US 4,660,150 AGrant
Spectrum Analyzer With Improved Data Analysis and Display Features
Issue Date:1987-04-21
•2 Claims
•18 Drawing Sheets
Abstract
An automatic spectrum analyzer is provided with enhanced features for measured data handling and analysis. One feature allows the user to establish a sequence of operations to be performed at the end of each sweep of the instrument. Another feature enables the user to specify a particular signal amplitude and to have a marker placed on a displayed signal trace at that amplitude. Still another feature provides a marker that can be positioned on the next signal peak to the right or the left of the current marker position, enabling the display of the amplitude and frequency of the signal on which the marker is placed.
Metadata
Assignee
- Hewlett-Packard Company
Inventors
- Gregory A. Anderson
- Paul M. Eason
- Michael S. Marzalek
- Lynn M. Wheelwright
Application Information
Application Number:US 5809916
Filing Date:1984-02-16
Priority Date:1984-02-16
Art Unit:234
Classifications
IPC:
G01R 2316G01R 2300
Field of Search:
36432473485;487;518;521;57977 B;77 D;57 SS659
Patent Drawings (18 sheets)
Description
Background of the Invention
A spectrum analyzer is essentially a receiver that is tuned or swept across a band of frequencies, and the amplitude of received signals is displayed on a cathode ray tube as a function of frequency. It is often desirable to make adjustments to the displayed signal information for the purposes of calibration, and it is also desirable to quantify the characteristics of the displayed information such as signal magnitude and frequency. These functions are most effectively performed if the information is converted from analog form to digital form and stored in a memory. Such digitization of the information also facilitates the automation of spectrum analyzer functions.
An automatic spectrum analyzer is disclosed in U.S. Pat. Nos. 4,253,152; 4,264,958; 4,257,104; and 4,244,024 which has a number of advanced data handling and analysis features, several of them connected with markers that can be placed on the display screen. These features allow the user to determine, for example, the frequency and amplitude of a particular point on a signal trace in the display where he has placed a marker. The user can also request the marker to be placed at the highest signal shown in the display with a peak search function, and the display will show the frequency and amplitude of the point found by the search. Another prior art spectrum analyzer, the Hewlett-Packard model 8566A, had an additional feature that allowed the user to have a marker placed on the next highest peak.
Another feature of the prior art device allows the user to specify a particular frequency on the display and then to cause a marker to be placed on the signal trace in the display at that frequency. The spectrum analyzer will then display the amplitude of the signal at the chosen frequency.
The prior art spectrum analyzer can be controlled from a keyboard on the front panel of the instrument or from a remote computer connected to the analyzer by the Hewlett-Packard Interface Bus (HP-IB). Using commands that represent functions available from the front panel of the instrument, a remote computer can thus programmatically control the spectrum analyzer.
Summary of the Invention
The preferred embodiment of the present invention is an improvement on the prior art spectrum analyzer referenced above and provides a number of functions that are not available in the prior art spectrum analyzer. These new functions provide significant advantages in the acquisition and use of spectrum analyzer data.
One of these new functions provides for the automatic execution of user defined functions at the end of each sweep of the spectrum analyzer. In the prior art, further processing of data from the spectrum analyzer by the user essentially had to be done by or under the command of an external computer, and there was a time delay associated with such processing. In the preferred embodiment of the present invention, processing can be done much more quickly since it is done by the spectrum analyzer itself when it reaches the end of a sweep. This function, for example, allows the user to cause the analyzer to indicate if the results of a test being performed are out of bounds set by the user or if a particular signal being sought has been found.
Another new function provided allows the user to specify a particular signal amplitude and request a marker be placed on the displayed signal at that amplitude. This amplitude marker function greatly facilitates the testing of electronic devices such as filters and amplifiers and also aids in the location or identification of signals. It is a significant improvement over the previously available feature that allowed only the manual placement of a marker on the display or the specification of the frequency at which a marker was to appear, because it essentially eliminates trial and error attempts to find specific signal amplitudes.
Frequently a number of signals of varying magnitudes will be displayed on the screen of a spectrum analyzer, and the user will want to determine the frequency and amplitude of each one. A prior art device provides a convenient function for placing a marker on the highest signal and displaying the frequency and amplitude of that signal. A next highest marker function can be used to measure each of the other signals in the order of their magnitudes. However, since the amplitudes of the signals may have no relationship with their order on the display, the marker may jump around on the screen in an inconvenient way as it is moved from signal to signal.
One of the new features of the preferred embodiment allows the user to cause the spectrum analyzer to place the marker on the next peak to the right or the left of the current marker position. By providing a next peak left or next peak right marker function, the preferred embodiment of the present invention provides a significantly improved way to measure the frequency and amplitude of each signal in the order it appears on the display.
Brief Description of the Drawings
FIGS. 1 and 2 show the combination of FIGS. 1A-1F and 2A-2E respectively.
FIGS. 1A-1F and 2A-2E are schematic diagrams of a controller assembly for a spectrum analyzer in accordance with the preferred embodiment of the present invention.
FIGS. 3 and 4 are flow diagrams of a function for performing user defined programs at the end of a sweep of the spectrum analyzer.
FIG. 5 shows the combination of FIGS. 5A-5C.
FIGS. 5A-5B show a flow diagram of an amplitude marker function.
FIG. 6 illustrates the operation of the amplitude marker function of FIG. 5.
FIG. 7 is an overall flow diagram illustrating the operation of several marker peak functions.
FIG. 8 shows the combination of FIGS. 8A-8C.
FIGS. 8A-8B show flow diagrams of a function for placing a marker on the next peak to the right or left of the current marker position.
FIG. 9 illustrates the operation of the function of FIGS. 8A-8B.
Detailed Description of the Preferred Embodiment
The present invention comprises improvements on the apparatus disclosed in U.S. Pat. No. 4,253,152, which is incorporated herein. That apparatus comprised a spectrum analyzer with circuits for digitizing measured analog signals and for displaying and analyzing the digitized signals in various ways. The following description is of those aspects of the preferred embodiment of the present invention that differ from the prior embodiment. The first difference is in the controller assembly, shown in FIGS. 38 and 39 of the prior patent and described principally starting at column 24, line 65 and continuing through column 25, line 47 thereof.
FIGS. 1A-1F and 2A-2E show the controller assembly of the preferred embodiment which replaces the controller assembly of the prior patent. Components shown in block form in these figures are identified in Appendix III. Block 102 is a clock with a crystal oscillator that runs at approximately 14.7 megaherz. The oscillator output signal is divided once for the processor clock signal and twice for the clock to run the HP-IB chip.
A reset circuit 104 verifies that power supply 106 is within tolerance before the processor starts operating so there is not an inadvertent loss or destruction of memory contents. Read/write decode buffering circuits 108 provide buffers for the memories (RAM and ROM) to enable high bite or low bite select outputs.
An address decoder 110 comprises some standard type decoders (models LS139 and LS138) and a couple of programmable array logic type decoders that do the ROM and RAM addressing. Address decoder 112 provides address decoding for several other circuits in the controller assembly, such as a parallel interface timer chip, an HP-IB chip and data bus buffers that buffer the instrument bus. The buffers are shown in block 114, and these buffers isolate the instrument bus from the processor data bus to protect from RFI inside the instrument. The output of those buffers is what drives the rest of the instrument. The buses and the types of signals available on those buses are the same as those in the apparatus described in the prior patent.
Block 116 includes buffer controls for the buffers in block 114. Block 118 has strobe generators for the input/output (I/O) bus to select between the display section and the RF section, each of which has its own set of addresses.
A peripheral interface and timer chip 120 is used to take care of some miscellaneous control lines and to drive some LEDs for signature analysis for fault diagnosis.
In block 122 is a Hewlett-Packard Interface Bus (HP-IB) chip along with two buffer chips. In addition, there is a set of address switches and a buffer U14 (an LS244) to read them back on to the processor data bus.
Block 202 is an interrupt decoder (an LS148). Self test block 204 includes some registers that are used in a self-test procedure check. These are registers where one can read an address in and then write them back out on the data bus to find out if they are really what they ought to be.
A microprocessor 206 is used to control the operation of the spectrum analyzer. A Motorola model 68000 microprocessor running at 8 megahertz is used instead of processor 2015 shown in FIG. 38A of the prior patent. The model 68000 microprocessor is a standard, commercially available one; and the instruction set for this microprocessor can be found in any number of published books, such as "MC68000 16-Bit Microprocessor User's Manual," third edition, 1982, published by Prentiss-Hall.
Since the microcode instruction set for this processor is different from the instruction set for the processor disclosed in the prior patent, the programs given in the appendices of the prior patent have been recompiled to run on the MC68000 microprocessor. However, all of the functions disclosed in the prior patent are also performed by the present embodiment in essentially the same manner, taking into account the differences between the two processors and their instruction sets. It should be noted that the processor reset signal referred to as POP in the prior patent is now called RESET in this embodiment, and the starting address of 40 for program execution has now been changed to 0. It should also be noted that the HSTM line is no longer used; and a stop circuit 3801 is no longer needed, this function now being performed by the microprocessor.
Block 208 comprises some interrupt decoding logic, and block 210 comprises some signature analysis jumpers that can be connected in such a way to force microprocessor 206 to perform a number of preselected operations repetitively for diagnostic purposes.
Block 212 contains read only memories (ROMs) that contain the programs for the various functions performed by the spectrum analyzer. These programs include those disclosed in the prior patent and those disclosed below. Block 214 contains random access memory (RAM) where data gathered by the spectrum analyzer as well as the results of data manipulations and various alterable parameters and instructions are stored. Both the RAM and ROMs are included on the same circuit board with the microprocessor rather than on separate circuit boards as in the device of the prior patent.
While the foregoing controller assembly using the MC68000 microprocessor is disclosed as the preferred embodiment, it will be understood by those skilled in the art that the following programs could just as easily be used in the embodiment disclosed in the prior patent by compiling them to run on the processor in that embodiment.
FIGS. 3 and 4 show flow diagrams for a function that automatically performs a user defined program at the end of each sweep of the spectrum analyzer. For simplicity, this function is referred to as "on end of sweep" or ONEOS.
The ONEOS feature actually comprises two temporally separate functions. One is the entry of a string of standard remote commands into the spectrum analyzer by the user. The spectrum analyzer stores those commands in memory for later use, and sets a flag in memory to indicate that such a command string has been stored. The other function occurs thereafter at the end of every sweep of the spectrum analyzer, when the spectrum analyzer accesses the command string and performs the commands. Examples of these command strings are limit checking and reprogramming the center frequency of the instrument. Any legal spectrum analyzer command can be put in the command string.
FIG. 3 shows the command string entry sequence. First the user selects an entry device, typically the keyboard of a remote computer connected to the spectrum analyzer through HP-IB, although it is possible to enter such command strings from the spectrum analyzer front panel as well. The command string is entered by first entering a command "ONEOS" followed by a string of commands in double quotes for the spectrum analyzer to perform. The string of commands is then stored in the spectrum analyzer memory and a flag is set saying that this command string has been entered.
An example of such a command string is the following:
This command string causes a marker peak search to be done and then to test the amplitude that the marker found to see if it is less than -30 dBm. If it is below that level, then the spectrum analyzer displays on the screen the message "SIGNAL TOO LOW" to alert the operator to the undesirable condition, so the operator can take appropriate action. It also generates a service request through the remote port on HP-IB to notify a controller that something is wrong so it can give an alarm or take other predetermined action.
FIG. 4 is a flow diagram of the procedure for checking whether there is an ONEOS command string to be performed and causing it to be performed if there is one. In the spectrum analyzer programming there is a routine that runs all the time, which, every time the analyzer comes to an end of a sweep, restarts the sweep. It is called DOEOS and is listed at lines 125 through 140 in Appendix II. In that routine there is a check made to see if the ONEOS flag has been set, indicating that the ONEOS command has been entered. If it has been entered, a procedure called DOSOFTKEY(-1) (lines 553-567 in Appendix II) is called and the command string is passed to the HP-IB command interpreter to be executed as if it were any remote set of commands. When the execution of the command string is complete, the next sweep is then started. This process then runs on a continuous basis.
It should be noted that the ONEOS function is useful not only in spectrum analyzers, but in other types of swept frequency measurement instruments. One example would be automatic network analyzers that have built-in processors or controllers analogous to the controller in the preferred embodiment.
Another function performed by the preferred embodiment is the placing of a marker on a trace on the spectrum analyzer display at a specified amplitude. This amplitude marker function is illustrated in the flow diagrams in FIGS. 5A-5B and the result of the function is illustrated in FIG. 6.
The function is invoked by first specifying a marker of type "amplitude," and this sets the mode of operation. Next the amplitude level at which the marker is to be placed is specified, and the spectrum analyzer places the marker at that amplitude on the data trace on the screen (if it can find it). If it cannot find a trace that passes through that amplitude, it finds the closest thing to it so if everything is above or everything is below, it will find the highest spot or the lowest spot on the trace respectively.
The search routine for the marker is called MKASRCH in the flow charts and in the supporting program listings (see lines 107 and following in Appendix I). At the beginning of the routine some parameters are initialized, including an array called AVOID (see lines 155-209 in Appendix I). This array specifies positions on the display that are to be avoided so that the marker is not set on top of other markers that already exist. Next, the displayed trace is fetched and loaded into the buffer where it can be worked on. Once that is accomplished, some initial search parameters are initialized. In particular, the current marker position and amplitude as well as the amplitude that is being looked for are set up.
At the beginning of the search a test is made in decision block 510 to find out if the marker really has to be moved at all or if it is already at the desired amplitude. If the answer is yes, the search is done, and nothing more needs to be done. Normally that is not the case, so the search left and right from the current marker position routine is started. The purpose of this search is to search both sides of the current marker position to see if the trace crosses the desired amplitude level anywhere on the display. First the search proceeds in one direction to the edge of the screen and then in the other direction to the other edge of the screen, unless a point satisfying the search condition is found before the edge of the screen is reached.
The program for the search left and search right routines is given at lines 116 through 153 of Appendix I. The search proceeds one data value at a time. Each data value is fetched one at a time to find out if it is greater than or less than the desired value. The program looks for something that either equals the desired value or crosses through the desired value from one point to the next. If the data value is initially greater than the desired value, then the program is looking for something less than or equal to the desired value. When that is found, then the trace has crossed through the desired value between these two digitized points. Once a candidate value is found it is stored for later use in the procedure.
If the search left was successful but the data point found in an existing marker, that value is not stored and the search left continues because existing markers are to be avoided. Once the search left is completed, the search right is performed in the same manner. If two candidates have been found for the marker, one to the left and one to the right of the original marker position, the closest one to the original position in frequency is chosen. If only one candidate was found, that position is chosen as the position for the marker.
If no candidates have been found, then the program checks to see if the current marker position is greater than the desired position. If it is, a minimum search is performed to find the lowest thing on the screen because that will be the closest in amplitude to the desired value. Likewise, if the current marker amplitude is not less than the desired amplitude, a peak search will be performed, because everything on the screen now is below the desired amplitude. The marker will then be placed on the highest point on the display.
Once the marker position is located, the marker is written to the display, and the next sweep begins. This search is done every end of sweep while the function is enabled. FIG. 6 illustrates the results of a marker amplitude operation, with a marker 602 being placed at the specified level of -40 dBm on the skirt of a signal trace 604 on the display.
FIGS. 7, 8A-8B and 9 illustrate a marker function that identifies the next peak on the display to either the right or the left of the current marker position. To perform this function there are basically two routines, one of which calls the other one. The first routine is called marker peak (MKPK), and it determines which kind of a peak search is to be done. It begins on line 71 of Appendix I. As shown in FIG. 7 peak searches are possible for the highest peak and the next highest peak, which will not be further discussed here, and for the next peak to the right or the left, which are described below.
When the desired routine is specified by the user, the program fetches the proper trace data and places it into buffer to be searched on and then branches to the appropriate search routine. When the next right or next left functions are specified, the first routine calls the second routine and passes the appropriate parameters to it to tell it which direction to search. After the new position for the marker has been found, the marker is written to the display.
The next peak routine, shown in FIGS. 8A-8B and listed in Appendix I starting at line 34, uses a number of parameters; these include the starting position for the search, the direction of the search, the length of the array being searched, the threshold above which or below which data points are to be ignored and an excursion parameter that describes the height of a peak with respect to the surrounding data. For example, a peak could be specified as a signal in the data array that is at least 6 dB higher than the data on either side of the peak, making 6 dB the excursion for that measurement. These parameters are user specified so the user can decide what size peaks to look for and where to look for them.
Once the initial values have been set, the search is started in the indicated direction. The goal of the search is to find a data point that is higher than points on either side of it by at least the amount of the excursion. For the purposes of the search, there are three variables that are used to determine when a peak has been found: PEAK, the current highest value found in the search; VALLEY1 the lowest value found on the approach side of PEAK; and VALLEY2, the lowest value found on the far side of PEAK.
The search commences with the VALLEY1 search routine which looks for a low point or valley between the starting point and the peak that is to be found. The first value that is taken out of the data buffer is then saved as the first VALLEY1 value in block 802. Next a check is made to ensure that the data point is within the bounds of the parameters set. At the same time a check is made to see if a peak has been found, for if it has, the procedure is finished and the marker can be placed on the display at that data point location. Then the next data point is fetched from the buffer, and it likewise is tested to see if it is within bounds.
Since the instrument is still in the VALLEY1 search mode, it next checks to see if the data point is less than the currently stored VALLEY1 point. If it is, that point becomes the VALLEY1 point because it is lower, indicating a descending slope in the direction of search. If, on the other hand, the point is higher than the current VALLEY1 point, then the data point is stored as the peak value, and the instrument changes to the PEAK search mode. Following that, the next data point is fetched.
Thus, for each data point, a check is first made to make sure that it is within bounds and then a three way determination is made to see which of the three search modes the instrument is in: VALLEY1, VALLEY2 or PEAK. Then further testing of the data is done depending on which search mode is being used.
Once the instrument is in the PEAK search mode, each new data point will be stored as PEAK so long as it is greater than the currently stored value of PEAK. When a data point is found that is less than PEAK, then the instrument checks the see if the difference between VALLEY1 and PEAK is greater than the excursion parameter. If that is the case, then the instrument changes to the VALLEY2 search mode, since the current PEAK value has passed one of the tests for being a peak.
In order to qualify as a peak to be marked on the display, the difference between PEAK and VALLEY2 must also be greater than the excursion parameter. If that is the case with the first data point after the peak is found, then the search is finished. If not, then successive data points need to be tested by the VALLEY2 search. Once a point is found that is low enough, the search is finished, and the marker can be placed on the display. If a data point is found that is greater than PEAK before the VALLEY2 condition is met, the instrument again changes to the PEAK search mode.
When the instrument is in PEAK mode, the next data point may be less than PEAK minus the excursion parameter, thus failing to satisfy the first condition for a true peak. As can be seen from the blocks in FIG. 8B following block 820, the instrument will change back to the VALLEY1 search mode until a new peak candidate is found. This portion of the routine allows the instrument to avoid labeling small bumps or saw teeth on the side of a trace as a peak if the height of the bump does not exceed the excursion parameter.
As mentioned above, once a data point is found that satisfies the search conditions of PEAK minus VALLEY1 and PEAK minus VALLEY2 both being greater than the excursion parameter, then the data point stored as PEAK is the desired point, and a marker can be placed at that point on the display.
FIG. 9 illustrates the operation of the next peak right function. The marker starts out on peak 910. When the next peak right function is performed, the marker will be moved to peak 911, the next peak to the right of peak 910. It can be seen that if the next peak right function is performed again, the marker will be on peak 912, whereas if the next peak function in the prior patent were used, the marker would be on peak 913 instead.
It should be noted that, while the amplitude marker feature and the next peak right and left features have been disclosed in connection with a spectrum analyzer, they could be used in other swept frequency instruments such as network analyzers. Furthermore, the next peak right and left features could be adapted to locate valleys or minimum points for use in network analyzers.
The following Appendices I and II give the program instructions for performing the functions described above and illustrated in the Figures. Most of the programs listed in Appendix II are ones that are called by the programs in Appendix I while they are running. These programs are written in the ALGOL language, and can be compiled to run on any appropriate processor. Appendix III is a list of the integrated circuit components used in the controller assembly shown in FIGS. 1A-1F and 2A-2E. ##SPC1##
Claims
What is claimed is:
1. A swept frequency measurement apparatus comprising: input means for receiving a signal; sweep generating means for generating a sweep signal and an end of sweep signal; variable frequency measurement means connected to the input means and the sweep generating means for producing a measurement signal in response to receipt of a signal by the input means; signal processing means connected to the variable frequency measurement means for processing the measurement signal to produce a processed signal; display means connected to the signal processing means for displaying processed signals; user input means for inputting by a user a command string corresponding to a sequence of operations to be performed; storage means for storing the command string; and control means connected to the storage means, the signal processing means and the sweep generating means for executing the command string to cause the stored sequence of operations to be performed in response to the end of sweep signal.
2. A method of making swept frequency measurements comprising the steps of: inputting by a user a command string corresponding to a sequence of operations to be performed on a measured signal resulting from a swept frequency measurement; storing the command string; receiving an electrical signal to be measured; generating a sweep signal and an end of sweep signal; performing a swept frequency measurement on the received electrical signal to produce the measured signal in response to the sweep signal; and automatically executing the command string to perform the sequence of operations on the measured signal in response to the end of sweep signal.
Patent Citations (14)
| Patent | Date | Inventor | Cited By |
|---|---|---|---|
| USRe31750 | 1984-11-01 | Morrow | |
| US3544894 | 1970-12-01 | Hartwell et al. | |
| US3641515 | 1972-02-01 | Sues | |
| US3904959 | 1975-09-01 | Britton, Jr. | |
| US3999004 | 1976-11-01 | Newton | |
| US4090126 | 1978-05-01 | Poticha et al. | |
| US4244024 | 1981-01-01 | Marzalek et al. | |
| US4253152 | 1981-02-01 | Holdaway | |
| US4255790 | 1981-03-01 | Hondeghem | |
| US4257104 | 1981-03-01 | Martin et al. | |
| US4264958 | 1981-04-01 | Rowell, Jr. et al. | |
| US4321528 | 1982-03-01 | Reichel et al. | |
| US4399512 | 1983-08-01 | Soma et al. | |
| US4533866 | 1985-08-01 | Zirwick |
Non-Patent Literature (3)
- Bliss et al., "A Low-Cost Spectrum Analyzer from Basic Modules" Electronic Engineering, Dec. 1970, pp. 45-47.
- Kraus, K. "Programmable Frequency Selector" Electronic Engineering, vol. 49, No. 588, p. 24, Feb. 1977.
- Klein, A. "Microprocessors Continue to Add Versatility to New Test Instrumentation" Control Engineering, Jul. 1978, pp. 41-42.