Electricity: measuring and testing – Impedance – admittance or other quantities representative of... – Calibration
Reexamination Certificate
2000-03-06
2002-07-09
Sherry, Michael (Department: 2829)
Electricity: measuring and testing
Impedance, admittance or other quantities representative of...
Calibration
C324S608000, C324S637000
Reexamination Certificate
active
06417674
ABSTRACT:
BACKGROUND
The present invention concerns radio frequency and microwave network analyzers and pertains particularly to automatic calibration of a network analyzer.
A radio frequency (RF) network analyzer system consists of a network analyzer and may include a multi-port test set. The network analyzer integrates a synthesized radio frequency source with built-in couplers for signal separation, a narrow band receiver, a display and a processor.
Measurement calibration is a process that improves measurement accuracy by using error correction arrays during signal processing to compensate for systematic measurement errors. Measurement calibration is also called Cal, accuracy enhancement, and error correction. Measurement errors are classified as random and systematic errors. Random errors, such as noise and connector repeatability are non-repeatable and not correctable by measurement calibration.
Systematic errors, such as tracking and crosstalk, are the most significant errors in most RF measurements. Systematic errors are repeatable and for the most part correctable, though small residual errors may remain. These systematic errors may drift with time and temperature.
Systematic errors are due to system frequency response, isolation between the signal paths, and mismatch in the test setup. Frequency response errors (transmission and reflection tracking) are gain errors that are a function of frequency.
Isolation errors result from energy leakage between signal paths in transmission measurements. This leakage is due to crosstalk. In reflection measurements, the leakage is also due to imperfect directivity. Directivity is the ability of the signal separation devices to separate forward traveling signals from reverse traveling signals.
Mismatch errors result from differences between the port impedance of the device under test (DUT) and the port impedance of the network analyzer. Source match errors are produced on the source (network analyzer RF OUT) side of the DUT; load match errors on the load (network analyzer RF IN) side. If the DUT is not connected directly to the ports the mismatch errors due to cables, adapters, etc. are considered part of the source or load match errors.
The network analyzer has several methods of measuring and compensating for these test system errors. Each method removes one or more of the systematic errors using equations called an error model. Measurement of high quality standards (for example, short, open, load, through) allows the network analyzer to solve for the error terms in the error model. The accuracy of the calibrated measurements is dependent on the quality of the standards used and the stability of the measurement system. Since calibration standards are very precise great accuracy can be. achieved.
To perform a transmission calibration, four measurement standards are utilized: for example, an open, a short, a load, and a through cable. The network analyzer measures each standard across a defined frequency band using a pre-defined number of points. The measurement of these standards are used to solve for the error terms in the error model and to remove systematic errors caused by frequency response and source match.
To perform a reflection calibration a one-port calibration is performed using three measurement standards: an open, a short, and a load. The network analyzer measures each standard across a predefined frequency band using a pre-defined number of points. The measurements of these standards are used to solve for the error terms in the error model and to remove systematic errors caused by directivity, source match and frequency response.
For further information about calibration of network analyzers, see for example, the HP 8712C and HP 8714C RF Network Analyzer User's Guide, Part No. 08712-90056, available from Hewlett-Packard Company, October, 1996, pp. 6-1 through 6-14.
Switching test sets can extend the measurement capability of network analyzers from a single pair of ports to multiple ports, and allows measurement of devices under test in the forward and reverse directions. The test sets allow significant increase of throughput when using a network analyzer to test a device by eliminating manual changing of device connections and enabling complete automation of the test process.
However, the addition of a test set after the network analyzer can significantly degrade the raw performance of the network analyzer system. The characteristics of the test set also drift with temperature. Vector error correction allows the system (composed of the network analyzer and test set) to achieve very good performance, but the drift of the test set makes frequent re-calibrations necessary. For some devices multi-port calibrations can take over 30 minutes to perform, and may need to be repeated frequently, for example, once or twice during an eight-hour shift. This significantly reduces the throughput improvement provided by the switching test set.
Two-port calibration of a network analyzer system using a test set requires external, open, short, load and through standards. The open, short and load must be connected to each port and measured. The through standard must be connected to each possible port pairing and measured in each direction. A four port test set requires six external through connections to complete a two-port calibration. A 12 port test set requires 66 external through connections to complete a two-port calibration. An N-port test set requires N(N-1)/2 external through connections to complete a two-port calibration. Therefore, the time required to perform a two-port calibration increases geometrically with the number of ports used in a multi-port test set. Thus as the number of ports increases, throughput is reduced due to the time spent for calibration.
In order to reduce the time required for calibration various systems have incorporated some automated features. For example U.S. Pat. No. 5,434,511, U.S. Pat. No. 5,467,021, U.S. Pat. No. 5,537,046, U.S. Pat. No. 5,548,221, U.S. Pat. No. 5,552,714 and U.S. Pat. No. 5,578,932 discuss electronic calibration accessories which perform computer-assisted calibrations with electronic standards, making the calibration process less time-consuming and error-prone. However, when using these electronic calibration accessories it is necessary to manually connect a module to the measurement ports. U.S. Pat. No. 5,587,934 also sets out an electronic calibration module that requires manual connections.
U.S. Pat. No. 5,578,932 sets a technique for including electronic calibration in a multi-port test set. This technique, however, appears to use precision-characterized electronic standards. This does now allow for calibration anywhere other than the network analyzer front panel ports.
U.S. Pat. No. 5,548,538 discloses a technique for including calibrations internal to the network analyzer. This technique involves the addition of a precisely characterized two-port module in front of the test set ports. An error matrix is determined for this two-port module relative to known standards at the test set ports. This error matrix can then be used to calculate the reflection coefficients of the electronic calibration standards inside the two-port module and to enable future automatic calibrations.
However, all of these existing techniques, when applied to a multi-port test set, are still very time consuming and complex when used for a full transmission calibration of a multi-port test set.
SUMMARY OF THE INVENTION
In accordance with the preferred embodiment of the present invention, a network analyzer system is calibrated. For example, the network analyzer system includes a radio frequency network analyzer or a microwave network analyzer. An initial calibration is performed for each test port in the network analyzer system. The initial calibration is performed by first performing a measurement calibration using calibration standards. The calibration standards are located in a reference plane located outside the network analyzer system. The measurement calibration yields initial calibration coefficient
Mastin Peter F.
Rowell, Jr. Joseph
Agilent Technologie,s Inc.
Patel Paresh
Sherry Michael
LandOfFree
Two port self-calibration for an N-port network analyzer does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Two port self-calibration for an N-port network analyzer, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Two port self-calibration for an N-port network analyzer will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2888445