JPH02104124A - Pulse width modulation system - Google Patents

Abstract

PURPOSE:To maximize S/N at the time of demodulation by specifying the size of pulse width to be changed by modulation in accordance with the period of a repeated pulse signal. CONSTITUTION:An input signal (a) is sampled by a repeated pulse signal (b) with a period T and the pulse width of the repeated pulse signal (b) is changed by the sample value. When it is defined that the size of the pulse width to be changed by modulation is DELTAW and the degree (m) of modulation is m=DELTAW/T, the degree (m) of modulation is proportional tp the amplitude of the input signal. On the other hand, the frequency component (spectrum distribution) of a signal can be found out by Fourier analysis and a repeated pulse signal can be found out by Fourier series expansion. Consequently, the value of DELTAW maximizing the S/N is DELTAW=(2/3)T. When the size of the pulse width to be changed by modulation is set up to 2/3 of the period of a repeated pulse, the S/N at the time of modulation can be maximized.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a pulse width modulation method for transmitting or recording information of an input signal by changing the pulse width of a repetitive pulse. One of the pulse modulation methods that changes the parameters of the repetitive pulse according to the information of the technical input signal is the pulse width modulation method (e.g., communication method, IEICE, PP2).
29-233).

In the pulse width modulation method, an input signal is sampled, and the pulse width of a repetitive pulse is varied depending on the sample value.

Specific methods of modulation and demodulation are described in numerous documents including the above-mentioned document, so they will be omitted here.

Problems to be Solved by the Invention In the pulse width modulation method described above, the signal-to-noise ratio during demodulation depends on the size of the pulse width that changes due to modulation and the band of the transmission line. , not enough attention was paid.

The present invention has been made in view of this point, and it is possible to transmit or record high-quality signals by setting the magnitude of the pulse width that changes for modulation to an optimal value for the band of the transmission path. The purpose of this study is to provide a pulse width modulation method that allows for

Means for Solving the Problems In order to solve the above problem, the present invention provides the following equation: where T is the period of a repetitive pulse signal that samples an input signal, μ is the magnitude of the pulse width that changes depending on the sample value, and ΔW is the period of the repetitive pulse signal that samples the input signal. ΔW=(X)T.

Operation The present invention can transmit or record input signal information with high quality by using the above-described method.

Example In order to explain the method of the present invention, one of the pulse width modulation signal waveforms will be explained.
An example is shown in FIG.

As shown in FIG. 1, in the pulse width modulation method, an input signal a is sampled using a repetitive pulse signal with a period T, and the pulse width of the repetitive pulse signal υ is varied according to the sampled value. Figure a shows a case where the falling edge of the pulse of the repetitive pulse signal S is changed. Here, the magnitude of the pulse width that changes due to modulation is ΔW, and the modulation degree m is m
=ΔW/T (1), then the modulation degree m is proportional to the amplitude of the input signal.

On the other hand, the frequency component (spectral distribution) of the signal is found by Fourier analysis, and the repetitive pulse signal is found by Fourier series expansion.

An example of the calculation results of the spectral distribution when the pulse width is changed for a repetitive pulse signal with period T is shown in Fig. 2 (a) to (d). are (a) where the pulse width γ is the period T of the repetitive pulse signal, respectively.
34-(G)%, (C)%, (4%) Each signal waveform and its spectral distribution are shown. In the graph showing the spectral distribution, the horizontal axis is the frequency, and the vertical axis is the frequency for each frequency component. It is a complex amplitude.As is clear from the figure, in a repetitive pulse signal, high frequency components attenuate rapidly.

Furthermore, as the pulse width becomes narrower, its frequency components spread to higher frequencies. Therefore, for example, if the pulse width r is Z with a repetition period T (c), the spectral components are fo, 2
．． fo, 3. fo (fo=1/T), and becomes almost zero at 4%0 or more. Therefore, the signal (0) is transmitted,
When demodulating, the practical frequency band of the transmission path is 4fO=f
Just oh. This fo is called the cutoff frequency. in general,
If the pulse width r is γ = T/N, the cutoff frequency fc for that pulse signal is fo = 1/r = Nx (1/
T), which is approximately the reciprocal of the pulse width γ.

From this, when the magnitude of the pulse width of the repetitive pulse is changed by ΔW by pulse width modulation, the minimum pulse width is (T-ΔW)/2, so the cutoff frequency f of the transmission path.

is f = 2/(T-ΔW) ・・・・・・・・・
...(2).

Here, since the noise on the transmission path is usually random, the cutoff frequency f0 of the transmission path is equal to the noise band.

From the above, the relative signal power versus noise power during demodulation (
The relative SN ratio is d/fo using the modulation degree m and the cutoff frequency f0 of the transmission path.

According to equations (1) and (2), the relative S/N ratio frI/fo
Calculate 1 and rearrange as follows: f = (TΔW2−ΔW')/2T2...
...(@). (From the Shizuru formula, the relative SN ratio is a function of the cube of the changing pulse width ΔW. As an example, the calculation result when T = 8 ns is 3
As shown in the figure. Note that in FIG. 3, the values are standardized so that the maximum value on the vertical axis is 1.

As shown in Figure 3, as m, that is, the magnitude of the pulse width that changes due to modulation, increases, the value of the relative SNR also increases, but when the magnitude of the pulse width that changes becomes larger than a certain value. The relative signal-to-noise ratio value decreases.

Therefore, in the pulse width modulation method, there is a pulse width size ΔW at which the S/N ratio is maximum for the transmission band fo.

The value of ΔW that maximizes the S/N ratio is obtained by differentiating the equation (@d Crrf/f0)/d (ΔW)=02TΔW−Δ
W2=.

Δ'W=(2/3)T That is, when the magnitude of the pulse width that changes due to modulation (% of the period of the repetitive pulse), the S/N ratio during demodulation becomes maximum.

Effects of the Invention As stated above, by setting the magnitude of the pulse width that changes due to modulation to approximately % of the period of the repetitive pulse signal, the S/N ratio during demodulation is maximized, which is extremely useful in practice. A pulse width modulation method can be provided.

[Brief explanation of drawings]

FIG. 1 shows one example of a pulse width modulation signal for explaining the present invention.
Figure 2 shows an example of the spectral distribution of the modulated signal.
FIG. 3, which is a diagram showing an example, is a diagram showing calculation results of the relative S/N ratio during demodulation taking into consideration the band of the transmission path. Name of agent: Patent attorney Shigetaka Awano and 1 other person 1st
Figure //7 = 2 strings i

Claims (1)

[Claims] When the cycle of a repetitive pulse signal that samples an input signal is T, and the maximum amount of change in the pulse width that changes depending on the sample value is ΔW, the relationship ΔW=(2/3)T holds true. A pulse width modulation method characterized by: