JPS6342966B2 - - Google Patents

Description

[Detailed Description of the Invention] This invention relates to a filter applied to signal waveform processing, etc., and its purpose is to realize a squared cosine filter that matches theoretical values with a simple circuit. be.

Figures 1A and 1B show examples of squared cosine filters conventionally used in television broadcasting equipment or signal generators. Both are lattice-type filters matched to a matching impedance R.

Several stages of low-pass filters are connected in cascade between the input terminal a and the output terminal b.
It forms a raised cosine filter.

However, such a filter can only achieve approximate characteristics and cannot obtain perfect raised cosine characteristics. In addition, in order to get closer to the squared cosine characteristic, the values of the coil and capacitor must be special values, and therefore each coil and capacitor must be adjustable.
Therefore, although the circuit configuration appears simple at first glance,
This method requires time-consuming and difficult adjustments, and has the disadvantage that only approximate characteristics can be achieved.

The present invention provides a filter that eliminates the above-mentioned conventional problems, and the circuit shown in FIG. 2 represents an embodiment of the present invention.

In this figure, a is a signal input terminal, b is an output terminal, and DL ₁ and DL ₂ are delay circuits for delaying the input signal by a certain time T (sec).
W ₁ , W ₂ , W ₃ , and W ₄ are weighting circuits, which function as attenuators in this embodiment. AD is an adder that adds and synthesizes three input signals.

The signal applied to input terminal a is a baseband signal, and first, consider an impulse signal having ideal frequency characteristics band-limited up to frequency ₀ as shown in A in FIG.

This signal and the output signals of the two delay circuits DL ₁ and DL ₂ are respectively sent to weighting circuits.
After being weighted _1/2 , 1, 1/2 by W 1 , W ₂ , and W ₃ , they are added and combined in an adder AD, and then weighted by 1/2 in a weighting circuit W ₄ to be sent to the output terminal. taken out from b.

Transfer characteristic G(ω) between input terminal a and output terminal b
is expressed as the following equation.

G(ω)=1/2{e ^-j 〓 ^T +1/2 (1+e ^-j2 〓 ^T )
} = 1/2 (1 + cosωT)e ^-j 〓 ^T = cos ² (ωT/2)e ^-j 〓 ^T ...(1) Here, the delay time T (sec) is expressed as frequency ₀ and the following formula T = 1/2 ₀ ... (2) When setting the relationship expressed by (2), the amplitude characteristic of G(ω) becomes a characteristic as shown in B in FIG. 3, which takes a value of zero at frequency ₀ . When a signal like A in FIG. 3 is applied to input terminal a, an output signal like C in FIG. 3 is obtained at output terminal b. Furthermore, assuming that the phase characteristic of the input signal is zero, the phase characteristic of the filter of the present invention corresponding to B in FIG. 3 is linear, so the phase characteristic of the output signal is D in FIG. Since it is clearly a linear phase, no group delay distortion occurs due to passing through the circuit of FIG. The dotted line characteristic in the region of ₀ or more at D in FIG. 3 indicates that the signal component has already been lost at G in FIG. 3, so the signal returns to zero phase.

From equation (1) and G in Figure 3, the amplitude characteristic is as follows: |G(ω)|=cos ² (ωT/2)...As shown in (3), the frequency is in the range of 0≦≦ ₀ ...(4) It clearly shows accurate raised cosine characteristics within the range.

Furthermore, when an impulse signal that has already passed through a low-pass filter, as shown in E in Fig. 3, is applied to the circuit in Fig. 2, the amplitude characteristic of the signal observed at the output terminal will be as shown in F in Fig. 3. As shown, B and E in Figure 3
It is expressed as the product of the characteristics of

E in Figure 3 shows an example in which the signal exists in a frequency range greater than ₀ , but as shown in F in Figure 3, in the range 0≦≦ ₀ , the circuit in Figure 2 becomes a squared cosine filter. working as.

Components remaining in the range > ₀ are unnecessary components. In this example, the order is sufficiently negligible, but
Furthermore, in order to suppress this unnecessary component, the delay time T (sec) may be adjusted so that ₀ at F in FIG. 3 becomes a larger value.

In this way, in FIG. 2 configuring the raised cosine filter, the delay circuits DL ₁ and DL ₂ are, for example, BBD,
It can be made using a so-called charge transfer device such as a CCD, a lumped constant delay circuit consisting of a coil and a capacitor, or a distributed constant delay circuit, etc. Also, the weighting circuits W ₁ to _{W 4} are amplifiers whose use can be set. A resistor attenuator or the like can be used, and the adder AD can be configured by a combination of a resistor adder and an amplifier.

With the circuit configuration shown in Figure 2, almost no adjustment can be made, but the area where fine adjustment may be necessary is the weighting settings of weighting circuits W ₁ to W _4. Observe the output waveform of each weighting circuit. However, it is extremely easy to adjust the respective values so that the amplitude ratio is 1:2:1.

Furthermore, although the circuit configuration shown in Figure 2 appears to be more complicated than that shown in Figure 1 at first glance, current IC technology makes it possible to complete the circuit configuration shown in Figure 2 with an extremely small number of parts.

In addition, the circuit of the present invention can not only be used as a filter that requires accurate squared cosine characteristics, but also can be used in general It can also be used as a low-pass filter.

Also, from the perspective of application, the first
It can be used in television signal transmission systems. That is, it can be applied to waveform processing of video signals or teletext signals in broadcasting stations, relay stations, etc., waveform processing in various signal generators, waveform processing in various receiving devices, etc.

A second possible application is waveform processing in various data transmission circuits that handle digital signals and waveform processing in various digital signal application devices.

In addition, as mentioned above, it has a wide range of applications as a simple low-pass filter.

This kind of multi-faceted application is made possible by
The squared cosine filter itself has a wide range of uses, and the reason why it is so valued is because the resulting output waveform has little ringing and no waveform distortion due to phase distortion.

As explained above, the present invention is capable of realizing a filter that matches the theoretical value of a raised cosine filter, and that the squared cosine filter can be realized with a relatively simple circuit configuration, and furthermore, It is possible to make no adjustment, the circuit structure is easy to integrate into an IC, so it can be miniaturized, and it has an extremely wide range of uses.For the above reasons, the filter of this invention can be used very effectively and industrially, making it very effective and useful. It is.

[Brief explanation of the drawing]

1A and 1B are circuit diagrams of a conventional raised cosine filter, FIG. 2 is a circuit diagram of a filter of the present invention, and FIG. 3 is an operational diagram. a...Input terminal, b...Output terminal, DL ₁ , DL ₂
...delay circuit, W ₁ , W ₂ , W ₃ , W ₄ ... weighting circuit, AD ... adder.

Claims (1)

[Claims] 1. A filter that suppresses high frequency components by providing a raised cosine filter for a baseband signal whose highest frequency is band-limited to a predetermined frequency, comprising: As a signal, this first
a first delay circuit that outputs a second signal that is a signal delayed by a certain time T (sec); and a third signal that is delayed this second signal by a certain time T (sec). a second delay circuit; a first delay circuit which is supplied with the first, second, and third signals separately, and which weights the three signals at an amplitude ratio of approximately 1:2:1; It consists of three weighting circuits, second and third, and an addition circuit that is supplied with the output signals of these three weighting circuits and performs additive synthesis, and has a frequency characteristic of 2 from zero frequency to a predetermined frequency.
A filter characterized in that it is a raised cosine filter.