CN106100770A - A kind of noise coefficient measuring method based on two kinds of detecting ways - Google Patents

Abstract

The invention discloses a kind of noise coefficient measuring method based on two kinds of detecting ways, belong to technical field of measurement and test, the power that the present invention obtains according to mean value and two kinds of detecting ways of virtual value, calculate the noise power that vector network analyzer self produces, according to the relation of noise power and noise coefficient, obtain the noise coefficient of vector network analyzer；Then measured piece is connected, the total noise power that measurement measured piece and vector network analyzer produce and the gain of measured piece, calculate measured piece and the noise coefficient of vector network analyzer cascade；Cascading equations finally according to noise coefficient calculates the noise coefficient of measured piece.The present invention does not needs the measurement that accurate scaled noise source just can realize noise coefficient, it is not necessary to configuration Noise Factor Analyzer, just can realize the high accuracy of noise coefficient, quick scanning survey, save testing cost；And noise bandwidth can offset in calculating process, overcome the adverse effect to certainty of measurement for the bandwidth degree of accuracy.

Description

A kind of noise coefficient measuring method based on two kinds of detecting ways

Technical field

The invention belongs to technical field of measurement and test, be specifically related to a kind of noise-factor measurement side based on two kinds of detecting ways Method.

Background technology

Noise coefficient is sign receiver and building block processes small-signal energy in the presence of having thermal noise One of key parameter of power.The bit error rate (BER) in receiver system and noise coefficient (NF) all characterize receiver and process little The ability of signal, wherein noise coefficient is possible not only to characterize the performance of whole receiver system, can also be used to characterize receiver The noiseproof feature of all parts (such as parts such as preamplifier, frequency mixer, intermediate frequency amplifiers) in system, therefore noise coefficient Measurement is all very crucial for research and development and the manufacture of product.

Current noise-factor measurement most common method is Y factor method (being also cold/heat source method).Y factor method uses one The noise source of accurate calibration, provides two known input noise level, by noise under noise source on an off two states Power measurement, can calculate the noise coefficient of measured piece.Noise Factor Analyzer uses Y factor method, when noise source has well Source mate and when can be directly connected to DUT, the certainty of measurement of this method is high.

The other method of noise coefficient is referred to as low-temperature receiver method, and the method is applicable to vector network analyzer test noise coefficient, This method combines based on the unique source calibration method of vector error collimation technique and PNA-X, can obtain industry rs most Noise-factor measurement precision.

Both the above method user must configure Noise Factor Analyzer and (or be built in the special of vector network analyzer and make an uproar Acoustic receiver module) and the accurate noise source calibrated, testing cost is high.

Implementation the most close with the present invention in prior art is as it is shown in figure 1, the method is based on signal generator and letter Number analyzer carries out noise-factor measurement, it is not necessary to the noise source of accurate calibration.

(1) measuring principle:

The method is input one the 50 Ω matched load of termination at measured piece, measures the noise power of its output (N_o), according to the definition of noise coefficient, calculate noise coefficient by equation below:

${\mathrm{NF}}_{\left(dB\right)}=10\lg \left(\frac{N_o}{{\mathrm{GkT}}_{0}B}\right)$

Wherein: the gain of G measured piece；

K Boltzmann constant (1.38 × 10^-23J/k)；

T₀Standard noise temperature (290K)；

B receiver bandwidth (unit Hz).

(2) measuring process:

1) such as Fig. 1 connecting test system, and preheat fully.The frequency arranging signal analyzer is Frequency point of interest, The bandwidth reducing signal analyzer connects matched load to 1kHz, the input port at signal analyzer, it is ensured that carry out enough surveys Examination number of times is to obtain optimal mean values；

2) reading is converted into W from dBm, and divided by resolution bandwidth to obtain signal analyzer at respective frequencies point normalizing The noise power spectral density numerical value changed, uses N_saRepresenting, the reception channel gain of signal analyzer is 1, the machine of signal analyzer Noise coefficient is designated as F₂, then:

$F_2=\frac{N_{sa}}{{\mathrm{kT}}_o}$

3) matched load is changed, connect the output of measured piece to signal analyzer；Export one by signal generator Low power pumping signal (typically requires that the amplitude of input signal should be lower than measured piece 1dB compression point to the input of measured piece 10dB), the amplitude output signal of measured piece is measured with signal analyzer, such that it is able to obtain the gain G of measured piece_DUT；

4) driving source taking off measured piece input port is changed to matched load, obtains connecting total output noise during measured piece Power, by result divided by resolution bandwidth to obtain measured piece and signal analyzer normalized total noise power spectrum density number Value, uses N_mRepresent；The noise coefficient of measured piece and signal analyzer cascade is designated as F₁₂, then:

$F_{12}=\frac{N_m}{G_{DUT}{\mathrm{kT}}_o}$

5) by above-mentioned F₂、F₁₂And gain G_DUTNumerical value brings the cascading equations of noise coefficient into, can obtain the noise coefficient of measured piece F_DUT:

$F_{DUT}=F_{12}-\frac{F_2-1}{G_{DUT}}=\frac{N_m}{G_{DUT}{\mathrm{kT}}_o}-\frac{(N_{sa}/{\mathrm{kT}}_o)-1}{G_{DUT}}$

To step 5) in formula arrange and can obtain the noise coefficient that measured piece dB represents:

${\mathrm{NF}}_{DUT}=10\lg (\frac{N_m-N_{sa}}{{\mathrm{kT}}_0}+1)-10\lg G_{DUT}$

The implementation of Fig. 1 the most close with the present invention, major defect:

1) can only point-frequency measurement, measuring speed is slow, and measurement process is complicated, and the manual operand of user is big；

2) certainty of measurement is had a certain impact by the Measurement bandwidth degree of accuracy of signal analyzer.

Content of the invention

For above-mentioned technical problem present in prior art, the present invention proposes a kind of making an uproar based on two kinds of detecting ways Sound coefficient measuring method, reasonable in design, overcome the deficiencies in the prior art, measuring speed is fast, measurement process is succinct.

To achieve these goals, the present invention adopts the following technical scheme that

A kind of noise coefficient measuring method based on two kinds of detecting ways, uses vector network analyzer, including walk as follows Rapid:

Step 1: vector network analyzer is preheated, and its ginseng including measurement frequency range and bandwidth is set Number；

Step 2: source power output calibration and full two-port calibration are carried out to vector network analyzer；

Step 3: start virtual value detecting way and the mean value detecting way of vector network analyzer, according to two kinds of detections The measurement result of mode, calculates noise power N that vector network analyzer self produces_VNA-add；

Step 4: according to formula (1), calculate the noise coefficient of vector network analyzer, be designated as F_VNA；

N_VNA-add=(F_VNA-1)G_VNAkT₀B (1)；

Wherein, N_VNA-addThe noise power producing for vector network analyzer self；G_VNAConnecing for vector network analyzer Receive the gain of passage；K is Boltzmann constant；T₀For standard noise temperature；B is the band of the reception passage of vector network analyzer Wide；

Step 5: access measured piece between the source output terminal mouth and reception passage of vector network analyzer, start S parameter Measurement thread, the available gain G of measurement measured piece_DUT；

Step 6: start virtual value detecting way and the mean value detecting way of vector network analyzer, according to two kinds of detections The measurement result of mode, calculates measured piece and total noise power N of vector network analyzer generation_DUT+VNA；

Step 7: according to formula (2), calculates measured piece and the noise coefficient of vector network analyzer cascade, is designated as F₁₂；

N_DUT+VNA=(F₁₂-1)G_DUTG_VNAkT₀B (2)；

Wherein, N_DUT+VNAThe total noise power producing during for measured piece and vector network analyzer cascade；F₁₂For measured piece Noise coefficient with vector network analyzer cascade；G_DUTAvailable gain for measured piece；G_VNAConnecing for vector network analyzer Receive the gain of passage；K is Boltzmann constant；T₀For standard noise temperature；B is the reception of measured piece and vector network analyzer Bandwidth during passage cascade, requires that when measurement the band of measured piece is wider than the bandwidth of the reception passage of vector network analyzer, Cascade bandwidth depends on the bandwidth of the reception passage of vector network analyzer；

Step 8: by F_VNA、G_DUTAnd F₁₂Bring the cascading equations of noise coefficient into, calculate the noise coefficient of measured piece:

$F_{DUT}=F_{12}-\frac{F_{VNA}-1}{G_{DUT}}.$

Preferably, in step 3, specifically include

Step 3.1: starting virtual value detecting way and the mean value detecting way of vector network analyzer, vector network divides Frequency component before analyzer detection includes signal and noise two parts, is designated as x_i, then

x_i=s+n_i(3)；

Wherein: x_iAmplitude summation for signal and noise；S is signal amplitude；n_iFor noise amplitude；

Step 3.2: start the average value detection mode of vector network analyzer, can obtain:

$|x_{AVG}{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_i{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\left(s+n_i\right){|}^2=|s{|}^2+|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{n}_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(4\right);$

Step 3.3: start the virtual value detecting way of vector network analyzer, can obtain:

$x_{RMS}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|x_i{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|s+n_i{|}^2=|s{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|n_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(5\right);$

Step 3.4: formula (4) and (5) are subtracted each other, can obtain:

${x_{RMS}}^2-|x_{AVG}{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}|n_i{|}^2-|\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}{n}_i{|}^2---\left(6\right);$

Step 3.5: by the formula (6) in step 3.4, noise power N of vector network analyzer can be obtained_VNA-add:

$N_{VNA-add}=\frac{1}{R_L}({x_{RMS}}^2-|x_{AVG}{|}^2)---\left(7\right);$

Wherein, R_LRepresent system impedance.

Measuring principle of the present invention is as follows:

First the present invention carries out source power output calibration and full two-port calibration to vector network analyzer, passes through the present invention The power that the mean value proposing and two kinds of detecting ways of virtual value obtain, calculates the noise that vector network analyzer self produces Power, according to the relation of noise power and noise coefficient, calculates the noise coefficient that vector network analyzer self produces；Then Connect measured piece, the total noise power that measurement measured piece and vector network analyzer produce and the available gain of measured piece, calculate Go out measured piece and the noise coefficient of vector network analyzer cascade；Cascading equations finally according to noise coefficient calculates measured piece Noise coefficient.

The Advantageous Effects that the present invention is brought:

The present invention proposes a kind of noise coefficient measuring method based on two kinds of detecting ways, compared with prior art, originally Invention is suitable for vector network analyzer measurement noise coefficient, and the present invention does not needs accurate scaled noise source and just can realize making an uproar The measurement of sonic system number, it is not necessary to configuration Noise Factor Analyzer, just can realize the high accuracy of noise coefficient, quick scanning survey, joint About testing cost；And noise bandwidth can offset in calculating process, overcome side the most close with the present invention in prior art Case in power spectral density transfer process, the impact on certainty of measurement for the signal analyzer bandwidth degree of accuracy.

Brief description

Fig. 1 is existing and the hardware elementary diagram of the present invention the most close noise coefficient test system.

Fig. 2 is the hardware elementary diagram of noise coefficient of the present invention test system.

Fig. 3 is the FB(flow block) of a kind of noise coefficient measuring method based on two kinds of detecting ways of the present invention.

Detailed description of the invention

Below in conjunction with the accompanying drawings and the present invention is described in further detail by detailed description of the invention:

Measuring principle of the present invention is as follows:

The principle of the method is the relation of the noise power based on Linear Network self generation and noise coefficient, needs accurately The noise power that the gain of measurement measured piece and measured piece self produce.

N_add=(F-1) GkT₀B (8)

Wherein: N_addThe noise power that measured piece self produces；

The noise coefficient of F measured piece；

The gain of G measured piece；

K Boltzmann constant (1.38 × 10^-23J/K)；

T₀Standard noise temperature (290K)

B receiver bandwidth.

The present invention is suitable for vector network analyzer, by the error correcting technology that vector network analyzer is advanced, permissible Accurately measure the gain G of measured piece_DUT。

The computational methods of the present invention for convenience of description, represent arrow with the wave filter in Fig. 2 dotted line frame, low-converter, A/D etc. Amount network analysis instrument receiver core cell, can be regarded as a unified entirety.Frequency component before detection includes signal It with noise two parts, is designated as x_i:

x_i=s+n_i(3)；

Wherein: x_iAmplitude summation for signal and noise；S is signal amplitude；n_iFor noise amplitude；

By average value detection, can obtain:

$|x_{AVG}{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_i{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\left(s+n_i\right){|}^2=|s{|}^2+|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{n}_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(4\right);$

By virtual value detection, can obtain:

$x_{RMS}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|x_i{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|s+n_i{|}^2=|s{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|n_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(5\right);$

Can draw the difference of two kinds of wave detector square values by formula (4) and (5):

${x_{RMS}}^2-|x_{AVG}{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}|n_i{|}^2-|\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}{n}_i{|}^2---\left(6\right);$

Equation (6) the right Section 1 represents the root-mean-square value of noise voltage, and Section 2 represents the average of average noise voltage Value, from the regularity of distribution of the voltage of broadband noise, when average time M is enough, the second entry value is zero.

Thus can get noise power N of network building-out_add:

$N_{add}=\frac{1}{R_L}({x_{RMS}}^2-|x_{AVG}{|}^2)---\left(7\right);$

R_LRepresent system impedance, the additive noise power of measured piece can be accurately obtained by calibration of power technology.

The noise coefficient of measured piece can be obtained by formula (8) network building-out noise power, noise coefficient and gain relationship.

The described noise coefficient measuring method based on two kinds of detecting ways (as shown in Figure 3), comprises the steps:

Step 1: connect noise coefficient test system according to Fig. 2, remove measured piece, vector network analyzer is carried out pre- Heat, and its parameter including measurement frequency range and bandwidth is set；

Step 2: carry out source power output calibration and full two-port calibration to vector network analyzer；

Step 3: start virtual value detecting way and the mean value detecting way of vector network analyzer, according to two kinds of detections The measurement result of mode, calculates noise power N that vector network analyzer self produces_VNA-add；

Step 4: according to formula (1), calculate the noise coefficient of vector network analyzer, be designated as F_VNA；

N_VNA-add=(F_VNA-1)G_VNAkT₀B (1)；

Wherein, N_VNA-addThe noise power producing for vector network analyzer self；G_VNAConnecing for vector network analyzer Receive the gain of passage；K is Boltzmann constant；T₀For standard noise temperature；B is the band of the reception passage of vector network analyzer Wide；

Step 5: access measured piece between the source output terminal mouth and reception passage of vector network analyzer, start S parameter Measurement thread, the available gain G of measurement measured piece_DUT；

Step 6: use the method the same with step 3, starts the virtual value detecting way and averagely of vector network analyzer Value detecting way, according to the measurement result of two kinds of detecting ways, calculates measured piece and the overall noise of vector network analyzer generation Power N_DUT+VNA；

Step 7: according to formula (2), calculates measured piece and the noise coefficient of vector network analyzer cascade, is designated as F₁₂；

N_DUT+VNA=(F₁₂-1)G_DUTG_VNAkT₀B (2)；

Wherein, N_DUT+VNAThe total noise power producing during for measured piece and vector network analyzer cascade；F₁₂For measured piece Noise coefficient with vector network analyzer cascade；G_DUTAvailable gain for measured piece；G_VNAConnecing for vector network analyzer Receive the gain of passage；K is Boltzmann constant；T₀For standard noise temperature；B is the reception of measured piece and vector network analyzer Bandwidth during passage cascade, requires that when measurement the band of measured piece is wider than the bandwidth of the reception passage of vector network analyzer, Cascade bandwidth depends on the bandwidth of the reception passage of vector network analyzer；

Step 8: by F_VNA、G_DUTAnd F₁₂Bring the cascading equations of noise coefficient into:Calculate tested The noise coefficient of part:

$F_{DUT}=F_{12}-\frac{F_{VNA}-1}{G_{DUT}}.$

The present invention uses mean value and two kinds of detecting ways of virtual value to carry out calculation process, it is not necessary to the noise of accurate calibration Source, as long as providing continuous wave signal, the difference to signal and noise response by average value detection and virtual value detection, counts respectively Calculate vector network analyzer self produce noise power and measured piece and vector network analyzer cascade produce always make an uproar Acoustical power；Then the relation according to noise power and noise coefficient, calculate respectively the noise coefficient of vector network analyzer with And noise coefficient when measured piece and vector network analyzer cascade；Again by starting vector network analyzer S parameter measurement line Journey, the gain of measurement measured piece；Cascading equations finally according to noise coefficient calculates the noise coefficient of measured piece.

The present invention does not needs the measurement that accurate scaled noise source just can realize noise coefficient, it is not necessary to configuration noise coefficient divides Analyzer, just can realize the high accuracy of noise coefficient, quick scanning survey, saves testing cost；And noise bandwidth is at calculating process In can offset, overcome scheme the most close with the present invention in prior art in power spectral density transfer process, signal divides The impact on certainty of measurement for the analyzer bandwidth degree of accuracy.

Certainly, described above is not limitation of the present invention, and the present invention is also not limited to the example above, and this technology is led Change, remodeling, interpolation or the replacement that the technical staff in territory is made in the essential scope of the present invention, also should belong to the present invention's Protection domain.

Claims (2)

1. the noise coefficient measuring method based on two kinds of detecting ways, uses vector network analyzer, it is characterised in that: bag Include following steps:

Step 1: vector network analyzer is preheated, and its parameter including measurement frequency range and bandwidth is set；

Step 2: carry out source power output calibration and full two-port calibration to vector network analyzer；

Step 3: start virtual value detecting way and the mean value detecting way of vector network analyzer, and according to both detections The detection result of mode, calculates noise power N that vector network analyzer self produces_VNA-add；

Step 4: according to formula (1), calculate the noise coefficient of vector network analyzer, be designated as F_VNA；

N_VNA-add=(F_VNA-1)G_VNAkT₀B (1)；

Wherein, N_VNA-addThe noise power producing for vector network analyzer self；G_VNAReception for vector network analyzer is led to The gain in road；K is Boltzmann constant；T₀For standard noise temperature；B is the bandwidth of the reception passage of vector network analyzer；

Step 5: access measured piece between the source output terminal mouth and reception passage of vector network analyzer, start S parameter measurement Thread, the available gain G of measurement measured piece_DUT；

Step 6: start virtual value detecting way and the mean value detecting way of vector network analyzer, and according to both detections The detection result of mode, calculates measured piece and total noise power N of vector network analyzer generation_DUT+VNA；

Step 7: according to formula (2), calculates measured piece and the noise coefficient of vector network analyzer cascade, is designated as F₁₂；

N_DUT+VNA=(F₁₂-1)G_DUTG_VNAkT₀B (2)；

Wherein, N_DUT+VNAThe total noise power producing during for measured piece and vector network analyzer cascade；F₁₂For measured piece and arrow The noise coefficient of amount Network Analyzer cascade；G_DUTAvailable gain for measured piece；G_VNAReception for vector network analyzer is led to The gain in road；K is Boltzmann constant；T₀For standard noise temperature；B is the reception passage of measured piece and vector network analyzer Bandwidth during cascade；

Step 8: by F_VNA、G_DUTAnd F₁₂Bring the cascading equations of noise coefficient into, calculate the noise coefficient of measured piece:

$F_{DUT}=F_{12}-\frac{F_{VNA}-1}{G_{DUT}}.$

2. the noise coefficient measuring method based on two kinds of detecting ways according to claim 1, it is characterised in that: in step In 3, specifically include

Step 3.1: start virtual value detecting way and mean value detecting way, the vector network analyzer of vector network analyzer Frequency component before detection includes signal and noise two parts, is designated as x_i, then

x_i=s+n_i(3)；

Wherein: x_iAmplitude summation for signal and noise；S is signal amplitude；n_iFor noise amplitude；

Step 3.2: start the average value detection mode of vector network analyzer, can obtain:

$|x_{AVG}{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_i{|}^2=|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\left(s+n_i\right){|}^2=|s{|}^2+|\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{n}_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(4\right);$

Step 3.3: start the virtual value detecting way of vector network analyzer, can obtain:

$x_{RMS}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|x_i{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|s+n_i{|}^2=|s{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}|n_i{|}^2+\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Re}\left({\mathrm{sn}}_i^{*}+s^{*}{n}_i\right)---\left(5\right);$

Step 3.4: formula (4) and (5) are subtracted each other, can obtain:

${x_{RMS}}^2-|x_{AVG}{|}^2=\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}|n_i{|}^2-|\frac{1}{M}\underset{i=1}{\overset{M}{\Σ}}{n}_i{|}^2---\left(6\right);$

Step 3.5: by the formula (6) in step 3.4, noise power N of vector network analyzer can be obtained_VNA-add:

$N_{VNA-add}=\frac{1}{R_L}({x_{RMS}}^2-|x_{AVG}{|}^2)---\left(7\right);$

Wherein, R_LRepresent system impedance.