JPS6055763B2 - thickness measuring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention generally relates to a contact type thickness measurement technique, and more specifically, to a range adjustment technique in a thickness measurement device that indicates the thickness deviation between the thickness of an object to be measured and a thickness standard piece. Regarding.

Prior art non-contact thickness measuring devices usually use a suitable radiation source and a radiation detector placed opposite the radiation source, with the detector placed between the radiation source and the object to be measured. They are exposed to the radiation that comes out.

The object to be measured is usually in the form of a movable thin plate, and as it moves, it blocks the radiation beam from the radiation source. The intensity of the radiation emerging from the object to be measured is received by a detector and amplified before forming the basis for comparing the thickness of the object to be measured with a standard piece of a certain thickness. In order to establish a thickness standard value, such a thickness measuring device injects one or more elements or standard pieces of precisely known thickness into the radiation beam, depending on the command. It is common to use standard single-piece magazines. The thickness measuring device is calibrated such that an analog indicator coupled to the output of the detector indicates zero at a certain nominal thickness of the standard in the beam. The thickness deviation of the object to be measured that deviates from the nominal thickness value is displayed as the indicator deviation. This type of thickness measuring device is intended for use in measuring thickness over a relatively wide range.

In order to reduce the number of standard pieces required by such thickness measuring devices, conventional thickness measuring devices use so-called one-complementary (C
It is structured around the J standard piece system. Briefly, the system divides the total measurement range into a number of smaller sub-ranges, typically into quarters, and for each of those sub-ranges, a suitable predetermined precision thickness of four. It uses two disks. Calibration of each auxiliary range is normally performed on one maximum thickness standard piece (so-called one basic standard piece J). If the basic standard piece is removed from the radiation beam, it is possible to measure the thickness of a workpiece thinner than the basic standard piece within its auxiliary range. This is accomplished in a thickness measuring device that utilizes complementary standards by appropriately energizing the standard magazine so that a certain complementary standard, which is thinner than the basic standard, enters the beam at the same time as the object to be measured. . The thickness at which this indicator is calibrated so that the analog indicator initially points to the "zero point" is the thickness of the object to be measured plus the thickness of the complementary standard piece, i.e., the thickness of the basic standard piece in that specific thickness range. These complementary standards are selected to compensate for the thickness of the workpiece so that it is equal to .The deviation in the thickness of the workpiece is
It is indicated as the deviation from the zero point, or midpoint, of the scale of the indicator. As is well known, the quality of a thickness measuring device of the type described here is indicated by its signal-to-noise ratio.

The noise factor in this ratio has, by definition, a value of zero when there is only a basic standard in the beam. The reason for this is that it is this basic standard piece that the thickness measuring device indicates as zero. Therefore, as the relative difference in thickness between the object to be measured and the basic standard increases, the magnitude of the noise factor in the signal-to-noise ratio increases and the quality of the thickness measuring device decreases accordingly. do. Problems to be Solved by the Invention The present invention solves the problem by solving such a one-complementary problem.
J) It is not a standard piece system.

As mentioned above, a basic standard piece and a complementary standard piece are required. The cage is zeroed with a base standard, and then measurements of the thickness of the workpiece are made by placing a complementary standard along with the workpiece in the beam path. This approach increases the noise figure. The present invention does not insert a standard piece into the beam path together with the object to be measured. The object to be measured is the only material in the beam path. It is possible to measure the same measurement range without requiring too many standard pieces than the number of standard pieces required by a thickness measuring device that targets a certain thickness measurement range, which is constructed around a complementary standard piece system. It is clear that it is advantageous to have a constant minimum noise rate over a range. Furthermore, according to conventional thickness measurement techniques, calibration of non-contact radiation thickness measurement devices generally includes alloy compensation, zero point adjustment, and range adjustment.

However, in conventional non-contact radiation thickness measuring devices, these calibrations, particularly range adjustments, have been mainly performed manually. Manual proofreading not only introduces errors by the proofreader; Calibration takes a long time, which is disadvantageous in situations where rapid calibration is required for various alloyed metals of different nominal thicknesses, such as encountered in rolling operations. Therefore, it is desirable to quickly and completely recalibrate the thickness measurement device as often as desired, with as little human involvement as possible. The purpose of the present invention is to
The object of the present invention is to automate the range adjustment for the calibration of a non-contact radiation thickness measuring device to speed up and improve the accuracy of the calibration work, and to eliminate the drawbacks of the conventional device that performs manual calibration as described above. Means and Function for Solving the Problems In order to achieve the above object, the thickness measuring device of the present invention includes:
A range adjustment means is provided that includes a range setting mechanism, a control circuit, an analog-to-digital converter, a digital signal generation circuit, a digital-to-analog converter, a feedback loop, and a range adjustment confirmation circuit. .

First, the range setting mechanism sets the range of deflection of the pointer of the analog indicator to the first range set in the zero point adjustment mode.
Select the thickness as a percentage of the thickness of the standard piece in the set.

A control circuit generates a digital signal representative of the selected range, which causes a standards setting mechanism to place a second set of standards in the radiation path having a thickness corresponding to the digital signal. Since the analog indicator should have been range-adjusted using a different second standard strip group in the previous measurement, it does not indicate full scale continuation for the new second standard strip group.

An analog-to-digital converter converts an analog signal representing the difference between the actual pointer position and the zero point position of such an analog indicator into a digital output signal representing the absolute value and polarity of the analog signal. The digital signal generation circuit is composed of, for example, a subtraction circuit and a reversible counter, in which case the subtraction circuit subtracts the digital output value of the analog-to-digital converter from the digital setting value representing the full scale of the analog indicator. , the reversible counter increases or decreases the count from the initial value depending on the polarity of the subtracted value.

The count output of this reversible counter is converted into a corresponding analog signal by a digital-to-analog converter and is further fed through a feedback loop to a variable gain amplifier, thereby controlling the output gain of the detector. Further, the range adjustment confirmation circuit includes a counter, and counts the subtracted value output from the subtraction circuit, and when the count value falls within a predetermined range, generates a zero point adjustment confirmation signal, thereby adjusting the gain of the variable gain amplifier. Set the analog signal to the value corresponding to .

Effects of the Invention The range adjustment of the present invention is performed by the operation of a servo loop using the above-mentioned digital circuit.

Adjustment is therefore quick and accurate. Further, by providing a range adjustment confirmation circuit, the range adjustment operation can be terminated when a value substantially close to the arbitrarily determined full scale value is reached, so that the range adjustment can be performed with desired accuracy. In other words, it is possible to freely select whether to emphasize measurement accuracy or speed depending on the purpose of the work. In the example drawings, thickness information is given by the indicated deviation of the analog indicator for the deviation of the thickness of the material to be measured, such as sheet metal or sheet material, and the thickness information is automatically measured for different thicknesses of the same material or different alloys. A thickness measuring device for calibration is shown and will be described below.

Complete calibration of the thickness measuring device for the desired thickness is accomplished by the operation of two servo loops in addition to the alloy compensation of the present invention.

These loops operate in succession to zero the analog indicator to the desired nominal or apparent thickness;
Also, the indicator is calibrated with a programmed deviation from zero so that the thickness deviation from zero is indicated on the appropriate indicator scale and scale value. In particular, the first loop is a zero adjustment loop that adjusts either the high voltage of the radiation source or the amplification of the radiation detector to place the selected standard into the radiation beam emitted by the radiation source. to obtain the zero point of the indicator. The selected standard has a thickness equal to the nominal or, if necessary, apparent thickness of the object to be measured in the radiation. This apparent thickness is corrected by a coefficient that compensates for the influence it has on the radiation absorption coefficient of the alloy that is the object to be measured, and therefore on the intensity of the radiation beam coming out from the object to be measured. means the nominal thickness. The second loop is a range adjustment loop that adjusts the gain of the variable gain amplifier that couples the output of the detector to the input of the analog indicator to provide an analog indicator that covers the entire range of expected thickness changes of the workpiece. It operates to change the meter until a predetermined deflection is obtained. The indicator is also automatically calibrated to a predetermined sensitivity for the expected range of nominal thickness deviations. The nominal thickness used in this text is for both the standard piece and the object to be measured.

The nominal thickness is the thickness at which the object to be measured is manufactured to have, for example, 0.01 cm. Therefore, the nominal thickness of the object to be measured is 0.01CrfL in this case. Therefore, in order to test this object, a standard piece with a thickness of 0.01 CWL is selected. 1. Apparent thickness applies only to the object to be measured, not to the standard piece.

If the object to be measured is made of a different alloy from the standard piece,
Compensation, ie joint compensation, is required. The means for automatically performing alloy compensation, zero point adjustment, and range adjustment will be explained in detail below with reference to the drawings. The non-contact thickness measuring device according to the invention, shown in FIG. 1, includes a suitable radiation source 10.

This radiation source can use an X-ray generator or a radioisotope. A standard strip magazine 12 placed on top of the radiation source 10 contains a plurality of standard strips whose thickness and alloy composition are precisely known. It has a standard piece of metal.

For the purpose of calibrating this thickness measuring device, only a preselected standard of a plurality of standard pieces is inserted into the path of the radiation by actuation of the corresponding standard piece drive solenoid. Measurement mode is distinguished from calibration. The problem with this measurement mode is that a plate-shaped object S to be measured, whose thickness deviation from the preselected standard piece is to be measured, is placed in the path of the radiation beam instead of the standard piece. The standard piece consists of a number of discs, each mounted on a separate arm, which can be selectively positioned within the radiation beam by means of a number of corresponding solenoids. There is.
Each disc is made of metal of the same known composition and is coded to have a different thickness value. By giving the standard piece a thickness value based on the binary w-ary system, 1eIf9.
using a disc with a thickness of, for example, 0.0001 inch to 0.
．． 9999 inches (approximately 0.00025 to approximately 2.54 cw
It will be appreciated that any of the four w-adic combinations of thickness up to L) can be selectively inserted into the beam.

A 4-digit binary code (hereinafter referred to as BCD) representing the desired thickness value.
Standard strip magazines are available in the art that are responsive to an input signal (referred to as . . . ). Inside the standard magazine or the object to be measured S
The influence of radiation absorption by the selected disk is detected by a detector 14, known per se, placed opposite the standard magazine 12 and the object S to be measured.

Detector 14 can be a conventional scintillation type photomultiplier tube. The voltage output of this photomultiplier tube is a function of the radiation absorption coefficient of the material inserted into the beam, and the magnitude of this output is inversely proportional to the thickness of the material inserted. Other types of detectors can also be used instead of scintillation type detectors. The output of the detector 14 is fed to a logarithmic amplifier 16
This amplifier 16 provides a voltage output that is approximately a linear function of the thickness of material placed in the path of the radiation beam.The output of amplifier 16 is passed through a variable gain amplifier 18 to a suitable Indicator, usually analog type instrument 2
Added to 0. A feedback control signal, described below, is applied via lead 22 to the input of amplifier 18 to automatically adjust the gain of amplifier 18 so that the scale of indicator 20 properly correlates to the thickness within the desired measurement range. Then, the indicator will be given a bias of the schedule. Next, with reference to FIG. 2, a digital logic circuit used to provide a BCD thickness signal input to a standard strip magazine for alloy compensation calibration of a thickness measuring device will be described.

Alloy Compensation: The selection of the nominal thickness for the object to be measured is performed by the thickness setting mechanism 30.

This mechanism is approximately 0
．． 00025C71 (0.0001 inch) steps with four manually operable knobs having a w-adic scale suitably graduated and logic providing parallel bit BCD output signals representing any particular setting of these knobs. It is equipped with a circuit. As mentioned above, the radiation absorption coefficient of a material is a function of the material's composition.

Therefore, if the composition of a material differs from that of the standard, the calibration of the thickness measuring device based on the nominal thickness of the standard will be inaccurate for measuring the thickness of that material. . The amount of compensation required for a given thickness of the object to be measured is based on the type of radiation used, the nominal thickness and composition of the standard. If the object to be measured is metal coming out of a rolling mill, the rolling mill provides useful information regarding the apparent thickness of the rolled metal. The difference between the nominal thickness and the apparent thickness is due to the absorption coefficient that the metal has for the type and intensity of radiation. This data can be expressed in terms of percent compensation required, which is applied to the thickness of the standard so that the compensated nominal thickness of the standard is equal to the nominal thickness of the workpiece. It is a kimono.

The required compensation, which can be expressed numerically to approximately 0.1%, is entered into the compensation setting mechanism 32.

The polarity of this compensation, i.e. positive or negative, is the compensation sign mechanism 3
It can be placed in 4. The mechanism 32 has a plurality of (
(for example, three) manual knobs and appropriate logic circuitry to provide parallel bit BCD output signals representing the various percentage settings of the knobs. The sign of the compensation is entered into a mechanism 34 and converted by the mechanism 34 to one of two different voltage output levels representing positive and negative compensation, respectively. The operation of placing in the beam a standard with a composite thickness that complements the alloy in the metal being measured involves the use of a pair of pulse counters from a common pulse source, such as a clock pulse oscillator, to measure the nominal and apparent thicknesses of the alloy. This is done automatically by sending at a relative speed that is a function of the desired ratio.

Once a certain number of counts is reached, one of the two counters (referred to herein as the Xa counter) will contain a count corresponding to the apparent thickness factor of said ratio. This counter provides an output for properly energizing the standards magazine to insert standards into the beam. This standard will attenuate the beam by an amount equal to the nominal thickness of the alloy being measured. For explanation,
This thickness measuring device measures approximately 0.254 cm (0.1
00 inches) is expressed as 1000 counts by three W-adic counters (hereinafter referred to as X-counters), and further assume that no alloy compensation is required for the standard pieces.

In this case, the X count and Xa counter count pulses from the oscillator at equal rates. Approximately 0.254 (0.254) corresponds to the digital voltage form as a basis for comparison.
Using a value of 100 inches) in the form of a corresponding digital voltage, 10 (4) pulses will be counted by the X counter and the same number of pulses will be counted by the Xa counter. When the Xa counter counts 1000, the standard strip magazine is activated and the total thickness equal to the nominal thickness of the object to be measured, in this case the apparent thickness, i.e. approximately 0.
254C77! (0.100 inch) is activated to insert a selected standard piece into the beam path having a total thickness equal to (0.100 inch). To further illustrate, assuming that +50% compensation is required for the apparent thickness of the standard to appear equal to the nominal thickness of the alloy being measured, the pulse applied to the The feed rate is 3
reduced by 1/2. Also, about 0.254Crf1 (
0.100 inch) thickness is 100 on the X counter
Assuming that it is represented by a 0 count, a 50% increase in pulses, or 1500 pulses, is applied to the Xa counter when the X counter registers 1000 counts. The standard thickness of 1500 counts in the Xa counter is approximately 0.381 cm (A). 150 inches), and the +50% compensation is done by causing the Xa counter to send K voltage signals equal to 1500 counts to the standard strip magazine. Then the magazine solenoid will operate and the rotation will be approximately 0.381c7rt (0.1
50 inches) thick magazine into the beam path. As yet another example, if we were to compensate by +100%, the pulse repetition rate from the oscillator applied to the x counter would be halved, so that the x counter would be 1
When the Xa counter counts 2000, the Xa counter counts 2000.

2000 counts in the Xa counter is approximately 0.508C77 in metal thickness! (stomach). 200 inches),
This thickness value is set in the standard piece magazine as before by the Xa counter. If negative compensation is required,
The X counter counts up to 1000 counts faster than the Xa counter at a faster speed proportional to the amount of compensation.

Therefore, the Xa counter counts a smaller number of pulses proportional to the amount of compensation and therefore a smaller percentage of pulses proportional to the amount of compensation than the X counter, and the Xa counter counts a final count proportional to the apparent thickness. give to The following is a description of a circuit for automatically inserting a standard into the beam path that provides the desired percentage compensation.

An oscillator 36 shown in FIG. 2 operates continuously to generate pulses of some suitable frequency, for example 100 MHz, and provides these pulses to the parallel inputs of two counters 38 and 40.

Counter 40 operates as a frequency divider. Therefore, if this counter is composed of three interconnected w-adic counters and operates as a 100-to-1 frequency divider, the 10 MHz pulse applied from the oscillator 36 is MM
The frequency is divided into Hz pulses. The counter 38 is also the oscillator 36
Clock pulse input from and percentage compensation setting mechanism 32
■ Count the sword output signals from.

The counter 38 can be composed of three W-adic counters that can be set by printers and a 1/2 frequency divider circuit composed of JK flip-flops. In such a case, counter 3
When 8 is programmed for 100% compensation from compensation setting mechanism 32, this counter operates as a frequency divider that divides the pulse output from oscillator 36 by 1 in 200. According to the selected value of % compensation, the divisor of counter 38 is 1.
It can be programmed to take a value between 000 and 2000. In practice, counter 38 divides the input pulse repetition rate by a factor of 1000+N. Here N has a value of 10 in absolute value for 1% of the required compensation. Thus, for example, if 100% compensation is required, N would be equal to 1000 and counter 38 would divide the clock pulse repetition frequency by 1/2000. On the other hand, if zero percent compensation is required, N is set equal to zero, and the counter 38 operates as a frequency divider that divides the frequency into 1 by 100. Sign setting mechanism 34 provides an O bit or one bit voltage level that determines the desired correction coefficient polarity.

Also included as part of the sign setting mechanism 34 is logic circuitry designated B. This logic circuit includes: (1) In response to positive or zero percent compensation in sign setting mechanism 34, output pulses from counter 38 are routed through a two-input NAND gate 42 to match circuit 4;
4 (X counter) and a direct circuit coupled to counter 40
(2) the output pulses of gates 42, 46 in response to negative % compensation in sign setting mechanism 34; counter 4 through each
Switch the direct circuit of the pulses leading to 4,48. In this way, the compensation setting mechanism 32 and the sign setting mechanism 34
Used to select the division ratio at 8, the counter 38
and 40 outputs are dialed in to positive or negative % compensation values, thereby conditioning the thickness measuring device for calibration of specific alloy thickness measurements. The counter 44 operates as a synchronous circuit and is programmed by the parallel hit BCD thickness signal from the thickness setting mechanism 30 until it reaches a Pascal count numerically equal to the thickness value. The received pulses are counted and when they are reached a match signal is generated and applied via lead 49 to controller 50. This match signal resets the X-count controller 54 (FIG. 3). , this controller causes an inhibit signal to be generated at terminal 55, thereby terminating any further pulses being applied to counter 44. This inhibit signal is connected to an OR gate 62 connected to terminal 55; It also appears at the second input of gate 46 and inhibits further pulses from being applied to Xa counter 48.
Therefore, both the X counter and the Xa counter are simultaneously prohibited from counting any further. The inhibit signal output of circuit 54 also triggers high voltage regulation controller 56 (FIG. 3) to initiate the next calibration mode, described below. X
The a counter 48 includes four cascaded decimal counters and counts from 0000 to 9999.

The output of this counter is B which corresponds to the recorded count.
With the CD signal Xa, this counter represents the apparent thickness of the object for which the thickness measuring device is to be calibrated. Xa
The output of the counter is coupled to a solenoid of the magazine 12 via a buffer register 61. This register has a load line coupled to terminal 55. The resistor 61 transfers the Km voltage output of the Xa counter to the solenoid when the X-count controller 54 is reset so as to supply current to the solenoid that inserts the corresponding one of the standards into the beam. The change in signal level that occurs at terminal 55 when controller 54 is reset is used immediately or after a suitable delay to generate a strobe signal.
This strobe signal is applied to register 61 by transfer line 63, causing the transfer of the Xa count to the magazine 12 solenoid. Transfer line 63 is connected to terminal 55 via a suitable delay circuit (not shown).

This delay circuit is responsive to the reset output signal of controller 54, but has a time constant sufficient to allow the Xa counter to stabilize before a strobe pulse is generated at the output of the delay circuit. This delay circuit can be composed of, for example, a monostable multivibrator. This monostable multivibrator is triggered into an unstable state by a trigger voltage applied to terminal 55, and when it returns to a stable state, gives a steep strobe pulse to transfer line 63. When a suitable standard is inserted into the beam path, the intensity of the radiation emanating from the inserted standard is detected by a detector 14 and its output signal is linearized by an amplifier 16. . The output of amplifier 16 is applied to an analog indicator via variable gain amplifier 18 to bias the pointer of the indicator to a position on the scale. When a standard is alloy compensated, the deviation indication on the analog indicator must be zeroed (adjusted) with respect to the newly selected standard; if this is not done, the indicator will be zeroed relative to the previously selected standard. Maintain zero point alignment.

Therefore, how automatic zero point adjustment is performed in the thickness measuring device of the present invention will be explained below. Zero point adjustment: The controller 50 shown in FIG. 3 is composed of five flip-flops connected in series.

One of these flip-flops forms part of the start calibration controller 52. This controller is reset by a stepped voltage provided by operation of a pushbutton start switch (not shown). This start calibration controller 5
Calibration flip-flop 2 is set when the pushbutton switch is pressed and produces an output signal. This output signal sets the X-counting flip-flop of the x-counting controller 54. When the X-counting flip-flop is set, it produces a voltage signal at terminal 55. This signal is connected to gates 42 and 4.
6 open at the same time. Since gates 42 and 46 are open,
As described above, the X counter counts at the same time as the Xa counter.

Due to the simultaneous counting in the x-counter, the x-count controller 54 (FIG. 3) is reset by a pulse applied from the x-counter via conductor 49. x counting controller 5
When 4 is reset, the voltage applied to terminal 55 changes level enough to close gates 42 and 46, thereby preventing X counter 44 and Xa counter 48 from being pulsed. Shortly after the X-count controller is reset, an Xa counter equal to K voltages is sent to the standard strip magazine solenoid by operation of register 61. Resetting controller 54 sets the flip-flop of high voltage regulation controller 56 (FIG. 3).

High voltage adjustment controller 56 adjusts the kilovolt level of the power supply to produce a beam of sufficient strength to zero indicator 20 when a new standard is inserted into the beam path. To generate a beam of such intensity, the power supply voltage is repeatedly adjusted as described below. The state in which the pointer of the indicator indicates the zero point corresponds to the state in which the voltage difference between the input terminals of the indicator is zero, so this voltage difference remains zero even with the thickness of the new standard piece inserted into the beam. There is no such thing as doing so. For example, if the thickness of the newly inserted standard is greater than the thickness of the previously inserted standard, the intensity of the radiation is insufficient to penetrate the object to be measured, and therefore: Indicator 20
The voltage applied to one terminal of j is not zero, but has a magnitude and polarity corresponding to its strength. The voltage at this terminal is determined by a voltage-pulse count type analog-to-digital converter 6.
It is converted to a number representing the pulse count by 4 and added to a reversible counter 66. The number of counts applied to counter 66 is proportional to the absolute value of the analog voltage across the indicator input terminal. Converter 64 also indicates the polarity of this voltage, and this information is also provided to counter 66 to control the counter to count forward or backward according to its polarity. In the present example, counter 66 is configured to count sequentially from some predetermined initial count. The output of counter 66 is transferred to zero adjustment storage register 68 following each sampling period.

This register provides an output to digital-to-analog converter 70. This converter converts the digital output of the register 68 into an analog signal of the same magnitude, and feeds this analog signal back to the voltage regulator 11 to adjust the power supply voltage.
In the example just described, the feedback signal increases the voltage on power supply 10 to increase the intensity of the radiation beam. Further, this feedback signal is sent to the detector 1 via a feedback path 71A indicated by a broken line.
In addition to 4, the gain of the detector 14 can be changed appropriately to zero the indicator. Because the servo action of the feedback loop 71 and its associated components drives the voltage at the input side of the converter 64 towards zero volts, and as the servo action continues said voltage passes through zero volts and reverses polarity. , the voltage is adjusted in the opposite direction again.

The polarity change of the analog input signal applied to the converter 64 is
The counter 66 is controlled to perform a reverse counting operation according to a change in the polarity output indication of the converter sent to the counter 66. As counter 66 counts back, the count stored in register 68 is decremented, thereby reducing the number of digital signals applied to converter 70. As a result, converter 70 reduces the magnitude of the analog feedback signal voltage applied to voltage regulator 11, reducing the intensity of the radiation beam. This servo action continues until the magnitude of the analog signal sampled by transducer 64 is approximately zero. After the converter 64 has sampled the analog voltage at the input of the indicator 20, provided that the error between the samples does not exceed a certain value, e.g. 0.01%, the verified zero adjustment signal is determined by the verification circuit. The voltage is then sent to the voltage adjustment controller 56. This signal resets controller 56, thereby terminating further zero adjustments. The confirmation circuit that operates like this has two counters. We have evenings 80 and 82.

These counters provide a signal on lead 83 indicating that sufficient power supply voltage adjustment has been made and that indicator 20 is zeroed to the standard in the beam. The following is a description of an example of a circuit device for automatically zeroing an indicator/meter with respect to a newly inserted standard piece.

This device includes the analog θ digital converter 64, a reversible counter 66, a high voltage zero point adjustment storage register 68,
A digital-to-analog converter 70 and a feedback loop 71 are provided which connect the output of the converter 70 to the control input of the regulating device 11 or to the detector 14 . Converter 64 is a conventional bipolar analog-to-digital converter of the conventional voltage-to-pulse-counting type that samples the analog signal applied to the input terminal of indicator 20 and converts the magnitude of the sampled analog signal. A series of output pulses having a proportional number of pulses is applied to lead 72, and a leading pulse of 0 or 1 bit is applied to indicate the polarity of the analog signal, ie, whether it is a positive or negative voltage with respect to ground potential.

A count transfer signal is generated by the transmission at the end of each conversion cycle, and this signal is applied via conductor 74 to one input of a two-input AND gate 76,9.6. A count output representing the magnitude of the analog voltage sampled by transmission 64 is provided to counter 66 via AND gate 78 which is opened by conductor 72.

AND gates 76 and 78 are opened by voltage from high voltage regulator 56 at the beginning of this operating sequence. This voltage appears on conductor 57 connected to the inputs of gates 76 and 78. At the end of each conversion cycle, a count transfer signal is applied to the input of register 68 through open gate 76, causing parallel transfer of the digital count accumulated by counter 66 to register 68. The output from register 68 is applied to digital-to-analog converter 70. This converter has a size of register 68
generates an analog output signal proportional to the count within.
The output of converter 70 is applied via feedback loop 71 to voltage regulator 11 to proportionally control the voltage level of radiation source 10 and proportionally control the intensity of the radiation beam emitted from radiation source 10. The voltage regulator 11 includes a DC power source that can programmably control the high voltage supplied to the radiation source 10 according to analog voltage changes in the feedback loop 71.

Voltage regulator 11 operates to maintain the voltage of radiation source 10 in a constant relationship to a given reference voltage input. Thus, in response to the analog output level from the digital-to-analog converter 10, the radiation intensity level of the radiation source 10 changes to increase or decrease the input to the detector 14, causing the pointer of the analog indicator to zero. to point to. Alternatively, as described above, the analog signal from converter 70 can be applied to detector 14 via loops 71 and 71A to vary the gain of detector 14 so that indicator 20 indicates a zero reading. At the end of each conversion cycle of analog-to-digital converter 64, the zeroing operation is repeated until indicator 20 indicates zero.

To detect zero, the pulse count appearing on lead 72 and the count transfer signal appearing on lead 74 are applied to w-adic input counters 80 and 82, respectively. The counter 80 is a reversible counter and controls the counter 82 by resetting the counter 82 every time an overflow count state occurs. Counter 80 performs forward or reverse counting in parallel with counter 66 in a direction determined by the polarity signal at the beginning of each pulse count applied to conductor 74 by converter 64, and is determined by its counting capacity. overflowing at a rate of If the polarity signal is either 0 bit representing positive polarity or 1 bit representing negative polarity, the counter 80 counts forward if there is 1 bit at the beginning of the added pulse, and counts backward if there is 0 bit. . As the pointer of the indicator swings in either direction from zero, the count provided to counter 80 may be sufficient to generate an overflow pulse output that is used to reset counter 82. As the indicator 20 approaches the zero reading, or as the pointer oscillates at the midpoint of the scale, the counts added to or subtracted from the counter 80 will cause an overflow within a given time interval. is sufficient so that counter 82 counts the transferred signals from analog-to-digital converter 64 without being disturbed by the reset signal from counter 80. As previously stated, the number of counts transferred by transducer 64 is proportional to the absolute value of the voltage appearing at the input terminal of indicator 20 to which transducer 64 is connected.

Since it is desirable to put the indicator 20 in the zero point indicating state,
The greater the number of counts transferred at the end of a given shift cycle, the greater the deflection of the pointer of indicator 20 from its desired zero point. Therefore, a zero count of the transferred count represents a complete zero reading of the indicator, and a full scale count represents a large unilateral deflection of the indicator pointer from zero. As the count output of the transmission 64 converges to zero, the non-periodic nature of the count transfer pulses generated by the transmission 64 becomes less, so that the count transfer signal is generated periodically when a complete zero point indication is performed. .

Counter 82 counts successive transfer signals from transmission 64 while counter 80 counts the polarity and number of counts generated by the transmission between two consecutive transfer signals. By appropriately configuring 80 and 82, the counter 82 will not run until the output count of the transmission 64 is zero or very close to zero.
Counter 80 is configured to reset counter 82 before it reaches its full counting capacity, thereby providing the desired calibration accuracy. When the output count of the transmission 64 reaches zero or a value very close to zero, the counter 8
Counter 82 can count a predetermined number of consecutively transferred signals without being temporarily reset by a zero overflow signal. Once this predetermined number of transferred signals has been counted, counter 82 overflows and provides a signal on line 83. This signal ends the zero check adjustment sequence. In a typical embodiment, counter 80 comprises a single programmable w-adic counter that is programmed to overflow when it receives a total of +20 counts.

Counter 82 includes two programmable one-pheast counters connected in series and overflows when receiving consecutive transfer signals. If counter 82 overflows without being reset by counter 80, counter 82 generates a signal on conductor 83. This signal resets the high voltage regulation controller 56 to its initial state. In this state, controller 56 generates a voltage on conductor 57.
This voltage closes gates 76 and 78 to stop further operation of the zeroing sequence and sends a trigger pulse to percent Xa thickness controller 58. This pulse sets the flip-flop of its controller for the reasons described below. Since gates 76 and 78 are closed, reversible counter 66 does not receive the count output from transmission 64, and therefore reversible counter 66 and zero adjustment storage register 68 are maintained as they are.

In other words, the count value is fixed. This fixed count value is added to the analog signal by a digital-to-analog transmission 70 and further applied via a feedback loop 71 to the radiation source voltage regulator 11 or to the detector so that the radiation source voltage or detector gain is adjusted accordingly. Fixed to the value corresponding to the analog signal. After checking the accuracy of the zero adjustments made during the zero adjustment sequence and adjusting the radiation source 10, i.e. fixing the gain of the detector 14, the deviation of the thickness of the workpiece from the nominal thickness is determined. It is necessary to calibrate, or range, the indicator 20 to cover the predicted range.

Range adjustment: The automatic range adjustment of the present invention will be explained below.

The maximum range is typically the full scale range of indicator 20. If this calibration is not performed, the deflection of the pointer of the indicator 20 can be compared to the actual thickness deviation of the workpiece because the thickness measuring device has not been previously calibrated to the desired measurement range for the new nominal thickness. It is actually impossible to correspond to the value of . This calibration operation is called a J range adjustment operation because it adjusts the range of the indicator. In particular, range adjustment is accomplished by adjusting the gain of a variable gain device or amplifier 18 in the detection circuit. In summary, this range adjustment uses a counter that counts pulses from the oscillator 36 at a pulse repetition rate that corresponds to the maximum normal deviation of the object from the nominal thickness. 9
This is done by controlling the X counter 44 and the Xa counter 48 by zero.

At the end of the nulling operation, the source 10 voltage or detector gain is fixed, so the Xa counter 48 adds a count representing this preselected maximum deviation. It will be done. This deviation represents the percentage deviation from the nominal value. When the Xa counter 48 is updated by this count value, the standard piece magazine is updated by the output of the Xa counter 48.
and insert another standard piece into the beam path corresponding to the selected percentage deviation in thickness. The operation of certain circuits previously used in the zeroing operation, as well as other circuits, apply an analog voltage to the ranging feedback loop 99 (FIG. 1). This voltage is applied to amplifier 18 and
The gain of the amplifier 18 is changed by the direction and magnitude that cause the pointer of the indicator 20 to swing by a certain amount in a certain direction from the zero point. This gain change can be made by changing the variable resistance in the feedback loop of amplifier 18. Typically, the amount of deflection is a full scale deflection of the pointer of the indicator in one direction from zero. Confirmation of the sensitivity of the indicator to thickness changes is provided by an analog-to-digital converter 64 operating in conjunction with counters 80, 82 in a manner similar to that described above for high voltage regulation operations. For example, once the indicator 20 has been calibrated to positive full scale, it is relatively easy to calibrate it to negative full scale. The reason for this is that in indicators with linear response characteristics, there is symmetry about the midpoint of the scale. The circuit that automatically performs the range adjustment operation will be explained in detail below.

During the zero adjustment operation in which the Xa counter is followed by the K1 value of the apparent thickness of the object to be measured, the pulse passing through gate 46 is also connected to the input of percent Xa thickness divider 84 and one input of NAND gate 86. added to.

The divider 84 has two w-adic counters. These counters are programmable in advance of the thickness deviation range from the nominal thickness and provide a full scale deflection of the indicator 20 over this range. Divider 84 is programmable by percentage Xa setting mechanism 85.
The mechanism includes a plurality of switches that can be selectively closed to set the division factor of divider 84 to any desired percentage. Alternatively, a divider can be used without programming to divide the count input by a constant factor. For example, if a deviation of 10% is desired, the divider 84 will divide by 1 for every 10 counts received.
output pulses, thereby providing a count output of 10% of the pulse count applied to Xa counter 48. This percentage pulse count is combined via a NAND gate 88.

Gate 88 is set by x-count controller 54 during the x-count operation. These pulses are applied to the forward count input of counter 90 through gate 88, which is open. The counter 90 is composed of a plurality of W-adic counters connected in series, and has an appropriate count capacity, for example, 99
Give 9. Therefore, during the counting operation, the counter 9
The zeros are filled to a count equal to a programmed percentage of the count in Xa counter 48 (eg, 10%). When the zero point adjustment operation is completed, the percentage Xa thickness controller 5
8 is triggered to the set state and controller 58 generates a gate open signal.

This signal is transmitted from the terminal 91 to the NAND gate 86 which is opened.
through gate 62 and gate 46 .
is added to open gate 46. The pulse count is then applied through gate 86 to the counter-count input of counter 90 and directly to Xa counter 48. Detector 92 is responsive to the count output of counter 90 and generates a signal output when counter 90 registers all zeros. This signal is applied via lead 93 to the percent Xa thickness controller 58 to reset the controller. As a result, this controller generates a signal on conductor 91,
This signal causes gates 46 and 86 to close, inhibiting further pulses from being applied to Xa counter 48 and counter 90. This action causes the programmed percentage of the initial Xa count provided by divider 84 to change the previous count in Xa counter 48 by that percentage. Terminal 91, like terminal 55, is also connected to transfer line 63 of register 61 via the delay circuit, so that when controller 58 is reset, the signal sent by terminal 91 is also updated in the form of BCD. It is also used to start the transfer of the Xa pulse output to the solenoid of the standard piece magazine. Since one of the solenoids is actuated for that purpose, the K] value of the total number of standards moved during the beam path is equal to the new BCD count value in the Xa counter 48. Therefore, the new deflection of the indicator pointer corresponds to a 10% change in the original thickness value programmed by the thickness setting mechanism 30 and the compensation setting mechanism 32. In the subsequent gain adjustment operation, the gain of the amplifier 18 is changed to swing the indicator 20 to full scale for the assumed 10% thickness deviation range. Percent Xa
When the thickness controller 58 is reset, the gain adjustment controller 60
is set, the operation of the range adjustment feedback loop 99 is started, and as a result, the range adjustment operation is started. As previously mentioned, the voltage across the analog indicator is sampled by analog-to-digital converter 64 and converted to a digital count proportional to its magnitude. In the adjustment operation just described, the digital output of transmission 64 is applied to digital subtraction circuit 94. The output of this circuit is connected to the input of counter 66. Subtraction circuit 94 has four reversible binary counters connected in series.

These counters can be reset to an initial starting count of 1000 by a reset signal applied to lead 95 by the gain adjustment controller. This reset signal is also applied to one input of AND gate 96 to open this gate. The reset subtraction circuit 94 includes the transmission 64
to provide a digital output representing the difference in counts between the actually measured count output shown on the indicator 20 and the full scale count output of the transmission corresponding to the full scale indication of the indicator 20. Operate. For example, in response to an analog voltage magnitude corresponding to a full scale deflection of the pointer of indicator 20, transmission 64
Assume that the configuration is configured to give an output of 00 counts. And actually 80 which corresponds to 0.8 of full scale indication
Assume that a 0 count is generated. In such a case, a value of 1000 counts corresponding to the full scale is given to the circuit 94 by a circuit not shown, and 800 counts corresponding to the actual deflection of the indicator 20 are subtracted from this, so these values are The difference between the two counts, i.e. 2
Generates a pulse output corresponding to 00 counts. These 200 counts are added serially to counter 66 during the analog-to-digital conversion cycle. At the end of each shift cycle, the count transfer signal from the transmission 64 is applied to the range adjustment storage register 98 through an open AND gate 96, activating this register.
This register is included within gain adjustment feedback loop 99.
As a result, this register receives and stores the final count transferred from counter 66. The contents of this register are applied to digital to analog converter 100. This converter converts the digital count in register 98 to an analog signal whose magnitude is proportional to the digital count. This analog signal is fed back to amplifier 18 to change the gain of amplifier 18, thereby changing the voltage output of amplifier 18 by the direction and amount that causes the pointer of indicator 20 to swing to full scale. In the example described here,
The gain of amplifier 18 is increased by a value corresponding to the 200 counts produced by converter 64. For that purpose converter 6
The magnitude of the analog signal applied to the input of 4 increases by an amount proportional to the gain increase. Thus, in the next conversion cycle, the count output of the converter 64 will reflect a stepwise increase in the magnitude of the analog input signal, and the count output of the converter 64 will reflect a stepwise increase in the magnitude of the analog input signal applied to the input side of the converter 64, i.e., across the input terminals of the indicator 20. Generates an output pulse count of approximately 1000 counts corresponding to a full scale value. The pulse output from the subtraction circuit 94 is sent to the reversible pulse counter 80 of the zero point detection circuit.
It can also be added to The zero point detection circuit operates as described below to detect the full scale indication condition of the indicator. During gain adjustment, analog indicator deflection causes circuit 94 to output counts in excess of the number required to overflow counter 80 (20 counts in the previous example);
Counter 82 is thereby reset. Indicator 2
0 approaches full scale deflection, or indicator 2
If the zero reading were to oscillate about the full scale reading point, the output of circuit 94 and thus the count added to or subtracted from the count of counter 80 would cause an overflow condition. is insufficient, and the counter 82 operates to count the transfer pulse signals. The operation for confirming range adjustment is the same as that described for confirming zero point adjustment. After a predetermined number of adjustment cycles without a reset signal, an overflow signal from counter 82 is sent to gain adjustment controller 60 via lead 97. This signal resets controller 60 so that the controller generates a voltage level on conductor 95 that resets subtraction circuit 94 to close gate 96, thereby initiating the ranging operation. Let it end. At this time, the start calibration controller 52 is also reset. Since the subtraction circuit 94 has been reset and the gate 96 has been closed, the pulse output from the subtraction circuit 94 is not input to the reversible counter 66, and its count integrated value is not transferred to the range adjustment storage register 98. Therefore, reversible counter 66 and range adjustment storage register 98 remain unchanged. In other words, the count value is fixed. This fixed count value is converted into an analog signal by the digital-to-analog converter 100 and is further applied to the variable gain amplifier via the feedback loop 99, whose gain is fixed at a value corresponding to the analog signal. When the range adjustment operation is completed, the ground voltage signal from the calibration controller 52 that has just been reset is applied to the conductor 10.
3 to all solenoids of the standard magazine 12.

This voltage signal returns all solenoids to their initial state, thereby removing all thickness standards from the radiation beam path. When a plate-shaped material is placed on the thickness measuring device so as to block the radiation beam, the radiation attenuation due to the plate-shaped material is detected by the detector 14 . amplifier 16,
The analog output signal is processed by 18 to provide a driving voltage to indicator 20. To calibrate the thickness measuring device for workpieces of different nominal thicknesses, the compensation setter 32 and the thickness setter 30 are programmed with percentage compensation and the desired thickness value.

When the start pushbutton (not shown) is pressed again, a voltage is developed on line 105, which sets the start calibration controller 52. Set controller 52 generates a delayed reset signal which is applied to X counter 44 and Xa counter 48 via lead 104. A trigger signal is sent to also set the X-count controller 54, initiating X-count operations and other calibration operations. In the above example,
In addition to the range adjustment operation, other calibrations such as alloy compensation and zero point adjustment are also performed automatically as a series of calibration operations, further improving the speed and accuracy of the calibration work.

In addition, the range adjustment confirmation circuit of the above embodiment uses two counters and performs range adjustment only when the integrated value of the subtraction circuit remains almost zero even after performing the conversion operation by the analog to digital converter 64 a predetermined number of times. Configure it to stop working.

This configuration is preferable because achievement of range adjustment can be confirmed with high accuracy. However, in order to simplify the configuration, only the count output of the subtraction circuit is counted, and when the integrated value is within a predetermined range, for example, 0 to 20 counts,
It can also be configured to stop the range adjustment operation.

In addition, the number of conversion cycles is counted by counting the transfer signal from the analog-to-digital converter 64, and when this count reaches a preset value, for example, 40 counts, the integrated value of the subtraction circuit is calculated. It is also possible to configure the range adjustment operation to be stopped when it is assumed that the temperature has reached almost zero.

[Brief explanation of drawings]

1 and 2 are block diagrams of an embodiment of a thickness measuring device constructed according to the present invention, and FIG. 3 is a block diagram of a control circuit that controls the circuits shown in FIGS. 1 and 2. 10... Radiation source, 12... Standard piece magazine, 14... Radiation detector, 18...
... Variable gain amplifier, 20 ... Analog indicator, 32 ... Compensation setting mechanism, 34 ... Compensation sign mechanism, 56 ... High voltage regulator , 68...
...Zero point adjustment memory register, 98... Range adjustment memory register.

Claims (1)

[Claims] 1 a radiation source 10, a detector 14 for detecting radiation transmitted along a radiation path and indicating a deviation in the thickness of a workpiece in the radiation path in response to the level of radiation emanating from the radiation source; a standard piece magazine 12 that sets a plurality of calibration standard pieces corresponding to the thickness in the radiation path; and an analog indicator 20 that indicates the deviation of the thickness of the object to be measured from the set thickness of the standard piece. , connected between the detector 14 and the analog indicator 20 and the detector 14
a variable gain amplifier 18 for varying the output of the analog indicator; and a zero point for controlling the radiation source in a calibration mode for setting a first set of standards in the radiation path to align the actual run of the analog indicator with the zero point position. and a range adjustment means for controlling the variable gain amplifier to adjust the range of deflection of the pointer of the analog indicator with respect to the thickness deviation of the object to be measured, wherein the range adjustment means comprises: a range setting mechanism 85 for selecting a range of deflection of the pointer of the analog indicator as a percentage thickness of the thickness of the first set of standards; and a range setting mechanism 85 for generating a digital signal representing the selected range; a control circuit 84, 88, 90, 92, 58 for placing a second set of standards in said radiation path having a thickness corresponding to said digital signal; an analog-to-digital converter 64 for converting a first analog signal representing the difference between the zero point and the zero point position into a digital output signal representing the absolute value and polarity of the analog signal; and a digital set point representing the full scale of the analog indicator. a digital signal generating circuit 94 that generates an output signal representing a value obtained by increasing or decreasing the initial value by the difference between the digital output value and the digital output value of the analog-to-digital converter;
66, 98 and the digital signal generation circuits 94, 66,
a digital-to-analog converter 100 for converting the output of 98 into a corresponding second analog signal; and a feedback for providing the second analog signal to the variable gain amplifier 18, thereby controlling the gain of the output of the detector 14. A loop 99 counts the difference between the digital setting value representing the full scale of the analog indicator and the digital output value of the analog-to-digital converter, and when the count value falls within a predetermined range, a range adjustment confirmation signal is generated. and a range adjustment confirmation circuit 80 for generating and thereby setting the second analog signal to a value corresponding to the gain of the variable gain amplifier 18.