201:Voltage Measurement Halit Eren Curtin University 01Technology,Perth, WesternAustralia, Australia

1353 1354 1354 1355 1356 1357 1358 1359 1359 1360 1361

1 INTRODUCTION The output signal fonTI to be measuredin the majority of sensorswill require the measurementof either a voltage (this article), a current (Artide 202, CorreDi Measuremeni, Volume 3), or a resistancelimpedance (Artide 203, ResistanceMeasurement, Volume 3). Thus, their explanation is fundamentalto most measurements. So far as voltage measuringdevicesare concerned,voltage measurementcan be classified as 1. low-voltage measurements,such as those generatedby sensors; 2. medium-voltage measurements,such as those that exist in the power mains, laboratories, and industriaI operations; 3. bigh-voltage measurements,sueh as a rise in power generators,transmissionlines. and with plasma effects.

-

Consequently,thereis a wide rangeof voltage measuring techniquesand devicesin use. This article concentrateson medium-voltageand high-voltagemeasurements. Low-voltage measurementsfor sensorsrequire sophisticated signal processingschemes- see Artide 121, Signals in the Presenceor Noise, Volume 2; Artide 176, Signals and Signal-to-noise Ratio, Volume 3; Artide 178, Noise Matching and Preamplifier Selection, Volume 3; Artide 179, Input Connections; Grounding and Shielding, Volume 3; and Artide 181, Amplitude Modulated Signals: The Lock-in Amplifier, Volume 3. Common to alI techniquesare three major aspectsthat characterizethe measurements: 1. Amplitude: li the voltage is smaller than a few millivolts. we may'òeed to use suitable electronic components to amplify the'signals. li the voltage is large. in the kilovolts and megavolts region. we may need to attenuatethe magnitudein order to bring it to manageable levels. 2. Frequency: The frequencyof a voltage signal plays an important role in configuring the appropriate components of a voltage measuring device. The frequency of interest can range from DC to a few gigahertz. If digital techniquesof measuring afe used. sampling of the voltage signals must conCorroto the Nyquist sampling criteria. The signal frequencywaveform also needscareful considerationtest errors be generated. 3. Duration: The duration of a signal is significant in deterrnining the appropriate technique to use for the measurement.Duration may vary from continuous signals; as in the case of power supplies. to impulses appearing for a few microseconds;as in the casesof surges. and corona effects in power transmission and distribution.

Handbook01Measuring SystemDesign. edited by Peter H. Sydenhamand Ricbard 'l'borA, e 2005 Jobo Wiley & Soo8.Ltd. ISBN: 0-470-02143-8. '

-

, ;"./ .

' ..

,.

'èi~~! ,'.;:;

,,'

.

. '. I -.. .",,."",.;i,~'~"'~1;.i.~~.;:..;;~".:~~.:rl;:~';';"c

..~, J~

1354 CommonMeasurands

voltage

interactions

the

the angle

an

pointer

input

indicates the

the

hence

of of

value reading

based

the

convoltage

wideIy

of

a

afe in as

square used frequency

not

the

ftowing

instruments high

for

afe

to

current

a proportional suitabie

instruments

is

of

These

scale.

the

by

The

and

by

causes

arrange4

restraining

a

investigation.

to

Such

by

under

suitably

of

action

use dispiaced torque,

be

balancing

the

balanced

voltage

proportional

conductors. torque

the

then

is graduated

given

measured.

driving

to

The

by

torque

obtained

the be

afe

required

means

.

ei' electr
Figure 7. O' Arsonval galvanometerstructure.

betweentbc core and correspondingpole tace. 1be suspension is a single fine strip of pbosphorbronzeand servesasa lead to tbc upper eod of the coil. 1be lower end is connected to a lead consisting of a spirai spring. In the most accurate executions, tbe restraining torque is given by tbe torsion of the suspensionstrip. In this way, the friction torque is also practically removed.A small minor, fixed to the suspension, reftects a narrow beam of light through a glass window in tbe outer case surrounding tbe galvanometer 00 to a scale placed a meter away, on wbich tbe deftection is measured.Simply winding tbc coil on aluminium frame provides eddy curreot damping. Resistive damping may also be obtained by connectinga variable resistor in parallel witb tbe deftection coil. Proper adjustmentof this resistancegives criticai damping, tbereby reducing measurement time. A typical galvanometerstructureis shown in Figure 7.

9 DIGITAL INSTRUMENTS Digital ammeters (multimeters) obtain tbe required measurements by converting tbe analog input signal into a sequenceof digital samplesuniformly spacedin time in tbe early stages of tbc signal processing.The input signals are tberefore processedin tbc discrete-time domain and tbe measurement results afe displayed in a digital formo It is worthwhile to note tbat tbe distinction between analog and digital meters is not becauseof tbc way tbc measurementresults are displayed.but becauseof

Current Measurement 1369

Current

i(t) -

DSP ~

S&H

Digital output

Figure8. Structure of a modero digital meter. the domain (continuous-time or discrete-time domain) in which the input signals are processedin the main body of the devices. A more modero approach to this type of measuremeni exploits the capability of a fully digital structure as shown in Figure 8 that represents a voltmeter. The same structure can be adapted to an ammeter operation, provided that a current-to-voltageconversion input stage is inserted. This structure samplesthe input signal Vj(t) at constant sampling rate fs and converts each sampled value into a digital code. The whole sequenceof converted codes is stored in the memory of the DSP and then processedfor the evaluation of information. Assume that the input signal is periodic, with period T, and that the frequency spectrum is upper-limited by the harmoniccomponentof order N. Digital signal processing theory, and in particular tbe sampling theorem,ensurestbat the inforrnation associatedwitb tbe input signal can be tota1lyretrieved from tbe sequenceof tbe sampled data if at least (2N + l) samplesare taken over a period T in such a way tbat (2N + I)Ts = T, where the sampling period Ts being Ts= li/s' If vi (kTs) is the kth sample, tbe rms value of tbe input signal is given by

V

=

1

2N

+

L vf(kTs) 1=0

lO STANDARDS The unit of current is defined as, 'the ampere is that constantcurrentwhich, if maintainedin two straightparallel conductorsof infinite length, of negligible circular crosssection,and placed l meter apartin vacuUID,would produce betweentheseconductorsa rocceequal to 2 x 10-7 Newton per meter of length'. There is no definedstandardfor electrical current as such as it is mainly derived from existing electrical standardsof resistanceand voltage, which are based on the quantum Hall effect and the Josephsoneffect respectively. However, there exist numerousdocument standardsfor the current generation,harmonic contents,durationsof currents in high-voltage applications,electrical and electronic circuits, batteries, grounding and safety aspects, circuit breakersana fuses,and so OD.Some examplesare

. .

2N

1

full resolution of the instrument without clipping or distorting the measuredsignal. To obtain the most accurate representationof the signal being monitored, it is important to use as much of the fun range of the ADe as possible.

(10)

.

As an example, a true rrns AC ammeterbasedon tbis approachcan feature an uncertainty as low as the 0.1% of the CulI-scalevalue with a 12-bit resolution analog-to-digital converter (ADe). According to the sampling theorem, the instrument bandwidth is limited to half the sampling frequency. This means that a 500-kHz bandwidth can be attainedwith modero devices.Although wider bandwidths can be obtained, this is paid for in terrns of a lower resolution of the ADC devices. In many cases,such as current measurementsin highvoltage and power quality applications, special arrangements can be made by using appropriate current transducers and supporting equipment. In these cases, carefuI consideration is needed when sizing the transducers (e.g. cfs) required so that they take advantage of the

.

."

IEEE C37.09-1979 for eircuit breakers; ASTM 0495-99 for test method for high-voltage. loweurrent, dry are resistaneeof solid electrieal insulation; mc 60255-8 for high-voltage eurrent-limiting fuses; AS 2024-1991for high-voltage AC switehgear.

FURTHER READING Dyer, S.A. (ed.) (2001) Survey of lnstrumentation and Measurement, Wiley, New York.

Eren,H. (2003) Electronic Portable Instrumems-Design and

Applications. CRC PressLLC, Boca Raton, FL. Eren, H. and Ferraro,A. (2003) Electronic Voltmetersand Ammeters, EncyclopediaoJ liJe Support Systems,EOLSSIUNESCO, http://www .eolss.netlE6-39 A-toc.aspx. Eren, H. and Ferraro, A. (2003) Galvanometersand Electromechanical Voltmetersand Ammeters, EncyclopediaoJ LiJe SuppOrI Systems,EOLSS/UNESCO,http://www.eolss.netlE6-39Atoc.aspx.