WO1994018637A1

WO1994018637A1 - Method of manufacturing utilizing master test result feedback to slave test stations

Info

Publication number: WO1994018637A1
Application number: PCT/US1993/009676
Authority: WO
Inventors: James Marvin Harris
Original assignee: Gardiner Communications Corporation
Priority date: 1993-02-11
Filing date: 1993-10-12
Publication date: 1994-08-18
Also published as: AU5537094A

Abstract

The present method of manufacturing provides improved accuracy inthe measurement and tuning of select parameters of electronic devices. The electronic components are assembled (10) on an assemblyline and tested by test stations (18, 22, 24, 26) wherein each test station is monitored by a control system (34). The method compares (108) slave station test performance to a master test station with feedback of the differences to the slave test station, thereby forcing it to emulate the master test station in future testing. The method further involves the monitoring (34) of assembled device quality. If the quality deteriorates, the control system can automatically change component selection (40) or contact a human operator (38). The control system automatically records the entire assembly line process to provide inventory control and an assembly history of each device for repair and assembly line process modification. Additionnaly, the control system monitors the entire assembly line for analysis and constraint (36) of preselected parameters such as production rate and product quality.

Description

METHOD OF MANUFACTURING UTILIZING MASTER TEST RESULT FEEDBACK TO SLAVE TEST STATIONS

TECHNICAL FIELD OF THE INVENTION

The present invention relates to a method of manufacturing utilizing master test station feedback to slave test stations. In particular, specific parameters of a reference unit are measured by both a master test station and by a slave test station. These results are compared and the slave test station is forced to emulate the master test station.

BACKGROUND OF THE INVENTION

Computerized control systems can be useful in controlling the quality of assembled electrical devices. Electronic devices, particularly those involving high frequency signals, can vary widely in their responses to a fixed input, and can require

tuning during the assembly process. For example, a low noise block downconverter

(LNB) converts a high frequency, low amplitude signal from a satellite into a usable signal for transmission through a coaxial cable using a low noise amplifier and a block downconverter. The performance of both the downconverter and the low noise amplifier can be affected by the physics of their circuit layouts. During the assembly of the LNB, both the low noise amplifier and the downconverter can be tuned. However, to properly tune these constituent circuits, accurate measurements must be made of their untuned responses. Therefore, assembly and tuning can depend upon accurate and consistent measurements by test equipment.

Testing of circuit response is carried out at a test station using specialized test equipment coupled to a computer. The test equipment, or testers, can measure any number of device response parameters such as circuit gain or noise. The computer, in turn, can process, store, and even display the readings of the test equipment. A problem occurs when, in a large production operation, multiple test stations are used to test a certain parameter. Each test station is equipped with unique test equipment which could provide significantly different test results from the same tested device. In other words, two seemingly identical testers can produce different test results from the same device, even if the testers are the same brand and model. These variations can occur due to variations in the components in each tester, or due to the wear

caused by use. The testing of several parameters, each on one of several testers complicates the matter further. The error in readings from each sequential tester can add, causing a greater deviation from desired performance for the assembled device. Also, in the assembly and tuning of certain electronic devices, the tuning of one constituent element of the device can optimize one parameter reading while diminishing the results of another. Therefore, the overall quality of the electronic device becomes difficult to control when several parameters are being tuned at several test stations.

An example of automated manufacturing using a computerized control system

is disclosed in U.S. Patent No. 4,827,423 to Beasley et al. entitled "Computer

Integrated Manufacturing System". The system includes several levels of computers to organize and disseminate the information for controlling shop floor level systems. Scheduling data and data relating to process, product, and material specifications, as well as bills of material, are generated in an upper level computer system and refined and down loaded as needed to lower level computers controlling the shop floor processes. The computers on the upper levels are capable of communicating with the computers on the lower levels and computers on the same level are capable of communication with each other as needed to pass information back and forth.

U.S. Patent No. 4,870,590 to Kawata et al. discloses a "Manufacturing Line Control System". In this control system, the hardware region of a main memory is divided on the basis of different pieces of information to be stored. A line computer synthetically judges and processes data transmitted from a manufacturing line. Various analog signals, digital signals and status signals issued from a product passing through the manufacturing line are input from the line computer into various equipment by a line controller including an input-output portion with which a plurality of card equipment is detachably provided for application to various types of manufacturing lines.

The goal of the control system for the assembly of electronic devices is to

ensure uniformity of product, recordation and optimization of the device's response characteristics, and maximization of product output. However, neither Beasley et al.

('423) nor Kawata et al. ('590) address the use of computers to compensate for the inaccuracies of test equipment used in a manufacturing process. Neither do they address the interactive dependencies of various tuning processes.

Therefore, a need exists for a method of manufacturing wherein each slave test station is forced to emulate a master test station, i.e. static calibration. Second, the method of manufacturing should allow for dynamic interaction between a control system, the computer test stations, and the human operators. In other words, there is a need for deficiencies in downline products to be analyzed and corrections fed to an upstream operator to eliminate the deficiencies and variations, i.e. dynamic calibration. There is also a need for a manufacturing system which utilizes an overlying constraint system so that, if the response characteristics are poor, the constraint system can select different constituent components for the electronic device. If the rate of assembly falls below a predetermined limit, the constraint system can automatically contact a human manager. Third, there is a need for a manufacturing system which can be used to record the response characteristics of devices. This would allow for inventory control as well as detailed performance reports for customers. SUMMARY OF THE INVENTION

The method of manufacturing of the present invention involves the assembly and tuning of electronic devices, wherein operators and computers work together to

assemble constituent elements and test subassemblies of the device as well as test the performance of the assembled device. The system provides: (1) a comparison of slave station test performance to a master test station with feedback of the differences to the slave test station, thereby forcing it to emulate the master test station, i.e. static calibration; (2) monitoring of downstream parameter values with immediate feedback upstream to effect changes in component selection, assembly, or tuning, as well as

constraint of deviant fluctuations, i.e. dynamic calibration; and (3) automatic recording of the entire assembly line process to provide inventory control and an assembly history of each device.

A first aspect of the invention involves a comparison of slave test station performance to a master test station. The slave test station can be either at least one first parameter tester, at least one second parameter tester, or at least one additional parameter tester. A reference unit is tested on the master test station for a first parameter such as circuit gain over a range of frequencies. The performance of the master test station is known, optimized and consistent. Likewise, the reference unit provides consistent output for a given input. Thus, confidence in the results of the master test unit is high.

Next, the same reference unit is taken to a first slave test unit for testing of the same parameter. The results of these tests are stored and compared to the results of the master tester. Any differences are noted and stored in a "delta" file. This file is stored in the slave tester's memory. The readings of that slave tester for a production unit are skewed by the value in the delta file. Thus, that slave test station is forced to emulate the accuracy of a master test station. This process is repeated for each slave test station. Therefore, the system interacts with its human operators

in such a way that deficiencies in downline products are analyzed and solutions are fed to the upstream test stations to correct the deficiency. The solution is presented to the operator in such a way that it does not disrupt his performance. He continues to tune the constituent circuit or subassembly, while the computer attached to the test equipment adjusts the values measured by the test equipment and displays the adjusted values. This process is also referred to as "static calibration" because the calibration is typically performed once before production begins and its effect remains constant throughout the production run.

A second aspect involves the "dynamic calibration" of the method of manufacturing. Downstream parameter values are fed upstream continuously to effect changes in component selection, product assembly, and product tuning. For example, if a deficiency in downstream product is detected, the control system automatically feeds back changes to an appropriate test station. These changes are stored in a "dynamic skew" file. The values in either the dynamic skew file and/or the static delta file automatically alters the values displayed at the test station. A deficiency in downstream product can also result in a change in the initial assembly unit which selects components for the electronic circuitry. In a typical assembly line process, the initial device, a pick and place machine, is programmed to select a number of different electronic components to be placed on a printed circuit board. If the desired parameter changes require new components, the computer controlling the assembly line process feeds back information to the component selection device which modifies its selection of components for the printed circuit (PC) board.

This dynamic calibration of the assembly process can also take the form of an automatic "constraint" of the production line based upon readings from the individual test stations. Each test station monitors the progress of the device during the assembly process. While automated aspects of assembly, such as with the pick and place machine, are easily determined, any human labor required may vary greatly from device to device. Backups, or queues, can occur, leaving certain workers without work. The dynamic control system takes information from each test station

along the assembly line. This information can include the number of devices passing through a specific test station or the time spent between test stations. When these values vary from a predetermined range, the constraint system contacts a human manager on the assembly floor. The contact informs the manager of the nature of the constraint on the assembly line. A third aspect of this method of manufacturing involves the storage of the measured response characteristics of each device as well as assembly data. Each device going through the assembly process is coded, as with a bar code, so that there is a history for each assembled item stored in memory. This process has several advantages. First, a complete inventory is provided to generate reports and information about the assembly line output. Next, a history of each device is provided so that defective devices returned for repair can be analyzed more effectively and defects in the assembly line procedure can be quickly pin-pointed. Also, the specific performance of each assembled device can be printed for labeling purposes or reports of individual device performance can be included with the assembled device for delivery to the purchaser.

BRIEF DESCRIPTION OF THE DRAWINGS

For a more complete understanding of the present invention, and for further details and advantages thereof, reference is now made to the following Detailed Description taken in conjunction with the accompanying drawings, in which: FIGURE 1 is a flow diagram of the assembly process for a low noise block downconverter (LNB);

FIGURE 2 provides a flow diagram of the method of calibrating the test

stations for the manufacture of a LNB in accordance with the present invention;

FIGURE 3 is a flow diagram of a generic method of calibrating every slave

station in accordance with the present invention to emulate a master test station;

FIGURE 4 illustrates the dynamic calibration in the method of manufacturing according to the present invention; and

FIGURE 5 illustrates the constraint system in the method of manufacturing according to the present invention.

DETAILED DESCRIPTION OF THE DRAWINGS

The present invention relates to a method of manufacturing that overcomes

many of the disadvantages found in the prior art. One use for this manufacturing

system is in the assembly and tuning of low noise block downconverters (LNBs) for use with satellite receivers. However, the method can be utilized in the assembly of any device. The present method will be described by a series of flow charts which diagram the various steps involved.

FIGURE 1 illustrates the general assembly process 10 for a low noise block downconverter (LNB). Each LNB is comprised of a low noise amplifier (LNA) circuit and a block downconverter (BDC) circuit attached to a carrier board. The low noise amplifier increases the gain of the relatively weak satellite transmission. The downconverter circuit utilizes an oscillator to convert the relatively high frequency satellite signal into a stable, lower frequency signal capable of transmission down a standard coaxial cable. In the assembly of the LNB, both the block downconverter and the low noise amplifier circuit are assembled 12, each typically on a separate PC board. Due to the high frequency of the input signal from the satellite, the physics of the board arrangement has a significant impact on both the gain and the noise level of the output signal from either the block downconverter or the low noise amplifier. Both circuits are comprised of a plurality of constituent elements. These elements are placed on their respective PC board by a conventional pick and place machine using conventional assembly techniques. These elements are soldered into place, and the PC board cleaned. After assembly, the low noise amplifier is installed 14 on a carrier board.

Next, a static sensitive transistor can be installed 16 on the PC board which is attached to the carrier board/wave guide assembly. The LNA is then tested 18 at one

of at least one amplifier tuning stations ("amp tune stations"). The wave guide is attached to a test fixture and subjected to an input signal. This signal is guided to the

LNA which alters the signal by increasing its gain. Test equipment measures the gain and noise response of the LNA. These results are processed and stored in a computer coupled to the test equipment. A graphic display attached to the computer displays the LNA's untuned gain and noise response over a range of frequencies. A human operator then tunes the LNA by shaving conductive material from the circuit, or other appropriate means, while observing the gain and noise response on the graphic display. Typically, suitable gain is in the range of 35 ± 3 decibels (Db) over the frequency range of 3700 Mhz to 4200 MHz, with a preferred gain of 37 dB. A desired gain flatness is approximately three dB. In other words, the difference between the highest and lowest gain should not exceed three dB over the range of tested frequencies. An acceptable range of noise is typically between 20 to 70 degrees Kelvin, with lower values being preferred. The noise level of a circuit is measured in degrees Kelvin which rightfully implies that the temperature of the device can also affect the circuit's noise. Therefore, the temperature of the assembly environment should be maintained within a known range.

The block downconverter is then installed 20 on the same carrier board as the low noise amplifier. The downconverter is comprised of a tunable circuit coupled to a resonator in a tunable resonator cavity. The resonator is subjected to a first oscillator tuning 22. A satellite signal is typically at a frequency of approximately 3.7 to 4.2 GHz. The preferred resonator oscillates at approximately 5.15 GHz and produces an intermediate frequency (IF) of 950 to 1450 MHz. The 950-1450 MHz

fiequency range is used, since a signal within this frequency range can be transmitted

down a standard coaxial cable. Thus, at the first oscillator tuning, the LNB is attached to a test fixture and subjected to a frequency emulating a satellite signal.

The downconverter' s output signal frequency is measured and displayed. If the output signal is not within a predetermined frequency range, the resonator cavity can be tuned by advancing or retreating a tuning screw which penetrates the resonator

cavity. The device including both the low noise amp and the block downconverter is next subjected 24 to a known input signal over a range of frequencies. Test equipment measures the gain and noise response of the block downconverter/low noise amp assembly. These results are processed and stored in a computer coupled to the test equipment. Again, a graphic display attached to the computer illustrates the untuned gain and noise response of the downconverter/low noise amp over a range of frequencies. A human operator can then tune the block downconverter by shaving conductive material from the circuit, or other appropriate method, while observing the gain and noise response on the graphic display. The operator strives to obtain a gain of approximately 65 dB, a gain flatness of approximately six dB or less, and the lowest noise value possible. After tuning the block downconverter/low noise amp assembly, the device is subjected to a second oscillator test 26 to ensure that the resonator still produces an acceptable output frequency from the higher frequency input frequency. Upon completion of these tests and tunings, the device undergoes final LNB preparation 28. This final preparation can include placing the LNB into a heated chamber for a period of time and then retesting gain and noise responses. For example, the LNB may be heated to approximately fifty degrees Celsius for a period of thirty to thirty-five minutes. The LNB can also be placed in a "burn-in" room overnight where it can be subjected to current and/or increased temperatures for a period of time. This is typically done overnight because during burn-in the circuit

can emit sufficient noise to affect the testing of other devices on the assembly line.

After burn-in, each LNB cover is sealed, and painted. After the paint cures, a title plate can be attached to the LNB identifying the manufacturer and model number.

After final preparation, the LNB is subjected to a final test 30 on either the master final tester or on a slave final tester. Gain and noise are measured to ensure that they have not significantly changed in value. Note that the final test can be performed on the device up to several days after its assembly. Final test results are printed and the device is packaged 32. The final test results can be printed on an adhesive label which is then applied directly to the LNB or to the LNB package. The label can also include the LNB's individual identification number with which it has

been tracked through the entire assembly process.

FIGURE 2 provides a flow diagram of the steps involved in the calibration 100 of the amp tune station, the block tune station and the slave final test stations according to a preferred embodiment of the present invention. As discussed, a single

LNB can produce different gain and noise readings when tested at different test

stations. Likewise, the constituent low noise amp (LNA) or block downconverter (BDC), either individually or coupled, can also individually display varied responses to a fixed input signal depending upon the test station used for the measurements. In

other words, deviations in test readings may be caused by the test equipment as well

as the devices. The deviations typically occur due to variations in the components within the tester or wear and tear of the tester. Therefore, the first step of the present method involves maintaining both a reference LNB unit, comprised of a reference LNA and a reference BDC with known parameter values, and a carefully controlled and consistent master test station. A goal of the present method is to force the slave test stations to emulate this master test station. A second step involves measuring 102 the gain and noise for the reference unit with the master test station. The initial measurements by the master test station can be repeated several times to produce average parameter values. The reference unit is then taken to each of the block tune stations for testing 120 of the same parameters. For each block tune station, the difference between each parameter value at each block tune station compared to the parameter value determined at the master station for each frequency is determined and stored 122 in a "delta" file. This difference in values will be referred to herein as the "skew value. " The delta file is stored in the memory of that block tune station, and that station is programmed to alter measured values of nonreference units by the skew value in that delta file. Thereafter, when a nonreference LNB is tested at that station, the parameter values of the production LNB are altered automatically by the skew values in the delta file. For example, if the reference unit on the master test station produces a gain of 65 dB and a noise level of 35 degrees at 1.4 GHz, but only a 63 dB gain and 38 degree noise level at that frequency on a block tune station, then the delta skew values are +2 dB and -3 degrees at that frequency. A nonreference test unit may produce test results at that block tune station of 62 dB and 42 degrees. These values are automatically altered and the values of 64 dB gain and 39 degrees noise are graphically displayed to the human operator to aid in his tuning procedure. This process can be repeated 126 for each block tune station, thereby forcing each one to emulate the accuracy of the master tune station.

This process is repeated for each slave final test station. In other words, the reference unit is measured 102 at the master final test station for various parameters

such as noise and gain across a range of frequencies. The same reference unit is

taken to each of the slave final test stations for the same measurements 130. Any differences in values become the skew values which are read into a unique delta file in the memory of each of these slave final test station 132. Thereafter, the measurements of a nonreference unit at the slave final tester are automatically altered by the skew values in its unique delta file. This process is repeated 134 for each slave final test station.

In contrast, each amp tune station is not calibrated directly against the master final tester. Instead, each amp tune station is paired with an individual block tune station. This change in procedure occurs because at both the block tune station and the final test station, a fully assembled LNB is tested. However, at the amp tune station, the block down converter has not yet been installed on the carrier board. A reference unit is tested at a first block tune station and its gain and noise response is measured over a range of input signal frequencies 136. The block downconverter circuit on the reference unit is then removed 138. The same unit is then taken to the paired first amp tune station and tested 140 for gain and noise response over the same range of frequencies. These results are compared to those obtained with the reference unit with a BDC board attached. A delta file is created 142 with a skew value

representative of the differences between these response values. This delta file is stored with the amp tune station. Thereafter, the results of any nonreference unit test on that amp tune station will be varied 144 by the skew values contained in that station's delta file. These altered values are displayed by display means to aid the human operator tune the low noise amp. This process is repeated 146 for each amp

tune station.

By pairing an amp tune station with a block tune station, and then calibrating those stations together, overall measurement accuracy is increased. However, to maintain this level of accuracy, a nonreference device that is tested and tuned at a particular amp tune station should then be tested and tuned at the block tune station paired with that particular amp tune station. The accuracy of each station does not significantly degrade over the course of a single day. Still, the calibration of each test station can be performed every day before production begins.

FIGURE 3 provides a generic representation of the steps involved in calibrating 100 a slave tester to emulate a master tester in accordance with the present invention. A master final test station and a reference unit with known parameter responses are maintained 102. Initially, measurements are made 104 of a reference unit for a first parameter. Next, the same parameter of the reference unit is tested 106 on a first slave tester. These values are compared 108 and the differences are skew values which become part of a first delta file 110. The first delta file is stored 112 with the first slave tester. The first slave tester is also programmed 114 to alter any measured parameter values from a nonreference unit by the skew value in the delta file. This process is repeated 116 for each additional slave tester. Likewise, this process may be repeated 104a to 116a for a second parameter value, a third parameter value and so forth. Up to "N" parameters can be measured and compared 104n to 116n. After, the slave testers have been programmed to emulate the master

tester, production can begin 118.

FIGURE 4 illustrates the "dynamic calibration" aspect of the present invention. As discussed, the assembly of a low noise block downconverter involves a number of steps performed 12 to 32 in assembly line fashion. These steps are illustrated in simplified form and are more thoroughly discussed above. Each step in the assembly process is monitored 34 by the assembly line control system.

The control system comprises a computer which is programmed to monitor the progress of devices through the assembly system as well as the test results from each device being assembled. The control system is coupled to each assembly and test station. To track the progress of each device through the assembly process, the carrier board or other suitable element is marked with an identifying bar code. At each station along the assembly line, this code is scanned, thereby informing the control system of the location of each device in the assembly process. If testing is performed at that station, then the scanned code also allows the control system to associate the measured test data with the particular unit being tested. The flow of information to the control system is represented by the arrows pointing from the individual assembly or testing station to the control system block 34. As mentioned, the control system reads the results of every test run on each device being assembled. The control system specifically monitors the results of the

final test 30. If the performance characteristics of the assembled devices begins to

diminish downstream, then the control system can respond immediately by creating at least one dynamic skew file. The dynamic skew file operates in the same way as the static delta file, in that it alters the value measured by the tester. The altered value rather than the measured value is displayed to the human operator who is tuning the device. The dynamic skew file, however, differs from the static delta file in that the dynamic skew value can change after every assembled device undergoes its final test, whereas the static delta file is usually only changed at calibration. However, the dynamic skew file need not ever change during the course of an assembly run if the performance characteristics of the assembled devices are satisfactory.

This dynamic skew file is preferably copied to any station along the assembly line. For example, if the control system determines that downstream product performance would be improved if the testing and tuning of the LNA differed, it creates a first skew file 42 and loads this file to the LNA testing station computer. The loading of the skew file is represented by the arrow pointing from the skew file block 42 to the LNA test station block 18. If the LNA tester measured a noise value of 37 degrees and the skew file contained a value of -2 degrees, then the visual display of the tester would display a noise value of 35 degrees to the operator. The operator tunes the LNA based upon the values shown on his visual display.

Likewise, if the control system determines that the downstream product performance would best improve if the results of the first oscillator tuning station were altered, it creates a second skew file 44 and loads this skew file at the first oscillator test station. Again, the visual display at that station is altered by the value in the second skew file. A third, fourth, and fifth skew files 46, 48, 50 could be created and loaded to the appropriate test station 24, 26, 28.

The dynamic skew file may work in conjunction with, or exclusively of, the static delta file. When the two files are used together, the sum of the values of both files provide the total skew to be applied to a measured value at a test station to change the value displayed at that station. When the dynamic skew file is used exclusively, any value in the dynamic skew file will override the application of skew

values in the static delta file. The skew values in the static delta file would only be

used when the dynamic skew file is empty. Additionally, the control system may create one or several skew files or none at all. The skew files can be updated continually or not at all.

The control system 34 will be programmed to create various dynamic skew files whenever it detects final test results which are outside of a predetermined acceptable window. The triggering event may be the final test results for a single unit or the average results for many units. Experience with the device being manufactured will enable the programmer of the control system 34 to determine which of the testing and tuning stations in the process flow should be skewed in order to best compensate for inadequate final test results. Taking the LNB as an example, experience may show that it is best to do most of the gain flatness tuning at the initial tuning 18 of the LNA, while finer adjustments of the gain flatness may be made at the BDC tuning 24. In such a situation, the control system 34 would likely correct large deviations in gain flatness at final test by creating a dynamic skew file 42 for use while tuning 18 the LNA, while smaller deviations in gain flatness at final test would likely be corrected by creating a dynamic skew file at tuning 24 of the BDC.

Obviously, such decisions will be dependent upon the characteristics of the component

being manufactured. Furthermore, it will be understood by those skilled in the art that the term "final test", as used herein and in the appended claims, does not necessarily mean the last test that is performed on the device before it leaves the factory, but rather any test which is performed after the device is substantially assembled.

FIGURE 5 illustrates the effect of the "constraint" system on the present invention. The control system monitors 34 the assembly process. The constraint system, in turn, monitors 36 the control system. The constraint system comprises a program within the control system computer which recognizes problems, or "disrupts", in the assembly process. In other words, the constraint system program monitors the test data measured at the test stations and compares these values to predetermined values. If the measured values differ from an acceptable range, then a disrupt occurs. These disrupts include deteriorating product quality, backups in device assembly at a particular stage of production, and the like. For example, a disrupt condition might be an assembled device with a gain flatness greater than six degrees even after creation and downloading of the appropriate dynamic skew files. Another disrupt condition might be a queue of more than ten LNAs at the LNA tuning station.

Once a disrupt condition is recognized, the constraint system responds in one of several ways. One possible response could be that the constraint system signals 40 the pick and place machine to select a different component during the assembly of the LNA or BDC boards. For example, the constraint system may determine that a 60 ohm resistor in the LNA would produce better overall gain flatness than a 50 ohm resistor presently being placed in the circuit. Such a decision would be based upon information programmed into the constraint system about the relationships between various component values and device performance. Such information will obviously depend upon the device being manufactured. Alternatively, the constraint system may contact a human operator 38 on the assembly line floor. For example, the constraint system may dial the number of a portable phone carried by a manager on the floor and play a prerecorded message telling him that a backup has occurred

at the BDC installation station. The manager is then able to investigate the situation personally. The constraint system allows all information about the device assembly and test flow to be accessed by a single control system. This is an important advantage because many manufactured items, such as high frequency electronic devices, have performance characteristics which are determined by both manufacturing processes and post-manufacturing tuning processes. A control system which has access to information about both processes and is able to vary both processes according to that information will result in more consistent final product performance.

Although preferred embodiments of the invention have been described in the foregoing Detailed Description and illustrated in the accompanying drawings, it will be understood that the invention is not limited to the embodiments disclosed, but is capable of numerous rearrangements, modifications, and substitutions of parts and elements without departing from the spirit of the invention. Accordingly, the present invention is intended to encompass such rearrangements, modifications, and substitutions of parts and elements as fall within the scope of the invention.

Claims

CLAIMS:

1. A method of manufacturing a device with a plurality of tunable parameters comprising:

(a) measuring a selected parameter of a reference device with known parameter value on a master test station; (b) measuring the selected parameter of the reference device on a slave test station with a memory;

(c) comparing said parameter value from step (b) to that found in step (a)

to determine a difference in parameter values; and

(d) storing said difference into the memory of said slave test station.

2. The method of manufacturing of Claim 1 further comprises:

(e) measuring said parameter value of a nonreference device on the slave test station; and

(f) altering the measured parameter value by the difference stored in step (d);

3. The method of manufacturing of Claim 2 further comprises:

(g) displaying the altered measured parameter value on a display means; and (h) tuning the nonreference device in response to the displayed, altered, measured parameter value.

4. The method of manufacturing of Claim 3 further comprises:

(i) storing the tuned parameter value for the nonreference device; and

(j) labeling the device with the tuned parameter value.

5. The method of manufacturing of Claim 4 further comprises monitoring steps (a) through (h) and constraining steps (a) to (h) between predetermined limits such that any variation from said predetermined limits creates a disrupt.

6. The method of manufacturing of Claim 5 further comprises responding to said disrupt by contacting a human operator to investigate the disrupt.

7. The method of manufacturing of Claim 5 further comprises responding to said disrupt by altering the selection of components for the nonreference unit.

8. The method of manufacturing of Claim 2, wherein the step of measuring said parameter values -further comprises measuring circuit gain.

9. The method of manufacturing of Claim 2, wherein the step of measuring said parameter values further comprises measuring circuit noise level.

10. A method of manufacturing a device with a plurality of tunable parameters comprising:

(a) measuring a selected parameter of a reference device with known

parameter value on a master test station; (b) measuring the selected parameter of the reference device on a slave test station with a memory;

(c) comparing said parameter value from step (b) to that found in step (a)

to determine a difference in parameter values;

(d) storing said difference into the memory of said slave test station;

(e) measuring said parameter values of a nonreference device on the slave test station; (f) altering the measured parameter values by the difference stored in step (d);

(g) displaying the altered measured parameter values on a display means;

and (h) tuning the nonreference device in response to the displayed, altered, measured parameter values.

11. The method of manufacturing of Claim 10 further comprises:

(i) storing the tuned parameter value for the nonreference device; and (j) labeling the device with the tuned parameter values.

12. The method of manufacturing of Claim 11 further comprises monitoring steps (a) through (j) and constraining steps (a) to (j) between predetermined limits such that any variation from said predetermined limits creates a disrupt.

13. The method of manufacturing of Claim 12 further comprises responding to said disrupt by contacting a human operator to investigate the disrupt.

14. The method of manufacturing of Claim 12 further comprises responding to said disrupt by altering the selection of components for the nonreference unit.

15. The method of manufacturing of Claim 10, wherein the step of measuring said parameter values further comprises measuring circuit gain.

16. The method of manufacturing of Claim 10, wherein the step of measuring said parameter values further comprises measuring circuit noise level.

17. A method of manufacturing a device with a plurality of tunable parameters at an assembly line having at least one test station and a final test station, each test station having a computer coupled to test equipment and a display, said

method comprising: (a) measuring a selected parameter of a first device at the final test station;

(b) creating a dynamic skew file with a compensation value to compensate for a variation between the measured parameter and a predetermined value;

(c) downloading the dynamic skew file to a selected test station;

(d) altering the measurements of the same parameter of a second device

by the compensation value to create an altered parameter value.

18. The method of manufacturing of Claim 17 further comprises:

(e) displaying the altered parameter value to an operator;

(f) tuning the second device based on the displayed altered parameter

value.

19. A manufacturing system for use in assembling and tuning an electronic

device, said system comprising:

(a) a master test station;

(b) a slave test station, wherein both the master test station and the slave test station comprise a computer means coupled to a testing means and a display means, and wherein the master test station and the slave test station are connected by a control system; and

(c) means within said control system to determine a difference between a first parameter test result from the master test station with a second parameter test result from the slave test station, said difference being stored in the computer means of said slave test station.

20. The manufacturing system of Claim 19 further comprises:

(d) means to alter the second parameter measurement by the slave test station by said difference.

21. The manufacturing system of Claim 20 further comprises:

(e) means to display the altered second parameter measurement on said slave test station display means.

22. The manufacturing system of Claim 19 further comprises:

(d) means to monitor the control system for a disrupt condition; and

(e) means to constrain said disrupt condition.