JPH0980100A - Measuring method by spectrum analyzer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simply show time progress of noise level change by displaying the spectrum waveform of an input signal on one half part of a display screen and displaying noise on the other half part in a time domain. SOLUTION: A carrier frequency is positioned in the center of the left half of a display screen, and frequencies upward and downward separated by an offset value fof from the carrier frequency on both ends, and each spectrum therebetween is displayed. In other words, a spectrum waveform (frequency region) is displayed. In addition, the level of noise of a frequency fN upward separated by the offset value fof from the carrier frequency is displayed in time domain on the right half part of the display screen. In spectrum display for the frequency region of the left half part, frequency sweep is performed at least between fc ±Fof in regard to an input signal, necessary parts are taken out from data taken in a backup memory and are displayed. Taken-in display of the data for the left part display and the other taken-in display of the data for the right part display are performed alternately.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention uses a spectrum analyzer to measure S / N ratio (signal noise ratio), C / N ratio (carrier level / noise level of a specific frequency) of continuous waves and burst waves. On how to do.

[0002]

2. Description of the Related Art FIG. 4 shows a general configuration example of a spectrum analyzer. The output signal of the object 11 is input to the spectrum analyzer 15 as an input signal. In the spectrum analyzer 15, the input signal is supplied to the frequency mixer 17 through the input variable attenuator 16, and the frequency sweep oscillator 1
8 is frequency-mixed with the local signal, the mixed output is supplied to the bandpass filter 19, the output of the filter 19 is amplified by the amplifier 21, and the amplified output is the frequency mixer 2
At 2, the frequency is mixed with the local signal from the local oscillator 23,
The intermediate frequency signal is taken out by the bandpass filter 24, the output of the filter 24 is detected by the detector 26,
The detected output is passed through the low pass filter 27, and then A
The signal is converted into a digital signal by the / D converter 28 and stored in the buffer memory 29. The control unit 31 is a so-called CPU, and sets the amount of attenuation of the attenuator 16 according to the parameter set by the parameter setting means 32, and controls the ramp voltage generator 34 through the timing controller 33 so as to control the sweep frequency oscillator 18. Control, that is, setting of the sweep frequency band, setting of the bandwidth RBW of the filters 19 and 24, setting of the bandwidth VBW of the filter 27, setting of the sampling period of the A / D converter 28, etc. Display control of the stored data on the display 35 is performed.

Conventionally, for example, C of a continuous wave input signal is used.
When measuring / N, the carrier frequency of the signal you want to measure,
Noise level frequency f to be known for this signal
_N (usually specified according to the modulation format of the signal, for example) is set and input, the memory 29 is taken out from the data,
For example, it is displayed as shown in FIG. 1B, and the data value L _C of the carrier frequency and the data value L of the noise frequency f _N are displayed.
Displaying the ratio L _C / L _N with _N the display screen. in this case,
Since the noise level L _N changes randomly, the bandwidth VBW of the low-pass filter 27 is normally set to be relatively narrow, about 1/10 of the bandwidth RBW of the band-pass filters 19 and 24, and the measurement noise is reduced. The levels are averaged.

[0004]

Conventionally, the bandwidth VBW of the low-pass filter 27 is made small in order to know the noise level as accurately as possible. On the other hand, the measurement time, that is, the frequency sweep time T _S can be expressed by the following equation when the frequency span (measurement frequency range) is S _pan (Hz).

T_S= S_pan(Hz) / {RBW (Hz) × min (RBW, VBW) (Hz) × 0.5} (sec) ………… (1) min (RBW, VBW) is the smaller of RBW and VBW
You. As mentioned above, VBW = RBW / 10 is set.
Measurement time T_SIs T_S= 10 x S _pan/
{(RBW)²× 0.5} (sec), and the measurement time is relatively long.
There was the problem of being long.

Further, since the noise is displayed as shown in FIG. 1B, the noise is displayed at one point of the frequency f _N , and its level fluctuates up and down, but the integration of the low pass filter 27 and the display integration of the display 35 are performed. Due to the effect, it was displayed as an almost constant level, and it was not possible to know the randomly changing state.

[0007]

According to the present invention, the spectral waveform of the input signal is displayed on one half of the display screen, and the noise of the frequency associated with the input signal is displayed in the time domain. Display in half.

[0008]

1A shows a display example on a display screen when the present invention is applied to C / N measurement. In this example, the carrier frequency is located at the center of the left half of the display screen, and the frequencies at both ends are vertically separated by an offset value f _of from the carrier frequency to display each spectrum between them, that is, the spectrum waveform ( The frequency domain) is displayed, and the noise level of the frequency f _N , which is apart from the carrier frequency by the offset value f _of , is displayed in the time domain on the right half of the display screen.

That is, the spectrum waveform display in the frequency region of the left half is at least f _c ± f for the input signal.
_A necessary part is taken out from the data taken in the buffer memory 29 by frequency sweeping between “of” and displayed, and the oscillation frequency of the local oscillator 18 is taken in by the signal frequency f _N.
Fixed so that = f _c + f _of, and the buffer memory 2
Read the data captured in 9, and in the right half on the display screen, at the vertical axis position according to the level size,
The horizontal axis is sequentially shifted to the right for each data and displayed, and when the display reaches the right end, it is displayed again from the left end. The data acquisition display for the left part display and the data acquisition display for the right part display are alternately performed.

In this case, the bandwidth of the low pass filter 27
VBW is almost equal to the bandwidth RBW of the bandpass filter 19.
Made equal. When making such measurements,
Automatically determines various parameters of the system analyzer, and
It can also be set automatically. Toes
The center frequency is the spectrum waveform (carrier wave part).
Set the carrier frequency f_CAnd the noise
The center frequency of the time axis display is f_N= F _c+ F_ofTosa
It is. These determined center frequencies are stored as parameter values.
It is held by the section 36.

The frequency span is set to 2f _of when the carrier wave portion is captured, and is set to zero when the noise is displayed on the time axis. The pass band width RBW of the band pass filters 19 and 24 is determined as follows. The RBW is determined so that the bottom of the spectrum of the measurement carrier wave portion does not affect the noise at the measurement frequency f _N. This is uniquely determined from FIG. 2A showing the relationship characteristic between the dynamic range with RBW as a parameter and the offset value f _of . That is, FIG. 2A shows the filter 19,
Due to the influence of 24, the bottom of the spectrum of the carrier wave portion represents a limit where measurement noise is not applied. The vertical axis represents the noise level when the peak value of the carrier wave is 0 dB. Therefore, the absolute value of this vertical axis is the dynamic range. Given the dynamic range and offset value f _of ,
The smaller bandwidth RBW of the characteristic curve near this point is selected. For example, the noise level is -100 dBc / Hz,
That is, the dynamic range is 100 dB and the offset value f
_{When of} is 40 kHz, the point becomes point A in the figure, and the band width 3 kHz of the characteristic curve on the left side of this point A in the figure is RBW. If the RBW is smaller than 3 kHz, the problem of spectrum overlap does not occur, but the frequency sweep time (measurement time) becomes long. Therefore, the bandwidth of the curve on the left side as close as possible to the point A is set.

Further, the dynamic range is determined as follows. When the dynamic range is increased, that is, when the noise level is decreased, distortion and spurious are generated due to the influence of the frequency mixers 17 and 22 inside the spectrum analyzer. Normally, the limit of the maximum value (minimum value of noise level) of the dynamic range that does not cause such distortion and spurious is obtained by measurement in advance, and a margin is added to the limit value and stored in the spectrum analyzer.
This is given, for example, by the dashed curve in Figure 2A. The measurable range is the range above this dashed curve in FIG. 2A. For example, if the offset value is 40 kHz, 40
The dynamic range is smaller than the value (point B) 113 dB on the broken line curve at kHz, that is, the noise level is -11.
It is necessary to set it to 3 dBc / Hz or more.

When the RBW is large, the noise level becomes large and the dynamic range becomes small. When the noise generated in the spectrum analyzer is N _S and the attenuation amount of the input attenuator 16 is ATT, the noise level NL can be expressed by the following equation. _{NL = N S +10 log (RBW} ) + and detection limit of the maximum signal level that can be ATT input and SL is a SL-NL. This limit line is shown by a dotted line in the noise level-offset value characteristic diagram of FIG. 2B. If the absolute value of the dynamic range is larger than this, the signal is distorted or the peak protrudes from the display surface. Thus the dynamic range DR becomes _{DR <SL-N S -10 log} (RBW) -ATT. From this, the set value of RBW and the settable range of the dynamic range (however, the vertical axis represents the noise level value which is the inverted value of the dynamic range) are shown in FIG. 2B. When the internally generated noise level N _S increases, the lower limit value increases, so the dotted line of the noise level moves upward, and when N _S decreases, the lower limit value decreases, and the dotted line of the noise level moves downward. The range in which the dynamic range can be set for the offset value f _of is the range shaded in FIG. 2B, and the dynamic range can be set. If the dynamic range is set in this range, the value of RBW can be uniquely obtained.

The RBW value obtained as described above is also
It is held in the parameter value storage unit 36. Low pass filter
(Video filter) Bandwidth VBW of 27 is equal to RBW
To get rid of To set the sweep time, use the formula (1) for min (RBW, V
RBW as BW) and T _S= S_pan/ {(RB
W)²× 0.5} (sec).

The reference level is set so that the carrier peak value matches the upper end of the display screen. In this way, various parameters are determined and temporarily stored in the parameter value storage unit 36, and these are set by the control unit 31 with respect to the corresponding units, and the measurement for obtaining the display of FIG. 1A described above is performed. . The present invention can also be applied to S / N measurement. That is, S
In the A / N measurement, conventionally, as shown in FIG. 1C, the level of each spectrum of the signal portion (frequency f _{SL to} f _SH ) is read from the buffer memory 29, all of them are added, and the added value is obtained. Is divided by the number of data to be the level L _{S of the} signal part, and from the lowest frequency f _NL of the noise part to the lowest frequency f _SL of the signal part and the highest frequency f of the signal part.
The levels of all spectra of the highest frequency f _NH of the noise part from _SH are added, and the added value is divided by the number of data. In order to make this division result G _S the size of the signal band, the following calculation is performed to obtain the level L _{s of the} signal portion.

L _s = 10 log G _S +10 log ((f _SH −f _SL ) /
(1.2 × RBW) +2.5 RBW is multiplied by 1.2 by the bandpass filters 19 and 2
Correction based on 4 being a Gaussian characteristic, 2.5
Is added in order to correct the level drop due to logarithmic amplification performed in the previous stage of the detector 26. These ratio L _S / L
_N is calculated and displayed as the measurement S / N.

In the present invention, as shown in FIG. 1D, the signal portion is displayed on the left half of the display screen as a spectral waveform, and only specific noise, for example, the component of frequency f _NH is received, and the time elapses in the same manner as described above. Is displayed on the right half of the display screen. In this case, the frequency that becomes the peak value of the signal part and f _NL
Alternatively, the frequency span, RBW, VBW, etc. are determined as described above, using the difference from f _NH as the offset value f _of . The noise level L _{N in} this case is obtained as follows. That is, all level values (linear values) of data at each point displayed on the right half of the display screen are added, and the added value is divided by the number of data, and the noise of this division result G _N1 is uniform in the noise part. The noise level L _N is calculated by the following equation assuming that the noise level L _N is

L _N = 10 log G _N1 +10 log ((f _SL −f _NL +
f _NH −f _SH ) / RBW)) + 2.5 From this L _N and the above L _s , L _s / L _N is measured S / N. The present invention can be applied to C / N measurement of burst waves and S / N measurement. In this case, the sweep frequency is fixed to capture the center frequency component of the input signal, the signal is captured, the captured data is read from the buffer memory 29, and a trigger signal (in synchronization with the input burst signal as shown in FIG. 3B) is created, a burst waveform is displayed in the time domain on the display screen using this trigger signal as shown in FIG. 3D, and the range 41 to be measured is specified by a marker for the displayed waveform. Corresponding gate signal 42 is generated as shown in FIG. 3C. The repetition period T _r of the input burst signal and the pulse width T _W are known to the user in advance. The measured carrier frequency f _C or the frequencies f _SL and f _SH indicating the signal portion are also known and set and input, and various parameters are determined in the same manner as described above, but the bandwidth VBW of the low pass filter 27 is 1 / T _G (T _G is the pulse width of the gate signal 42).

[0019]

As described above, according to the present invention, noise measurement and display are performed in the time domain, and the data are added to obtain an average value. Further, since the time domain display is performed, the noise level changes. , The average level is also easy to understand intuitively, and there is no need to integrate the low-pass filter 27 as in the conventional case so that the display of the noise level does not change so much. Pass filter 2
The bandwidth VBW of 7 can be matched with RBW, and the measurement time (sweep time) is remarkably shortened to 1/10 of the conventional one.

In the case of burst wave measurement, the measurement time cannot be shortened, but since the time domain display is used, the fluctuation state of noise can be clearly understood. In either case, the spectrum of the signal portion or the vicinity of the carrier signal is also displayed as a waveform at the same time, and the same observation as in the past can be performed.

[Brief description of drawings]

1A is a diagram showing a display example in which the present invention is applied to C / N measurement, B is a diagram showing a display example of a conventional C / N measurement, and C is a display example of a conventional S / N measurement. FIG. 3D is a diagram showing a display example in which the present invention is applied to S / N measurement.

FIG. 2 is a diagram showing a relationship characteristic between a noise level or a dynamic range and an offset value with a bandwidth of a bandpass filter as a parameter.

3A is a diagram showing a waveform example of a burst signal, B is a diagram showing a trigger synchronized with the burst signal, C is a diagram showing a gate signal for extracting a part of the burst signal, and FIG.
FIG. 6 is a diagram showing an example of time domain display of a burst signal.

FIG. 4 is a block diagram showing a configuration example of a spectrum analyzer.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Osamu Aoyama 1-32-1, Asahimachi, Nerima-ku, Tokyo Within Advantest, Inc. In Stock Company Advantest (72) Inventor Yoshiaki Miyamae 1-32-1 Asahi-cho, Nerima-ku, Tokyo Stock Company Advantest (72) Inventor Toshiharu Kasahara 1-32-1 Asahi-cho, Nerima-ku, Tokyo Stock Association Company Advantest (72) Inventor Hiroaki Takaoku 1-32-1, Asahi-cho, Nerima-ku, Tokyo Stock company Advantest

Claims (3)

[Claims] 1. A spectral waveform of an input signal is displayed on one half of the display screen, and the other half of the display screen is displayed.
A method for measuring with a spectrum analyzer, characterized in that the noise of the frequency related to the input signal is displayed in the time domain. 2. The spectrum according to claim 1, wherein the bandwidth VBW of the low-pass filter for passing the detection output of the spectrum analyzer is made substantially equal to the bandwidth RBW of the band-pass filter for extracting the frequency converted signal. Measuring method with an analyzer. 3. The settable bandwidth of the band pass filter is used as a parameter so that the relational characteristics of the center frequency of the measurement input signal, the frequency interval of the frequency of the measurement noise, and the noise level or the dynamic range of the input signal are satisfied. 3. The method for measuring with a spectrum analyzer according to claim 1, wherein the RBW is determined from the difference between the center frequency of the input signal and the frequency of the noise.