of lamps 30 and 32 must be variable with the electronic camera heads 31, 33 sensitivity. Blue light from a direct current fluorescent light source is preferred for the electronic camera head as described above. The first shadow of the rod 10 is directed together with excess light outside the edges of the bar from lamp 30 and causes in the first electronic camera head 3l a first camera signal. This signal consists of an original camera pulse, mixed with interference and fed through the wire 34 to the electronics of the first camera 35. As described above in connection with Fig. 4, the first camera the signal to remove the interference and deliver digital signals regarding rod size and rod position, which via line 36 is supplied to the computer 27 '. Measurement release and others signals are fed via the cable 37 from the computer 27 to the first camera electronics 35.

It together with excess light from lamp 32 on either side of the edges of the rod 10 cast the second rod shadow, simultaneously causes the second electronic camera head 33 to generate a second camera signal. This signal consists of similar way of a primary camera pulse, which is mixed with interference signals and is fed via line 38 to a second camera ktronik 39. The second camera signal is processed to remove interference and provide digital bar size and position signals, the same as the first camera signal, and these are fed via the cable 41 to the computer 27 '. Measurement release and other signals are routed via cable 40 from the computer 27 'to it second camera electronics 39. I The computer 27 'at the electro-optical bar shown in Fig. 1A the measuring system also receives digital bar brush size signals vsøzessßo 4 50 10 I5 20 25 30 40 _from the thumbwheel selector 42 via the cable 43. Bore size signals, for example 4.445 cm, was used to determine the bar profile deviation and for other purposes described below. In addition, the computer 27 'also receives a digital signal regarding the rod 10 assembly from the thumbwheel selector 44 via the cable 45. the setting signal, in the example 0.230%, represents the percentage the content of carbon in the iron bar 10 and was used as a factor in calculation of the hot rod's brush size based on from the cold brush size and for other purposes described below.

Furthermore, the computer 27 'also receives from the source 46 via the line 47 appropriate order data signals containing the date, time and size tolerances for the bar iron 10. Alternatively, some or all can of the stock exchange signals, the composition signals and other data signals are emitted through a control system, which is directly connected with the rolling mill ll, depending on what the bar measuring system users wishes. In order to make temperature corrections for diameter measurements on the movable hot iron bar 10, a. optical pyrometer head 48 adjacent the sensing device 12, intended for the moving hot rod 10. It the optical pyrometer head 48 is arranged to provide a highly sensitive primary temperature signal, as via cable 49 is supplied to the pyrometer electronics 50. This original temperature signal is corrected by grading and linearization circuits in pyrometer electronics 50. The corrected temperature the signal, in the example 910 ° C, is fed via the cable 51 to the digital catheter 52 and via the cable 53 to the computer 27 ', where it is used for thermal expansion compensation. ' Installation problems can obstruct an optical pyrometer head 48 and pyrometer electronics 50 from providing a correct one temperature signal applied to the computer 27 'and the indicator 52 with the desired precision and sensitivity. If so, must an alternative arrangement is to replace the pyrometer with an optical one field sensing including pyrometer systems as mentioned above in the embodiment according to Fig. 1. Another distinguishing feature of the bar measuring system according to Fig. 1A is an automatic recalibration system. As described above, this device is started each time the forward end of the hot iron bar 10 is detected to leave the roller chair ll. For this reason, a hot metal detector detects 10 15 20 25 30 35 40 51 '7802955-0 tektor 55 attendance resp. the absence of hot rod and feeds one corresponding signal via line 56 to the hot metal detector electrode niken 57. An attendance / absence signal is supplied via cable 58 to computer 27 ', where it starts the above-mentioned automatic recalculation the brewing system.

All sensing position signals, first and second camera signals, brush size signal, carbon content signal, other signals, temperature signal and hot metal presence / absence signal, which are carried via associated cables 26, 36, 41, 43, 45, 47, 53 and 58, combined in the computer 27 'to perform a variety of functions when controlling a group referred to below off-line or on-line computer program. One of these functions is to generate start-stop the signal for sensing in the cable 28 and the sensing speed control signals in cable 29, both under automatic sense control ning. Another function is to feed bar diameter data, bar profile deviation data, superimposed on commercial tolerance and start-up data from the computer 27 'via the cable 59 to the CRT terminal 60 and to accept mutual influence between a standard control panel in the terminal 60 and the computer 27 ' via cable 61. Another function of the computer 27 'is to feed the bar diameter. meter data, bar profile data, superimposed on commercial tolerance reference values and work start data from the computer 27 'via the cable 62 to the printer terminal 63 and to accept mutual influence between a standard control panel in the terminal 63 and the computer 27 ' at the cable 64. The printer terminal 63 produces a printout 65, as shown in Fig. 3. Additional functions of the computer 27 ' is to feed bar profile data and measurement system histograms via cable 66 to the control system 67, in response to the corresponding polling signals, which are fed back to the computer 27 'via cable 68. proceeding to Figure 2, a cross-section is shown therein, illustrating the cross section of the rod 10. Dashed circles 69 and 70 show the maximum and minimum deviation from the setpoint diameter at commercial standard product. Furthermore, reproduced by dashed straight lines planes A-A, B-B, C-C and D-D, which are of particular interest for a rolling mill operator and a control computer for determining the rolling the gap and placement relations between the rolling mills 11, are shown in Fig. 11A. When no sensing occurs, it is preferable to allow sensing the resting device 12 rests, at least temporarily, so that the first the camera head 31 and the second camera head 33 measure the diameter in plans C-C and A-A respectively. An A-plane dimension of the rod l0 'rsozass-o p 52 10 15 20 25 30 35 40 shown at 7l as 4,450 cm and the C-plane dimension of the bar 10 is illustrated at 72 as 4.442 cm, the brush size being 4.445 cm for illustrative purposes.

During the rod sensing operation, it is preferred that the other the camera head 33 starts the profile sensing 73 in plane B-B, continued sets counterclockwise 900 through plane C-C and stays in plane D-D. Samti- the first camera head 3l starts the sensing in planes D-D, continues counterclockwise 900 through level A-A and stays in level B-B. On in this way, the first and second camera heads 31, 33 can sense l30 ° of the peripheral surface of the rod, and this sensing is plotted from planes B-B to C-C, D-D, A-A and ends again in B-B. Other sensing methods can be applied. For example, sensing tion clockwise instead of counterclockwise. Furthermore, sensing the order l2 start at any other plan or arbitrary point in between, sense 90 °, return to starting position and thereby allowing each 180 ° sector of the rod to be plotted by rotate cameras 31 and 33 only 90 °. ' The resulting representation of the profile of the rod 10, corrected to cold size, is shown in Fig. 3 in a computer printout 65. Here, the bar profile 74 has been superimposed on a specific size, size tolerance and bar position format, generated by the computer 27 of Fig. 1A. The computer-generated structure includes includes an operation data start. The data profile deviation from current cold brush size, selected by the device 42 in Fig. 1A, is the X-axis variable. Angular position of the sensing device 12 is the Y-axis variable. The Y-axis print is graded at 0.0254 mm parts above and below the dashed brush size baseline 75 and extends beyond the reference lines 76, 77 for tolerance maximum and minimum for commercial goods.

Reference lines 76, 77 are printed as sparsely dashed lines across the X-axis. In addition to this have maximum and minimum for semi-commercial tolerance reference lines 78 and 79 printed across the X-axis as alphanumeric lines with 15 degree angles division on the bar profile representation over 1800. At 0 and var The 45th degree part has the cross-sectional plane designations B, C, D, A and B from Fig. 2 are inserted, while intermediate 150 and 300 angles are printed relative to the A and C positions. _ It can be noted that the image on the CRT terminal 60 is essentially the same as the computer printout 65 with two exceptions.

These are the computer 27 'in addition to the deviation representation of the bar profile and computer-generated structure also generates an additional 10 15 20 25 30 35 see, 7802955-0 representation of those shown in Fig. 2 in broken lines the detection plan A-A, B-B, C-C and D-D as well as the current ones numerical bar sizes A and C, shown with reference to 71 and 72 in Fig. 2. Second, the limits of full tolerance not reproduced when half tolerance is the setpoint of the system. Thus, the CRT terminal 60 shows bar profile, bar diameter and bar sensing plan information about a form that is unique and very useful for an operator for the bar measurement system and also a rolling mill, where the bar measurement apply.

Electronic camera head A typical, backlit, first camera head that was used in the electro-optical bar measurement device shown in Fig. 1A. the procedure, is also shown in Fig. 4 as the camera head 31, located along an optical axis on the opposite side of the rod 10, seen from the lamp 30. The camera head 33 and the lamp 32 are the same as 31, and includes the second electronic camera body. vudet. Depending on installation requirements and user preferences, each first and second electronic camera may include one telecentric lens 85 ', an image analyzer tube 90', photocathode masking masks 94 ', 95', focusing and deflection coil devices 93 'and additional shields, all described above in connection with Fig. 4.

Camera electronics Typical camera electronics used herein electro-optical bar measuring systems, are also shown in Fig. 4 as first camera electronics 35. The second camera electronics 39 is a duplicate of the first camera electronics 35 with except for the two-way sweep generator 97, which is shared by the two camera electronics 35 and 39. Details of the camera electronics 35 resp. 39 is best understood in connection with Figs. 4 and 7 to 13 and the above description of the measuring system according to fig. l. As a consequence, the first and the second camera electronics 35, 39 separately belonging to primary camera signals and produces first and second digital bar size pulses and bar center position pulses, as via wires 36 and 41 are supplied to the computer 27 'under the control of the respective signals in lines 37 and 40. Computer correction of each and one of the bar pulses is described below. veozessfo a 10 15 20 25 30 40 1259: - ”.

A block diagram of the electro-optical shown in Fig. 1A the computer 27 'of the bar measuring system is shown in Fig. 14A. Computer 27 ' is a digital system, programmed to perform all of FIG. l4 shown functions and various other functions described below.

As mentioned above, a commercially available, programmable or firmware-equipped mini-computer can be used. Alternatively may, if desired, the computer 27 'form part of a control computer installation for the entire rolling mill. The computer 27 'is in the present example made by Westinghouse Electric Co., USA, model W-2500, with an extended operating system for machining a second rod size pulse and correct the same, sensing position control and profile reproduction, as well as for adjustment of different program levels as indicated below.

The computer 27 'is equipped with conventional main components. including input buffer memory l90 ', output buffer memory l9l ', disk memory l92', disk switch l93 ', core memory 194' and all communication devices for different channels with data The processing unit l95 '. The operations of the computer 27 'are controlled sequentially. according to off-line and on-line computer programs l96 '. These includes: computer cards l97 ', shown in Figs. 15, 16A and 16B, program l98 ', bar measurement data program l99', compensation program 200 ', calibration program 20l ', recalibration program 202', profile and position program 203 and histogram program 204 ', all described below. _ All communications to the bar measurement system computer 27 ' from external sources, takes place via the input buffer memory l90 ', which includes devices for converting analog and digital input signals to digital form. These include signals, fed via wires or cables to the computer according to the following: first camera electronics 35 on line 36; Other camera electronics 39 via calf 41; mechanical sensing position 23 via line 26, hot metal detector 57 via line 58, rod temperature 50 via cables 53, 54; rod brush size 42 via line 43; rod assembly 44 via conduit 45; Other data 46 via cable 47; control system 67 via cable 68; CRT- terminal 60 via cable 61; and printer terminal 63 via cable 64.

All communications from the bar measuring system computer 27 ' to external sources takes place via the output buffer memory l9l ', which also includes means for converting output signals 10 15 20 25 30 35 40 '7802955-0 55 to digital or analog form. These include signals, which are fed from the computer via wires or cables according to the following: sensing start-stop 16 via cable 28; sensory speed control 16 via cable 29, control system 67 via cable 66; first camera electronics 35 via cable 37; other camera electronics 39 via cable 40.

Individual wires in signal cables are used in the drawings and these have been connected in accordance with their sources and function, as described above.

The CRT terminal 60 includes a control panel for the operator must be able to intervene in the computer 27 '.

The printer terminal 63 includes a control panel for the operator to intervene in the computer 27 '. Printer terminal 63 computer printing 65 includes a graphical representation of bar profile deviations shown in Fig. 3 and shown below in the form of a list data tables.

It is generally allowed for both terminals 60 and 63 to graphically reproduce the same data. All interventions from either the control panel is done via program lists of abbreviated designations, for example as follows: OFF LINE MEASUREMENT SYSTEM THE ABBREVIATED NAMES ARE AS FOLLOWS: HS - HISTOGRAM FOR EACH HEAD MP - CREATES COMPENSATION CARD FOR EYE FIELDS PR - SWITCHES THE SENSING DEVICE 900 AND ESTABLISHES g PROFILE TABLE PL - ÅmERGER GRAPHIC PROFILE TABLE RP - CREATES A PROFILE TABLE FOR RIGHT MASK DATA CL ~ PERFORM CALIBRATION CHECK OF LEFT AND RIGHT MASK TY - TRXCKER SHORT, FIRE AND SHIFT FACTORS AS WELL AS THE MASK VALUES SC - ROTATES SENSING DEVICE TO DESIRED ANGLE OF - ALLOWS ENTRANCE IN FIRE AND DISPLACEMENT CORRECTION FACTORS ZE - RESETS ALL CARDS AND CORRECTION FACTORS EEE WARNING! LF - LEFT MASK OFFSET TEST RI - RIGHT MASK SHIFT TEST (ALSO ALLOWED WINDOW ENTRANCE) TR - DISC TRANSFER OF MEASUREMENT COMMON FOR STEERING SYSTEM AREA 7802955-0 56 XT - EXIT TO MONITOR AND ALLOW PRINTING ON COMMON AREA, INCLUDING CARDS, FIRE AND SHIFT CORRECTION FACTORS, MASK VALUES AND WINDOW VALUES TO DISK MEMORY. DISK MEMORY IS UPDATED ONLY IF THE MEMORY SWITCH l2 IS UP. THIS REGISTER IS READ FROM MEMORY WHEN THIS PROGRAM (20) IS CALLED OUT MONITORNf The memory switches 193 'include switches designated "switch l0" and "switch 12" in the programs below switches must be turned to "WRITING POSSIBLE" to date programs or data on the memory.

DATA PROGRAM The table below indicates in flow form flow tables for individual programs and program groups, associated with computer program l96 ', which is used here.

COMPUTER PROGRAM IDENTIFICATION USE 'OFF-LINE ON-LINE KQRT (l97 ') DISK MEMORY CARD DEAR fl MEMORY CARD SERVICE PROGRAM (19ß ') IDL HANDLER M: IDL cD = IDL EB: IDL GAGTSK SUBCLL GAGTRN BAR MEASUREMENT DATA PROGRAM (199 ') GAGEIN KQMPENSATIQNSPROGRAM (200 ')) GAGMAP CORDAT zero MAPRNT GAGTPC CMPNST CALIBRATION PROGRAM (201 ') CALIBR RECALIBRATION PROGRAM (202 ') RTMASK GAGRCL LFTMSK PROFILE AND POSITION PROGRAM (203) ENCNGL .GAGPOS PROFILE RTPROF PLor GAGPLT HEADER GAGPRO NN N M xxxwx x N w wxwwww w xwwxww www M xxxwx (continued) 10 15 20 25 35 7802955-0 57 SHORT (l97 ') IDENTIFICATION USE OFF-LINE on-LINE HISTQGRAMPRQGRAM (2o4 ') GAGHST x PROFILE AND HISTOGRAM INTERFACE man Steering system x X DISC MEMORY CARD, see Fig. 15: The program address in the disk memory 192 'is extended to handle the additional above-mentioned operating devices.

CORE MEMORY CARD, see Fig. 16A, B: the program address in the decimal core memory 194 'is also expanded to handle the above additional surgical devices.

SERVICE PROGRAM (l98 ') The IDL control program, including M: IDL, CD: IDL and EB: IDL as well as the GAGTSK, SUBCLL and GAGTRN programs, have all been described above in connection with Fig. 14, except that they are simple expanded to handle the above mentioned for Fig. 14A future surgical devices.

BAR MEASUREMENT DATA PROGRAM (199 ') GAGEIN is a help subprogram, the same as described above, with the exception that it is extended to accommodate the extra bar measurement data coming from the other camera electronics 39.

COMPENSATION PROGRAM (200 ') Programs GAGMAP, CORDAT, ZERO, MAPRNT, GAGTPC AND CMPNST is also the same as described above, with the exception that all are extended to accommodate the extra bar measurement data ooh niken 39.

CALIBRATION PROGRAM (2o1 ') the correction requirements arising from the second camera CALIBR is an independent program, the same as described above, with the exception that it is extended to suit for the additional bar measurement data and calibration requirements coming from the other camera electronics 39.

QM CALIBRATION PROGRAM (202 ') The subprograms RTMASK, GAGRCL and LFTMSK are same as described above, with the exception that also each and 7802955-0 se 10 l5 20 25 30 35 40 one is extended to suit the additional bar measurement data and the automatic recalibration claims that come from the other camera electronics 39.

PROFILE AND POSITION PROGRAM (203) ENCNGL is a new help sub-program, which has been added for each subprogram, which requires the Angle Position for the beginning of the bar diameter measurements. It reads the position code electronics 23, checks validity, picks both binary and decimal values for the position into common area and sets an error mark in case the decoder fails.

GAGPOS, a new sub-program belonging to the disk memory such as an overlay, runs under the off-line system and requires the intervention of the operator. It is called through the subprogram SUBCLL by the abbreviation SC. Its purpose is to drive the scanner to an angular position input via the terminal key table 60, 63. The following comments make it easier to understand the program. l. When the target angle is greater than l0 ° from the scanner position, full speed voltage is supplied across the cable 29 to the sensing motor control l6 to drive towards the target angle. Is the target angle smaller than 1o °, go to step 3. 2. Continue at full speed until the scanner is within 10 ° of the target. At less than 10 ° from the target angle, the voltage is reduced to half speed via the controller 16. 4. Within 0.3 ° from the target angle, the controller 16 is applied zero volt voltage and shutdown occurs.

The operator must now set the target angle via the keyboard.

PROFILE is another new program that runs offline the measuring system. It requires the participation of the operator. Its purpose is to detect the camera through a complete 900 cycle and that set up a profile table, Fig. 18, which contains the deviations with 20 division IBDGTL (l94 '). The program does not reproduce these data graphically. The PL (PLOT) playback program runs while the off-line system performs its work.

Three possible error situations can arise. 1.-Sensor motor fault - indicates that the motor did not run started or that there was no end to the sensing cycle (0 or lO l5 20 {\ J UI 30 35 40 59 7802955-0 2. Decoding error ~ occurs if the finished bit does not was generated by the decoder. 3. IDL errors occur if a timeout occurs in the case of an IDL transfer.

RTPROF is another new program, which is run under the off-line measurement system. Its purpose is to divert to it right mask on both cameras when the cameras scan over one complete 900 cycle and to set up a profile table, which contains the deviations with division IBDGT1 (94). The program does not render this data graphically. The plotting program PL is run while the off-line system performs its task.

There can be three possible error situations. l. Sensor motor fault - indicates that the motor did not run started or there was no end to the sensing cycle (0 or 900). 2. Decoding error - occurs if the finished bit does not was generated by the decoder. 3. IDL errors occur if a timeout occurs in the case of an IDL transfer.

This program redirects each electronics "R" scan to the right mask 94 in Fig. 5 before starting the profile and redirects back to the electronics "C" center scan after that the "R" scan is complete.

PLOT is another new program that runs under off ~ line measurement the system. It does not require the operator's participation. Its purpose is to plot, ie. graphically represent the data included in profile table IBDGT1, stored in the core memory l94 '. Y-axis put 10 rows above the shoulder and 1 row below the shoulder.

The scale flows with a minimum of 0.005l mm. The deviation is reproduced along the Y-axis and the angular position of the scanner are displayed along X ~ areln with a division of 40 per column. Data points, that is blank or out of range, represented by "".

GAGPLT, another new on-line program, takes 90-element the IBDGTI profile cable stored in the core memory l94 ', from a common area, termed MASGAG and compresses it into a 60-element table for use shown in Fig. 19. Each table entry now represents three degrees. The program scans the table and determines which Y-axis pitch to use, based on the maximum and minimum values in the profile table. This pitch is either D, 0254 mm or 0.0508 mm. After this '7802955-0 6 ° UI l0 15 20 25 30 35 40 the program writes the brush size tolerance lines on the cathode steel tube CRT and in the printer terminals 60.63. The program calculates then the position of the Y misplacement for every 3rd degree table entry and writes an asterisk on the CRT and the printer terminals 60, G3, which corresponds to this X and Y location.

Finally, it calls the HEADER program and exits. One bar profile display using the GAGPLT program is shown in Fig. 3 as a printout 65 from the print terminal 63.

HEADER, another new on-line program, prints the brush size of the rod in the cold state, carbon content and temperature on CRT 60. It then prints the date, time, maximum tolerance, minimum tolerance and roundness tolerance, also on CRT 60.

Then it profiles the profile table IBDGT1 and calculates plus measure, minus measure and roundness, based on respectively tolerance limits. Then it prints these values that are displayed in Fig. 3 and ends.

GAGPRO is another new program, which is running under the on-line measurement system. It does not require the operator's effect. Its purpose is to detect the camera heads 31 and 33 over a complete 900 cycle and to set up a profile table, containing the deviations with division IBDGT1 (194). The does not print this data.

There are three possible error situations that can occur. l. Sensor motor fault - indicates that the motor did not run running or that there was no end to the sensing cycle (01 or 90 °>. 2. Decoding error - occurs if the finished bit does not was generated by the decoder. 0 3. IDL error - occurs if a timeout occurs in the case of an IDL transfer.

HISTOGRAM PROGRAM (204 ') GAGHST is an additional new program, which is run under on-line and off-line measurement system. It is a modified version of program 204 and requires the participation of the operator. Its purpose is to collect a number of readings from each camera head 31, 33 while positioned in planes "A-A" and "C-C" in FIG. 3 or in another mode, to store the readings in tables lO 61 7802955-o IBGDT2 and IBGDT3 for the 194 'core memory and to print one histogram for each camera head 31, 33 with a division of 0.005l mm in a range between + 0.127 to ~ 0.127 mm as shown in Fig. 17. In addition to this, it calculates and prints mean ~ and standard deviation of all readings from each camera head 31, 33. The operator must enter the desired number readings and the brush size of the rod and order use of each histogram cell IBGDT2 and IBGDT3 together with the profile table IBGDTI with the etyr system 67 shown in Fig. 19.

Claims (1)

10 15 20 25 30 7802955-0 62 PATENT REQUIREMENTS Electro-optical device for continuous compensation of error signals, which in case they are not compensated prevents obtaining a desired transverse dimension of a longitudinally movable metal rod, subjected to disturbances in the form of lateral vibrations during the longitudinal movement, characterized by, for compensating for spatial movement of the rod, which leads to said lateral vibration disturbances, an electronic camera device is provided, which comprises a telecentric lens system for converting an image of the longitudinally movable rod, subjected to lateral vibrations, to a primary, unprocessed camera signal, combined with error-causing signals caused by noise and other error-causing signals resulting from temperature variations, optical or non-linear errors; electronic circuit means, comprising means for processing the raw camera signal to generate a bar size pulse, which varies as a function of one or more of said error signals, said means for processing the raw camera signal comprising differentiated circuits for detecting pulse edges for each raw camera signal and an autocorrelator for removing electrical noise signals from each differentiated raw camera signal, means for measuring the temperature of the rod to provide a temperature compensating signal, means for measuring the position of the rod in the field of view of the camera means for providing a compensation signal for the position of the rod, computer means for removing said second error-causing signals 1 depending on said temperature and rod position compensating signals and providing a completely corrected rod size pulse, which is stored in the computer, and means using the stored the pulse signals to achieve get enough 'ann measurement and control of the rod size. '7802955-U 53 ...._. «-. Electro-optical device according to claim 1 for measuring more than one transverse dimension of the moving bar, characterized in that the electronic camera device is modified to convert any bar image of each measuring dimension to a corresponding number of primary camera signals, of which at least one varies as a function of said one or more errors, that the electronic circuits are modified to process each variable primary camera signal to generate a corresponding number of variable bar size pulses, at least one of which may vary as a function of one or more of said errors, that the computer is arranged to receive each variable bar size pulse and compensate for at least one bar size pulse depending on one or more corresponding error signals received by the computer to generate and store a corresponding number of corrected bar size pulses, and that means of use are arranged to use the stored data to indicate and / or register each name d corrected rod size. 7802955-'0 w Device according to claim 2 or 2, for measuring one or more transverse dimensions of a moving rod 1 in different peripheral positions relative to the rod, characterized in that the measuring means comprise sensing means, arranged to controllably carry one or two electronic cameras 1 the camera device around a cross section of the rod, wherein a scan position signal is generated and said computer is further arranged to graphically reproduce and store the rod profile as a function of one or more corrected rod size pulses and said scan position signal and said use means is arranged to use the stored data. or register the bar profile with or without corrected - bar size. 4. Device according to claim 3, characterized in that the sensing means comprises a control device which is operated by a scanning control signal. Device according to claim 4, characterized in that the computer is arranged reversibly for automatic control of the sensing device along a predetermined peripheral part of the rod 1 in accordance with a suitable order signal. Device according to one or more of the preceding claims, which generates one or more variable rod size pulses which require correction for one or more linear or non-linear errors from optical or electronic sources, characterized in that the computer is modified to compensating one or more of the variable rod size pulses for said error in accordance with a corresponding error compensation signal received by the computer to efficiently output one or more corrected rod size data. Device according to one or more of the preceding claims, which emits one or more variable rod size pulses which require correction for one or more errors, regarding camera field of view, displacement factor, deviation factor, rod temperature or the effect of the rod composition on the rod temperature in relation to the rod cold size, k characterized in that it further comprises bar temperature and bar composition signal counts is that the computer is modified to compensate one or more bar size pulses for one or more of the above errors 1 depending on associated error compensation signals which the computer receives to efficiently emit one or more several corrected bar size data. Device according to one or more of the preceding claims, characterized in that at least one electronic camera illuminated from behind is included in said camera device. Device according to one or more of the preceding claims, characterized in that at least one electronic camera in the camera device comprises a telecentric lens system for allowing imaging of bar movements anywhere in a predetermined field of view. 10. Device according to any one or more of the preceding claims, characterized in that at least one electronic camera in said camera device comprises an image analyzer, which is intended to be scanned electronically and that the electronics further comprises a scan generator, which drives the scan of each image analyzer. Anory device according to claim 10, characterized in that the scan generator is connected for scanning the image analyzer. Device according to claim 10, characterized in that the sweep generator is connected to a linear, two-way sweep cycle, which has equal sweep-up and down-sweep half-cycles. Device according to claim 10, characterized in that the sweep generator is connected for a non-linear two-way sweep cycle. Device according to one or more of the preceding claims, characterized in that the electronic camera device comprises a variable gain image analyzer and that the electronics comprise a self-balancing 7802955-0 kb measuring loop with automatic gain control circuit for varying the image analyzer output current and maintaining the output current. Device according to one or more of the preceding claims, characterized in that the device for processing the primary camera signal comprises an autocorrelator for removing camera signal interference. Device according to one or more of the preceding claims, characterized in that the device for processing the primary camera signal comprises a differentiated pulse edge detection circuit for each primary camera signal and an autocorrelator for removing interference from each differentiated primary camera signal. Device according to claim 12, characterized in that the circuit electronics comprise means which are actuated by one or more variable rod size pulses to emit respective first error compensation signals in accordance with the centerline position data of the rod for each rod image and that the computer is modified to receive the first error compensation signals to effectively compensate for one or more variable bar size pulses according to a predetermined value for associated position data for the bar centerline. Device according to claim 17, characterized in that each position information for the center line of the bar is given depending on the detection of successively variable variable bar size indicating flanks in the associated sweep and downswing half of a two-way sweep cycle of the camera and determines the position of the bar centerline. of the distance between the successive flanks of the successive bar size pulses. Device according to any one or more of the preceding claims, in particular claim 3, characterized in that it further comprises a source of rod brush size data, modifies the computer to graphically reproduce and store rod deviations from the setpoint as a function of the rod size and the aforementioned weight. data according to an appropriate order signal and that the means of use use the stored data to indicate and / or record the rod size deviation from the brush size. Device according to claim 19, characterized in that it comprises a source for bar size tolerance data, that the computer recording device and memories are modified to superimpose bar size tolerance data from said source after a suitable command signal and that the user means uses the stored data to indicate and / or register rod size tolerances, superimposed on the rod deviation from the brush size. Device according to one or more of the preceding claims, characterized in that it further comprises a work data source, which is connected to and stored in the computer according to a suitable order signal and that the use means use the stored work data, for indicating and / or record work data with and without the above other data. IN Device according to one or more of the preceding claims, characterized in that the date is modified to provide calibration of the system, a standard bar being used to build a memory card or to recalibrate the measuring system without the use of a bar, in both cases after an appropriate order signal. Device according to claim 10, characterized in that the image-sensitive device comprises one or more calibration masks, that the circuit electronics further comprises means for displacing at least one sensing from a central rod imaging sweep to one of the calibration-wishing masks and further comprising means for recalibrating of the measuring system without any rod by checking the selection and use of each calibration mask. Device according to claim 23, characterized in that the computer is arranged to perform the selection and use each of the calibration masks according to a suitable order signal. 10 15 20 7802955-0 BB An electro-optical method for measuring a transverse dimension of a longitudinally movable rod, not drawn - of the steps, of imaging a rod dimension to be measured, in an electronic camera device, of converting the rod image into a primary one, unprocessed camera signal, which may vary as a function of electrical noise signals or other erroneous signals, to process the unprocessed camera signal with a telecentric lens system and electronic circuits to produce a rod size pulse which may vary as a function of said electrical noise signals or said second error-causing signals, removing the electrical noise signals from the differentiated unprocessed camera signals, compiling the variable rod size pulses in a computer and said second error-causing signals and calculating a correction factor to compensate for the variable rod size pulse to provide a corrected rod size pulse, to store said corrected rod size pulse and to use the stored data to indicate and / or register the correct rod size. h., 78029S5-U A method according to claim 25 for measuring more than one transverse dimension of the moving bar, characterized in that each bar size for measurement is imaged in an electronic camera and converted to a corresponding number of primary camera signals, of which at least one may vary as a function of said one or more errors that each variable primary camera signal be processed to produce a corresponding number of variable rod size pulses, at least one of which may vary as a function of one or more of said errors, that each said variable rod size pulse and one or more corresponding error compensation signals are compiled in the computer and that a correction factor is calculated to compensate each variable rod size pulse for each said source of error depending on a corresponding number of error compensation signals, after which each corrected rod size pulse is produced and stored and stored and register each cow rigged rod size. A method according to claim 25 or 26 for measuring one or more transverse dimensions of a movable rod in different peripheral positions relative to the rod, characterized in that it further comprises the step of sensitively sensing with one or more electronic cameras in the camera device a 7802955- Cross section of the bar while generating a sensing position signal, thereby modifying the compilation step to graphically display and store the bar profile in the computer as a function of one or more corrected bar size pulses and of the sensing position signal and to modify / use the use step to use the data or record the bar profile with or without a corrected bar size. 28. A method according to claim 27, characterized in that the sensing of said rod is reversibly controlled either manually or automatically depending on a sensing motion control signal. A method according to any one or more of the preceding claims, 25-28, characterized in that the assembling step is modified to compensate for one or more of the variable rod size pulses for one or more of the linear or non-linear errors from optical or electronic sources. A method according to any one of the preceding claims, 25-29, characterized in that it further comprises the step of generating a bar temperature signal and, if desired, the further step of generating a bar composition signal and that the assembling step is modified so as to , that the computer calculates one or more correction factors and compensates one or more of the variable rod size pulses for any of the errors, including camera field of view, displacement factor, deviation factor, rod temperature or the effect of rod composition on rod temperature related to the cold temperature of the rod. a A method according to any one or more of the preceding claims, 25-30, characterized in that it further comprises the step of generating one or more signal sources regarding the bar brush size, bar size tolerances or working data, modifying the assembly step to graphically represent and store the bar size deviation from the setpoint and, if desired, storing bar size tolerance and / or work data in the computer depending on an appropriate order signal 7802955-0 7 / and further modifying the usage steps to use the stored data to indicate and / or record the bar size deviation from the setpoint, that superimpose said bar size tolerance data and / or working data. A method according to any one or more of the preceding claims, 25-31, characterized in that it further comprises the step of calibrating the measuring system, using one or more standard size rods to provide the rod image for subsequent conversion to a rod size pulse and changing the compilation step of building a memory card in the computer in response to an off-line order signal, thereby allowing effective comparison of current rod size pulses with stored values depending on an on-line order signal. Method according to one or more of the preceding claims, 25-32, characterized in that one or more image analyzers are arranged in the camera device, each with one or more calibration masks adjacent to a central rod imaging part, to further electronically sense the central part of the image analyzer occurs during normal measuring operations, that a recalibration of the measuring system is added without bar by shifting the electronic sensing of an image analyzer by controlling the selection and use of each calibration mask 1 depending on a suitable order signal, thereby enabling an efficient comparison of each image analyzer output signal recalibration with a known standard for this.