RU2771481C1 - Method for vector calibration taking into account the intrinsic noise parameters of the meter - Google Patents

Abstract

FIELD: radar technology.SUBSTANCE: invention relates to the field of radar, radio navigation, long-range and ultra-long-range communications, television and radio broadcasting, scientific research and can be used to measure the noise coefficient of receivers, amplifiers, mixers, etc. HF and UHF devices. The method consists in carrying out vector calibration of the noise receiver of a vector analyzer of circuits with a noise generator and with at least six loads from a mechanical set of measures, the result of which is obtaining a vector of coefficients containing noise parameters of the noise receiver; the S-parameters of the device under study are measured; the relative noise power at the output of the device under study is calculated; the noise coefficient of the device under study is calculated based on the measured S-parameters and the relative noise power at the output of the device under study.EFFECT: increase in the accuracy of noise coefficient measurements by reducing the characteristic fluctuations in the measured frequency dependence of the noise coefficient of the device under study, due to the dependence of the meter noise on the input impedance.1 cl, 4 dwg

Description

The technical field to which the invention belongs

The invention relates to the field of radar, radio navigation, long-distance and ultra-long-range communications, television and radio broadcasting, scientific research and can be used to measure the noise figure of receivers, amplifiers, mixers, and other RF and microwave devices.

State of the art

One of the main parameters of microwave amplifying and receiving devices is their threshold sensitivity, which is related to the signal distinguishability limit determined by the receiver's own noise. One of the criteria describing the threshold sensitivity is the noise figure (NF), which is determined under the condition that the path in which the measurements are made is consistent. With a noise description of the device under study (DUT) in a mismatched path, the concept of “noise figure” alone is no longer enough, because signal re-reflections occur between the DUT and the load (or path), which introduces a mismatch error into the measurements. A vector network analyzer (VNA) that measures the complex reflection coefficient of the DUT allows the NR of the DUT to be estimated for an arbitrary load, compensating for the effects of mismatch. Another significant source of errors in the mismatched path can be the dependence of the intrinsic noise of the measuring equipment on the impedance of the connected DUT (see p. 28 in the book Almazov-Dolzhenko K.I. Noise figure and its measurement in the microwave / edited by academician N.D. Devyatkova, Moscow: Nauchny Mir, 2000, p. 240.), described by an equation with four noise parameters.

There are many methods and equipment that allow the measurement of NR and noise parameters. To measure the NR using spectrum analyzers or noise figure meters, as a rule, the Y-factor method is used (Noise figure meter. Kh5M-18: operation manual: in 3 hours / [CJSC NPF Mikran]. [Tomsk] : Mikran, cop. 2011. 3 hours URL: www.micran.ru"; "The Y Factor Technique for Noise Figure Measurements: Application Note / Mike Leffel, Rick Daniel; Rohde & Schwarz. Edition 1.2019. München: Rohde & Schwarz, 2019. 30, [1] p. URL: https://scdn.rohde-schwarz.com Doc. no. 1MA178_4E.").

This method does not provide for vector measurements, and, accordingly, cannot provide an accurate account of the mismatch effects and the dependence of the meter noise on the DUT impedance.

There is also a known method of "cold" source (Cold-method) for measuring NR using a VNA (patent US 10371733 publ. finding noise parameters and estimating the NR of the DUT using the Cold method for any load.

The disadvantage of this solution is that the calibration step does not provide for the measurement of VNA noise parameters, and this does not make it possible to exclude the influence of the DUT impedance on the VNA noise.

There is a known method for extracting the noise parameters of nonlinear devices (patent US 9929757 published on March 27, 2018), in which the calibration stage provides for measuring the S-parameters of the noise generator, DUT and VNA, as well as measuring the noise parameters of the VNA.

This method is designed to measure the NR of non-linear devices such as mixers, and represents them as multiport devices.

The disadvantage of this method is that the technical solution cannot be used for two-port linear devices. Moreover, the invention includes an impedance tuner and a specific vector measurement option with the ability to determine VC parameters, which is not available on all VNAs.

The closest analogue is the method of measuring the noise figure of the device under test using a network analyzer (Patent US 7804304 publ. its own power gain, the S-parameters of the DUT, and the GN and VNA reflection coefficients.

The disadvantage of this solution is that the dependence of the meter noise on the DUT impedance is not taken into account when estimating the NR.

The essence of the invention

The objective of the invention is to create a vector calibration method that takes into account the noise parameters of the VNA, which eliminates the influence of the dependence of the inherent noise of the VNA on the impedance of the DUT.

The technical result from the use of the invention is to increase the accuracy of measurements of the NR, by reducing the characteristic fluctuations in the measured frequency dependence of the noise figure of the device under study, due to the dependence of the noise of the meter on the input impedance.

To solve this problem and achieve the specified technical result, the present invention proposes a method for vector calibration of noise figure measurements, characterized in that, at the first stage, vector calibration of the VNA noise receiver is carried out by connecting a noise generator and at least six loads from a mechanical set of measures to the VNA measurement port , measuring the noise power from the switched on noise generator and the thermal noise power from the switched off noise generator and loads, the results of which calculate the noise parameters of the VNA noise receiver; at the second stage, the output of the device under study is connected to the port of the VNA with a noise receiver, and the input of the measuring device is connected to another port of the VNA, the probing signal is switched on and the S-parameters of the device under study are measured; at the third stage, the probing signal of the VNA is turned off and the noise dispersion at the output of the device under study is measured, while the VNA port acts as a matched load at the input of the device under study, the relative noise power at the output of the device under study is calculated from the data obtained; in the fourth step, the DUT noise figure is calculated based on the measured S-parameters and the relative noise power at the output of the DUT.

Brief description of the drawings

In FIG. 1 shows a variant of the sequence of actions (steps) for measuring the noise figure of the DUT for the implementation of the proposed invention, in Fig. 2 shows a diagram of the VNA vector noise calibration, FIG. 3 shows a diagram of the direct measurement of the NR of the DUT, in Fig. 4 shows the results of noise measurements obtained using the proposed invention and using the closest analogue (US 7804304), which are conventionally called "vector correction" and "scalar correction", respectively.

In FIG. 2, 3 marked:

1 - vector network analyzer;

2 - noise generator;

3 - the second port of the vector network analyzer with built-in noise receiver;

4 - a set of mechanical measures (of at least six loads);

"c1-c1" - section of the coaxial path, in which the connected noise generator, mechanical measures (or electronic calibrator, or impedance tuner) are connected to the second port of the VNA;

5 - device under study.

Implementation of the invention

The invention is a method for calibrating a vector network analyzer with a built-in noise receiver located on the first or second ports, in order to measure the noise figure of some two-port DUT. For measurements, use the VNA 1 with the option of measuring the NR, noise generator 2, loads from a set of measures, for example, a set of mechanical measures 4.

(комплексный коэффициент отражения либо от генератора шума в «холодном» или «горячем» состояниях, либо от нагрузок из набора мер) и

(дисперсия шума), вычисляемая ВАЦ 1 и связанная с входной мощностью и параметрами шумового приемника по формуле:To measure the NR of the DUT 5, at the first stage, for the vector calibration of the noise receiver of the VNA 1, the calibration scheme shown in FIG. 2, while the second port 3 of the VNA 1 is connected in turn to the noise generator 2 and at least six loads from the mechanical set of measures 4, measurements are taken for each connection on the VNA 1

(complex reflection coefficient either from the noise generator in cold or hot states, or from loads from a set of measures) and

(noise variance) calculated by VNA 1 and related to input power and noise receiver parameters as:

,

,

- дисперсия шума, вычисляемая ВАЦ 1, Вт;where:

- noise dispersion calculated by VNA 1, W;

;

;

- количество нагрузок из набора мер плюс два состояния генератора шума;

- the number of loads from the set of measures plus two states of the noise generator;

- мощность на входе ВАЦ 1, выделяемая на

, Вт, где

- опорный импеданс;

is the power at the input of VNA 1 allocated to

, W, where

- reference impedance;

- скалярный коэффициент, прямо пропорциональный коэффициенту усиления шумового приемника ВАЦ 1 и уровню теплового шума согласованной нагрузки, Вт;

is a scalar coefficient directly proportional to the gain of the VNA 1 noise receiver and the thermal noise level of the matched load, W;

- комплексный коэффициент отражения либо от генератора шума в «холодном» или «горячем» состояниях, либо от нагрузок из набора мер, отн. ед.;

- complex reflection coefficient either from the noise generator in the "cold" or "hot" states, or from the loads from the set of measures, rel. units;

- комплексный коэффициент отражения от шумового приемника ВАЦ 1, отн. ед.;

- complex reflection coefficient from the noise receiver VNA 1, rel. units;

- скалярный коэффициент, прямо пропорциональный коэффициенту усиления шумового приемника ВАЦ 1 и уровню прямой волны мощности его собственного шума (т.е., распространяющейся внутрь шумового приемника), Вт;

is a scalar coefficient directly proportional to the gain of the VNA 1 noise receiver and the level of the direct power wave of its own noise (i.e., propagating into the noise receiver), W;

- скалярный коэффициент, прямо пропорциональный коэффициенту усиления шумового приемника ВАЦ 1 и уровню обратной волны мощности его собственного шума (т.е., излученной шумовым приемником и отраженной от входного импеданса), Вт;

is a scalar coefficient directly proportional to the gain of the noise receiver of the VNA 1 and the level of the return wave of its own noise power (i.e., emitted by the noise receiver and reflected from the input impedance), W;

, отн. ед.;

, rel. units;

- комплексный коэффициент, прямо пропорциональный коэффициенту усиления шумового приемника ВАЦ 1 и ковариации прямой и обратной волн мощности его собственного шума, Вт.

is a complex coefficient directly proportional to the gain of the VNA 1 noise receiver and the covariance of the direct and backward power waves of its own noise, W.

At the end of the vector calibration, the system of equations is solved:

.

.

least squares method:

.

.

,

,

и комплексный

. При этом предполагается, что известны величины

и ENR генератора шума.The resulting vector X contains the parameters of the noise receiver VNA 1 - scalar

,

,

and comprehensive

. It is assumed that the values are known

and ENR of the noise generator.

и коэффициент передачи ИУ

.At the second stage, to measure the S-parameters of the DUT 5, the DUT 5 is connected to the ports of the VNA 1 in accordance with the diagram shown in FIG. 3, where the output of DUT 5 is connected to a port of VNA 1 with noise receiver 3, and the input of DUT 5 is connected to another port of VNA 1. Then the probe signal is turned on and the S-parameters of DUT 5 are measured: the reflection coefficient from the output of the DUT

and DUT gain

.

. на выходе ИУ 5. При этом порт 1 ВАЦ 1 выступает в качестве согласованной нагрузки («холодного» источника шума). Используя полученные на первом этапе калибровочные данные, вычисляют относительную мощность на выходе ИУ 5 по формуле:At the third stage, to measure the noise power at the output of DUT 5, the probing signal of VNA 1 is turned off and the noise dispersion is measured

. at the output of DUT 5. In this case, port 1 of VNA 1 acts as a matched load (“cold” noise source). Using the calibration data obtained at the first stage, calculate the relative power at the output of DUT 5 by the formula:

,

,

- мощность на входе ВАЦ 1, выделяемая на

, при подключенном ИУ 5, Вт;where:

is the power at the input of VNA 1 allocated to

, with connected OD 5, W;

- коэффициент отражения выхода ИУ 5, отн. ед.;

- coefficient of reflection of the output of the DUT 5, rel. units;

- дисперсия шума, вычисляемая ВАЦ 1, при подключенном ИУ 5, Вт;

- noise dispersion calculated by VNA 1 with DUT 5 connected, W;

.

.

After that, at the stage of calculating the NR of the MD 5, based on the measured S-parameters of the MD 5 and the relative noise power at the output of the MD 5, the NR of the MD 5 is calculated using the formula:

,

,

- коэффициент передачи по напряжению ИУ 5, отн. ед.where:

- voltage transfer coefficient TS 5, rel. units

FIG. 4 shows the noise measurement results that were obtained by measuring the noise figure of the cable assembly connected between the first and second ports of the VNA 1. From FIG. 4 it can be seen that the characteristic fluctuations associated with the mismatch effect and the dependence of the self-noise of the VNA 1 on the input impedance are noticeably smaller for the method according to the invention compared to the prototype method (ie, scalar calibration).

Thus, the proposed solution eliminates the error in measuring the noise figure of the device under study, due to the dependence of the intrinsic noise of the VNA 1 on the input impedance.

Claims (5)

A vector calibration method taking into account the intrinsic noise parameters of the meter, characterized in that - at the first stage, vector calibration of the noise receiver of the VNA vector network analyzer is carried out by connecting the noise generator and at least six loads from the mechanical set of measures to the measuring port of the vector network analyzer, measuring the noise power from the switched on noise generator and the thermal noise power from the switched off noise generator and loads , the results of which calculate the noise parameters of the noise receiver of the vector network analyzer; - at the second stage, the output of the device under study is connected to the port of the vector network analyzer with a noise receiver, and the input of the measuring device is connected to another port of the vector network analyzer, the probing signal is turned on and the S-parameters of the device under study are measured; - at the third stage, the probing signal of the vector network analyzer is turned off and the noise dispersion at the output of the device under study is measured, while the port of the vector network analyzer acts as a matched load at the input of the device under study, the relative noise power at the output of the device under study is calculated from the data obtained; - at the fourth stage, the noise figure of the device under study is calculated based on the measured S-parameters and the relative noise power at the output of the device under study.

Images