Electrical Noise - PowerPoint PPT Presentation

About This Presentation
Title:

Electrical Noise

Description:

Flicker noise. Flicker noise is due to contamination and crystal defects. ... flicker noise. Metal film resistors have no flicker ... Flicker noise modeling ... – PowerPoint PPT presentation

Number of Views:390
Avg rating:3.0/5.0
Slides: 25
Provided by: CSUS5
Learn more at: https://www.csus.edu
Category:

less

Transcript and Presenter's Notes

Title: Electrical Noise


1
Electrical Noise
  • Wang C. Ng

2
Nature of electrical noise
  • Noise is caused by the small current and voltage
    fluctuations that are generated internally.
  • Noise is basically due to the discrete nature of
    electrical charges.
  • Externally generated noise is not considered here.

3
Why study noise?
  • It sets the lower limit for the detectable
    signals.
  • It sets the upper limit for system gains.
  • Develop mathematical models to take the effects
    of noise into account when analyzing electrical
    circuits/systems.
  • Find ways to reduce noise.

4
Thermal noise
  • Due to random motion of electrons.
  • It is ubiquitous (resistors, speakers,
    microphones, antennas, )
  • It is directly proportional to absolute
    temperature.
  • White noise - Frequency independent up to 1013 Hz.

5
Thermal noise modeling
  • The noise amplitude is represented by the rms
    value

6
Thermal noise modeling
  • The rms noise voltage for a 1-KW resistor is
    about 4 nV/Hz1/2.
  • The amplitude distribution is Gaussian with m 0
    and s vn .
  • A series voltage source (vn) can be added to a
    resistor to account for the thermal noise.

7
Thermal noise modeling
  • Examples
  • A 1-KW resistor in a system with a bandwidth of
    100 MHz generates about 40 mV of noise voltage.
  • A 1-MW resistor in this system generates about 40
    mV of noise voltage.
  • 10 1-MW resistor in this system generates about
    0.4 V of noise voltage.

8
Shot noise
  • Shot noise is due to the random arrivals of
    electron packets at the potential barrier of
    forward biased P/N junctions.
  • It is always associated the a dc current flow in
    diodes and BJTs.
  • It is frequency independent (white noise) well
    into the GHz region.

9
Shot noise modeling
  • The noise amplitude is represented by the rms
    value

10
Shot noise modeling
  • The rms noise current for a diode current of 1 mA
    is about 20 pA/Hz1/2.
  • The amplitude distribution is Gaussian with m
    ID and s in .
  • A parallel current source (in) can be added to a
    diode to account for the shot noise.

11
Shot noise modeling
  • Examples
  • For a diode current of 1 mA in a bandwidth of 1
    MHz shot noise generates about 20 nA of noise
    current.
  • For a diode current of 10 mA in a bandwidth of
    100 MHz shot noise generates about 2 mA of noise
    current.
  • 100 diodes would generate .2 mA of noise current.

12
Flicker noise
  • Flicker noise is due to contamination and crystal
    defects.
  • It is found in all active devices.
  • It is inversely proportional to frequency (also
    called 1/f noise) .
  • DC current in carbon resistors cause flicker
    noise.
  • Metal film resistors have no flicker noise.

13
Flicker noise modeling
  • The noise amplitude is represented by the rms
    value

14
Flicker noise modeling
  • The constant K1 is device dependent and must be
    determined experimentally.
  • The amplitude distribution is non-Gaussian.
  • It is often the dominating noise factor in the
    low-frequency region.
  • It can be described in more details with fractal
    theory.

15
Other noise types
  • Burst noise (popcorn noise)

16
System Noise Analysis
  • Wang Ng

17
Introduction
  • Noise sources can be added to a device models to
    represent the effect of noise.
  • We need a means to characterize the noise
    performance of a system (black box).
  • Noise figure
  • Noise temperature

18
Noise figure
  • Used for resistive source impedance.
  • Most communication systems have a 50-W source
    impedance (Thevenin equivalent).
  • Signal-to-noise (S/N) ratio
  • Noise figure F (S/N)in / (S/N)out
  • F is a direct measure of the S/N ratio
    degradation caused by the system.

19
Noise figure calculations
  • For an ideal (noiseless) amplifier
  • Sout G Sin
  • Nout G Nin
  • For a real system
  • F (Sin/Nin)(Nout/Sout) Nout/GNin
  • or F (Total noise)/(Noise due to input)
  • F in in general frequency dependent.

20
System noise
  • Internally generated noise can be computed from
  • Nsys (F - 1)GNin
  • since Nout Nsys GNin

21
Cascade systems
  • Gain Gtotal G1 G2 GN
  • Noise figure
  • Ftotal F1 (F2 - 1)/G1 (F3 - 1)/G1G2
    (FN - 1)/G1G2 GN
  • What does this tell us?
  • We should pay most attention to the reduce the
    noise of the first system (Why???)

22
Noise temperature
  • It is the temperature at which the noise
    generated from the source resistance equals to
    the system noise.
  • The noise temperature of a system is a better
    measure when F is close to 1 (low-noise system)
  • Noise temperature Tn T(F-1)

23
Radiometer
  • A modern radiometer can measure noise temperature
    variation down to 100th or even less in ?K.
  • This instrument can be used for remote
    sensing/imaging.
  • Possible extra credit presentation.

24
Summary
  • System noise measure Noise figure and noise
    temperature
  • Internal noise calculation
  • Cascade system noise
  • First stage noise
Write a Comment
User Comments (0)
About PowerShow.com