Ch.

5
Signals and Noise

Categories of signals
Analog signala signal that has a continuous variation with respect to time rather than a pulsed or discrete nature. Digital signal (DS)(or it is a binary signal) a signal in which discrete steps are used to represent information. Digital signals are special cases of analog signals. Therefore, you can use analog techniques to measure and generate digital signals.
2

Types of analog signals

Analog DC signal static or varies slowly with time. We can use analog-to-digital converter (ADC) to convert the analog signal into a digital number that the computer interprets. Examples of DC signals are temperature, pressure and battery voltage, etc. Analog AC time domain signal varies periodically. The signals convey information in the level of the signal and how this level varies with time. We are interested in some characteristics of the shape of the waveform, e.g. slope, locations and shapes of peaks, etc.
3

Converting domains
To look at an analog signal, you can convert the waveform data to the frequency domain. You can use Fast Fourier Transform (FFT) to convert the time domain to the frequency domain on a digital computer. Conversely, you can change the frequency domain to time domain by Inverse Fast Fourier Transform (IFFT). Since many useful parameters are not apparent in the time domain waveform, we need to use FFT to transform to frequency domain. In LabVIEW digital signal processing (DSP) function library, you can use DSP functions to observe the frequencies that make up a signal and the distribution of noise.
4

Information about signals

Using the roadmap in the next slide, you can organize your information about each signal to logically design your LabVIEW system. The map shows the classification, typical interface hardware, processing requirements and display techniques.

Information about signals

This road map can help you organize your info. About each signal to logically design your LabVIEW sys.

Information about signals

You need to characterize each signal to properly determine the kind of I/O hardware youll need. With the help of numerical methods, you need to think about the signal attributes you need to measure or generate. Then, you can think about the user interface, i.e. LabVIEW controls and indicators. You may have data storage requirements. All this items is directly affected by the signal characteristics.
7

Signal conditioning
There are many sources of noise in the environment. So, when doing experiments, you must separate the good signals from the bad ones. To achieve this, you need to do signal conditioning. Signal conditioning accessories amplify low-level signals and then isolate and filter them for more accurate measurements. You can put a lowpass filter on the signal and then make sure that everything is properly grounded and shielded.
8

Signal conditioning
To design the right signal conditioning approach for your application, there are several steps:
1. Know all about your sensors and what kind of signals they are supposed to produce. 2. Consider grounding and shielding. 3. May need amplifiers and filters. 4. List your specifications and by the right data acquisition (DAQ) hardware.
9

Noise Sources
Internal circuit noisesshot noise, low frequency (excess) noise, but most common is thermal noise. External noises & interferencea.c. power circuits, switching transients, temperature changes and vibrations, etc.

10

Noise SourcesInternal noise

Thermal noise:
Caused by the random, thermally induced, motion of conduction electrons. It is also called Johnson or Nyquist noise.

Shot noise:
Results from the random passage of individual charge carriers across a potential barrier. Reduce the number of connection parts to reduce the noise.
(Electronic Noise And Low Noise Design, Peter J. Fish, 1993)
11

Noise SourcesInternal noise

Low frequency (excess) noise:
When current passes through a resistor or a semiconductor, noise is generated in excess of the thermal noise in a resistor or the thermal noise plus shot noise in a semiconductor. This excess noise has a spectral density which increases as the frequency decreases. It is most noticeable at low frequency.
(Electronic Noise And Low Noise Design, Peter J. Fish, 1993)
12

Grounding and shielding

You can eliminate most noise-related problems right at the sourcethe wiring. The way of connecting a sensor to its associated DAQ hardware greatly affects the overall performance of the system. However, the wire between sensor and data acquisition can never improve the signal quality. So, you can do grounding and shielding.
13

Ground
Ground does not necessarily mean the Earth ground. It is preferred to say it is a safety ground. You should always make sure that there is a reliable path from all of your equipment to safety ground as required by code. This prevents accidental connections between power sources and metallic objects from becoming hazards. Such fault currents are shunted away by the safety ground system.
14

Ground
When you are measuring voltage of signal, you need to know about a reference potential. Earth ground refers to the potential of the earth. Sometimes, it is called system ground. Reference ground, sometimes called a return path or signal common. It is not necessarily wired to earth ground. Any reference conductor can be used for a common return. It provides a low-impedance path for currents to return to source equipotential.
15

Ground

Symbols of different Grounds

16

Ground

Taking the sys. approach to grounding in a lab. Note the use of a signal common (in the form of heavy copper cable) to tie everything together. All items are connected to this signal common.
17

Ground
If there is large resistance during the measurement procedure, the measured voltage will be lower than expected. Then, it may cause a big error. Metallic instrument may produce self-inductance, which gives out large resistance, error becomes significant. A well-designed signal common can be effective at higher frequencies where second-order effects like skin effect and self-inductance become important.
18

ShieldElectromagnetic fields
Your measurement system may introduce some noise by electromagnetic fields. There are a few principles of electromagnetism that you use when connecting your DAQ system:
1. 2 conductors that are separated by an insulator form a capacitor. An E-field exists between the conductors. This is called capacitive coupling and is one way for coupling noise into a circuit. So, moving things apart reduces capacitive coupling.
19

ShieldElectromagnetic fields
2. An E-field cannot enter a closed and conductive surface. This kind of enclosure is called a Faraday cage, or electrostatic shield, and is commonly implemented by a sheet of metal or screen surrounding a sensitive circuit. Electrostatic shields reduce capacitive coupling. 3. A varying B-field will induce a current in a closed, conductive loop. This phenomenon is called inductive coupling, or inductance. 20 The magnetic of the induced current is

ShieldElectromagnetic fields
4. Magnetic shielding is not easy to design because the B-fields that we are most concerned about are very penetrating and require very thick shields of iron or even better magnetic materials.

21

ShieldElectromagnetic fields
Put sensitive, high-impedance circuitry and connections inside a metallic shield that is connected to the common-mode voltage of the signal source. This can block capacitive coupling to the circuit and the entry of stray E-fields. Avoid closed and conductive loops, which are known as ground loops, to prevent influence of stray B-fields. Avoid placing sensitive circuits near sources of intense magnetic fields, such as transformers, motors, and power supplies.
22

ShieldRadio-frequency interference (RFI)

There will be Radio-frequency interference (RFI) when a moderately intense RF source is present nearby. Common sources of RFI are transmitting devices, e.g. cellular phones. Radio frequencies radiate for great distances through most nonmetallic structures, and really high (microwave) frequencies can penetrate through cracks in metal enclosures. When RFI is present in your DAQ system, it results in unexplained noise of varying amplitude and frequency.
23

ShieldRadio-frequency interference (RFI)

Solutions to RFI are as follows:
Shield all the cables into and out of your equipment. Add RF-rejection filters on all signal and power leads. Put equipment in well-shielded enclosures and racks. Keep known RF sources and cables far away from sensitive equipment.
24

Thermoelectric EffectSeebeck Effect

In 1822, an Estonian physician named Thomas Seebeck discovered (accidentally) that the junction between two metals generates a voltage which is a function of temperature. (http://www.lanzo.se/applications/thermocouple.html) Thermojunction voltage (Seebeck voltage) is generated any time 2 dissimilar metals come in contact with one another in the presence of a temperature gradient.
25

Seebeck Effect

(http://users.telenet.be/educypedia/electronics/thermoelectric.htm)
26

Thermocouples
Thermocouples are the most popular temperature sensors. They are cheap, interchangeable, have standard connectors and can measure a wide range of temperatures. The temperature range depends on the material used.
(http://www.lanzo.se/applications/thermocouple.html)
27

Other error sources Thermojunction voltage

Problems occur in DAQ when you attempt to measure DC signals that are in the microvolt to millivolt range, e.g. those from thermocouples and strain gauges. If you connect your instruments with wires, connectors, and terminal screws made of differential metals or alloys, then there is a risk of adding uncontrolled thermojunction voltages to the signals.
28

Other error sources thermojunction

Ways to kill the parasitic thermojunction voltage:
Make all connections with the same metallic alloy as the wires they connect. Keep all connections at the same temperature. Minimize the number of connections in all lowlevel signal situations.

29

Thermoelectric EffectPeltier Effect

The Peltier effect is the reverse of the Seebeck effect; a creation of a heat difference from an electric voltage. It occurs when a current is passed through two dissimilar metals or semiconductors (n-type and p-type) that are connected to each other at two junctions (Peltier junctions). The current drives a transfer of heat from one junction to the other: one junction cools off while the other heats up. This effect was observed 13 years after Seebeck's initial discovery in 1834 by Jean Peltier. The effect is stronger in semiconductors.
(http://en.wikipedia.org/wiki/Peltier_effect)

30

Thermoelectric EffectPeltier Effect

When a current I is made to flow through the circuit, heat is generated at the upper junction (at T2), and absorbed at the lower junction (at T1). The Peltier heat absorbed by the lower junction per unit time, Q is equal to

Q = AB I = ( B A ) I

Where is the Peltier coefficient AB of the entire thermocouple, and A and B are the coefficients of each material. Ptype silicon typically has a positive Peltier coefficient (though not above ~550 K), and n-type silicon is typically negative. (http://en.wikipedia.org/wiki/Peltier_effect)
31

Thermoelectric EffectPeltier Effect

The conductors are attempting to return to the electron equilibrium that existed before the current was applied by absorbing energy at one connector and releasing it at the other. The individual couples can be connected in series to enhance the effect. The direction of heat transfer is controlled by the polarity of the current, reversing the polarity will change the direction of transfer and thus the sign of the heat absorbed/evolved. A Peltier cooler/heater or thermoelectric heat pump is a solid-state active heat pump which transfers heat from one side of the device to the other. Peltier coolers are also called TECs (Thermo Electric Converter).
(http://en.wikipedia.org/wiki/Peltier_effect)
32

Instrumentation amplifier
An instrumentation amplifier is a special kind of circuit whose output voltage with respect to ground is proportional to the difference between the voltages at its 2 inputs.

33

Instrumentation amplifier

Representation of the common-mode and differential components of the input signal to a difference amplifier. Note that V1=VCM-Vd/2 and V2=VCM+Vd/2.
34

Multiplexers
Multiplexer is an array of switching elements (relays or solid-state analog switches) that route many input signals to one common output. Many analog inputs are connected to one multiplexer.

35

Single-ended connections
Single-ended connections are the simplest and most obvious way to connect a signal source to an amplifier or other measurement device.

36

Single-ended connections

A single-ended amplifier has no intrinsic noise rejection properties. You need to carefully shield signal cables and make sure that the signal common is noise-free as well.
37

Single-ended connections
Single-ended connections are most often used in wide-bandwidth systems, e.g. oscilloscopes, video and RF, etc. Noise induced on any of the input wires, including the signal common, is added to the desired signal.

38

Differential connections
In a differential measurement system, neither input is connected to a fixed reference such as earth or building ground. In the next slide, it shows a differential measurement system. A pair of multiplexers (MUX) are connected to an instrumentation amplifier. It assumes that noise pickup usually occurs equally on 2 MUX that are closely spaced, e.g. a twisted pair of wires. You only need to take the difference between the 2 voltages, the noise cancels but difference signal remains.
39

Differential and single-ended connections

A differential measurement system

40

Use amplifiers and other signal conditioning

Even if the grounding and shielding are properly done, you should consider signal conditioning, which includes amplifiers and filters, among other things to reduce the noise relative to the signal level. Amplifiers boost the amplitude of smaller signals, improving the resolution of the measurements. They provide differential inputs, a technique to help reject noise.
41

Use amplifiers and other signal conditioning

In general, you should try to amplify your lowlevel signal as close to the physical phenomenon itself as possible. Then, the signal-to-noise ratio (SNR) can be increased. SNR is:
SNR = 20 log( Vsig Vnoise )

Where Vsig is the signal amplitude and Vnoise is the noise amplitude, both measured in volts rms
42

Use amplifiers and other signal conditioning

The 20 log() operation converts the simple ratio to decibels (dB), a ratiometric system used in signal processing. E.g. an SNR of 20 dB is the same as a ratio of 10 to 1.

43

2 broad categories of Signals

Ground-referenced analog inputsthe voltage signals are referenced to a system ground. You should use a differential input to avoid ground loops and provide common-mode noise rejection. Floating analog inputsbattery-powered equipment and instruments with isolated outputs do not supply a return connection to signal ground.
44

2 broad categories of Signals

Ground-referenced analog input

Floating analog input

45

Triggering
Triggering refers to any method by which you initiate, terminate, or synchronize a DAQ event. A trigger is usually a digital or analog signal whose condition is analyzed to determine the course of action.

46

Software triggering
You can control the trigger directly from the software. E.g. Using a Boolean front panel control to start or stop DAQ.

47

Hardware triggering
Internal triggersignals are generated within the hardwares. External triggersignals are generated by specialized hardwares.

48

Little noise is good?

When performing A/D or D/A conversion, there will be errors in the Least Significant Bit (LSB). If you add some noises to the signals, errors could be lowered.

49

Graphs A and B represent digitization of an ideal analog ramp of 4-LSB amplitude which results in objectionable quantization steps. Adding 1 LSB peak-peak dither noise and lowpass filtering (graphs C-E) improves results.

50