JPH04220570A - Fft analyzer - Google Patents

Abstract

PURPOSE:To achieve frequency zooming with a small amount of hardware. CONSTITUTION:A band-pass filter bank 1 is cascade tree structure which is constituted by a low-pass filter(LPF), a high-pass filter(HPF), and a decimation means of 2 of data 102, etc., in combination are provided, a signal transmission path selection means 301, etc., where a branch point is provided is controlled by a selection control signal ZS, and a passing route of signal is determined, thus enabling zooming of frequency to be made.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an FFT analyzer (including a digital spectrum analyzer), and more particularly to a technique for reducing the amount of hardware in a frequency zooming section (shift calculation section).

0002

2. Description of the Related Art FIG. 7 shows the basic configuration of a conventional FFT analyzer.

[0003] This FFT analyzer includes two local oscillators 200 and 201, two multipliers 202,
203, two digital filters 204, 205 for preventing aliasing, and data thinning (resampling) means 206, 207 for reducing the amount of data.
The frequency zooming is performed by multiplying the input waveform data X(n) by the sine wave and cosine wave generated from the local oscillators 200, 201 using the multipliers 202, 203, and shifting the frequency. It will be done.

0004

[Problem to be solved by the invention] The above-mentioned conventional FFT
Since the analyzer uses two local oscillators, two multipliers, and two digital filters for its frequency zooming function, it requires a large amount of hardware, leading to problems such as an increase in chip area when integrated into an IC.

The present invention has been made in view of the problems of the prior art, and its purpose is to realize the frequency zooming function of an FFT analyzer with a smaller amount of hardware.

[0006]

[Means for Solving the Problems] The FFT analyzer of the present invention includes a digital filter that performs frequency zooming, and an FFT processor that performs fast Fourier transform analysis on an output signal from the digital filter, The filter has a cascade tree structure configured by combining a low-pass filter, a high-pass filter, and data thinning means, and a transmission path for selecting a signal transmission path is provided at a branch point of the signal transmission path in this cascade tree. The present invention is characterized in that a selection means is provided, and the transmission path selection means is controlled by a selection control signal.

[0007]

[Operation] The low-pass filter (LPF) and high-pass filter (HPF) that make up the cascade tree are, for example,
It is a half-band filter (a filter with a cut-off angle frequency of approximately π/2), and by selecting a combination of these filters, you can essentially create multiple band-pass filters (band-pass filter bank) with different pass frequency bands. Can be configured. Therefore, by appropriately selecting and switching the transmission path of input waveform data using the selection control signal, signals in a desired frequency range can be selectively extracted.

[0008] A digital filter can be substantially composed of a delay element and an addition element, and an LPF and an HPF can be constructed by a slight modification in which signals having the same absolute value are added with their polarities inverted. That is, if a is a coefficient and x is data, the LPF is a0 x0 +a1 x1 +a2
If it is expressed as x2 +a3 x3, the HPF is a0 x0 -a1x1 +a2 x2 -a3 x3
It is sufficient to invert the polarity of every other term, and this can be easily achieved by using two's complement. Therefore, the above-mentioned bandpass filter bank can be realized by multiplexing one digital filter and passing the signal many times while considering the polarity of the signal.
There is no need to separately prepare two digital filters, one for imaginary numbers and one for real numbers. Furthermore, the calculations performed by the FFT processor do not require a complex FFT as in the prior art, but only a real FFT, which simplifies the configuration of the processor. Also,
The means for selecting the signal transmission path can be constructed by a simple switch, and the increase in the number of elements can be almost ignored. Therefore, the frequency zooming function can be realized with approximately 1/2 the amount of hardware compared to the conventional technology.

[0009]

Embodiments Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a diagram showing the configuration of an embodiment of the present invention.

The FFT analyzer 4 of this embodiment has a bandpass filter bank 1 and an FFT processor 2, and receives waveform data (time axis signal) from the A/D converter 3 and processes it by zooming. After narrowing down to a desired frequency range, waveform analysis is performed using FFT. The zooming area is designated by a control signal ZS applied from the outside.

FIG. 2 is a diagram showing the configuration of bandpass filter bank 1.

As shown in the figure, the bandpass filter bank 1 includes low-pass filters (LPFs) 101 and 103.
, 105, 111, 123, 125, 131 and high pass filter (HPF) 121, 107, 109, 11
3, 127, 129, 133 and data 2 thinning means 102, 104, 106, 108, 10112, 11
4,122,124,126,128,130,132
, 134 and signal transmission path selection means 300, 301,
It has a cascade tree structure in which 302, 303, 304, 305, and 306 are combined. The data 2 thinning means doubles the data acquisition interval and halves the sampling data amount. Further, the control signal ZS for specifying the zooming area mentioned above is given to the signal transmission path selection means, thereby switching an internal switch (not shown).

In this embodiment, 14 bands (passage frequency band, band00
,01,10,11,12,13,20,21,22,
23, 24, 25, 26, 27) are formed. Furthermore, the characteristics of the LPF and HPF are half-band filters, as shown in FIGS. 3(a) and 3(b), respectively.

FIG. 4 is a diagram showing a hierarchical structure of selectable pass frequency bands (bands) in this embodiment.

For example, when band 00 is selected, one of bands 10 and 11 can be further selected. For example, in the case of band 10, one of bands 20 and 21 can be further selected. In this way, the passband is appropriately narrowed down to obtain a signal in the desired band. For example, when band 11 is selected, the input waveform signal passes through the route indicated by the arrow in FIG. 5, and when band 26 is selected, the input waveform signal passes through the route indicated by the arrow in FIG. 6.

The output of the digital filter is (for example, 20
After outputting 48 points), FFT processing is performed by the FFT processor 2 at the subsequent stage, but in this case, only real number FFT is required for calculation, simplifying the configuration of the processor.

[0018]

Effects of the Invention As explained above, the present invention configures a bandpass filter bank using one digital filter, and unifies the conventional two-system zooming of imaginary numbers and real numbers to zooming of only real numbers, thereby making it possible to improve the FFT processor. Only one is required, which has the effect of making it possible to configure an FFT analyzer (including a digital spectrum analyzer) with a much smaller amount of hardware (approximately 1/2) than in the past.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a diagram showing the configuration of a bandpass filter bank 1.

FIG. 3 shows the LPF used in the embodiment of FIG. 2, respectively;
2A and 2B are diagrams showing the characteristics of the HPF, in which (a) shows the characteristics of the LPF, and (b) shows the characteristics of the HPF.

FIG. 4 is a diagram showing a hierarchical structure of selectable pass frequency bands (bands) in the embodiment of FIG. 2;

FIG. 5 is a diagram showing a signal passage route when band 11 is selected.

FIG. 6 is a diagram showing a signal passage route when band 27 is selected.

FIG. 7 is a diagram showing the basic configuration of a conventional FFT analyzer.

[Explanation of symbols]

1 Bandpass filter bank 2 FFT processor 3 A/D converter 4 FFT analyzer 101, 103, 105, 111, 123, 125, 1
31 Low pass filter (LPF) 121, 107, 109, 113, 127, 129, 1
33 High pass filter (HPF) 102, 104, 106, 108, 10112, 114
,122,124,126,128,130,132,
134 Data 2 thinning means

Claims (1)

[Claims] 1. An FFT analyzer that analyzes waveform data using a fast Fourier transform method, which includes a digital filter (1) that performs frequency zooming, and a fast Fourier transform method for the output signal from the digital filter (1). The digital filter (1) has a cascade tree structure composed of a combination of a low-pass filter, a high-pass filter, and a data thinning means, A transmission path selection means (300 to 306) for selecting a signal transmission path is provided at a branch point of the signal transmission path within
) is provided, and this transmission route selection means (300 to
306) An FFT analyzer characterized in that it is controlled by a selection control signal.