Scribd
Upload
29 views

Chap 4

The document discusses digital conversion from continuous time signals to discrete time signals. It covers sampling, where a continuous time signal is converted to a discrete time signal by taking samples at regular time intervals. It also discusses discrete-to-continuous conversion, where a discrete time signal can be reconstructed to a continuous time signal using various interpolation methods. The document provides examples of sampling sinusoids and discusses oversampling.

Uploaded by

김민성
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
29 views

Chap 4

The document discusses digital conversion from continuous time signals to discrete time signals. It covers sampling, where a continuous time signal is converted to a discrete time signal by taking samples at regular time intervals. It also discusses discrete-to-continuous conversion, where a discrete time signal can be reconstructed to a continuous time signal using various interpolation methods. The document provides examples of sampling sinusoids and discusses oversampling.

Uploaded by

김민성
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

23

Chap. 4 : Digital Conversion

4.1. Sampling : Conversion from Continuous Time(CT) to Discrete Time(DT)

Continuous time signal


x(t) → to → x[n] = {x(nTs )}
Discrete time signal
↑ Ts = 1/fs ( Sampling time=1/sampling freq)

Fig. 1. Conversion from continous time(CT) signal to discrete time (DT) signal
24

∗ Note that

• Time-index : n = 0, ±1, ±2, ±3, . . .


⇒ n is integer and cannot be 0.1, 0.5, 0.7, 1.8, . . . .
⇒ x[0.5], x[0.7], x[0.8] are NOT possible, i.e., NOT definable.

• EX : x(t) = cos(2πt), Ts :Sampling time


when Ts = 1 sec ⇒ x1 [n] cos(2πnTs ) = (1)n = 1
when Ts = 1/2sec ⇒ x2 [n] cos(2πn1/2) = (−1)n
when Ts = 1/3sec ⇒ x3 [n] cos(2πn1/3s)
when Ts = 1/π sec ⇒ x4 [n] cos(2πn1/π) = cos(2n)
⇒ x[n] = x1 [n] + x2 [n] is NOT possible.

Fig. 2. Discrete signal with different sampling times


25

• Sampled continuous signal with various sampling interval Ts

- For the fasten varying signals, the sampling interval should be more de-
creased. in other words, sampling Time(Ts ) should be inversely propor-
tional to signal freq.

Shannon Sampling Theorem (Nyquist Theorem)


A continuous-time signal x(t) with frequencies no higher than fmax can be
reconstructed exactly from its samples x[n] = x(nTs ), if the samples are
1
taken at a rate fs = 1/Ts that is greater than 2fmax , that is, fs = Ts > 2fmax .
26

• Corresponding discrete Signals

* Different sampling times


⇒ Different scales.
⇒ Maps with different scales can not be put into together.
⇒ x1 [n] ± x2 [n] are NOT possible!
27

• Sampling sinusoids

x(t) = A cos(ω0 t + φ)

⇓ Sampling with sampling time Ts

{x(nTs )} = {A cos(ωnTs + φ)}

x[n] = A cos(ω̂n + φ) = Re{Aejφ · ej ω̂n }

Normalized freq. : ω̂ = ωTs = ω/fs


28
29

4.2 Discrete-to-Continuous Conversion

Discrete time signal


y[n] −→ to −→ y(t)
Continuous time signal

∞
X
y(t) = y[n]p(t − nTs ), where p(t) is an interpolator.
n=−∞

• Various interpolators p(t)

⇒ Zero crossing occurs at every sampling point.


30


 1 : |t| < T /2
s
• Zero-order hold p(t) =
 0 : Otherwise


 1 − |t|/T : t
s
• First order linear interpolator p(t) =
0 : Otherwise
31

• Parabolic Interpolator p(t) : 2n d order polynomial with p(t) = 0 for


t = 0, ±Ts , ±2Ts

• Ideal Band limited Interpolator


32

• Oversampling
- Oversampling : sampling freq. is much greather than the Nyquist freq.
- With oversampling, the simples interpolator may reconstruct the continu-
ous signal.
- Ex : CD player, fmax ≈ 20 KHz ⇒ 4 or 3 times oversampling
33
34

You might also like