FLUKE®

The ABCs of Oscilloscopes

Outline Page
Introduction 3
Chapter 1 Oscilloscope Basics 5
1.1 Description and function 5
1.2 Analog oscilloscope block diagram 6
1.3 Vertical deflection 6
1.4 Horizontal deflection 9
1.5 Triggering 11
1.6 Additional functions 13
Chapter 2 Digital Storage Oscilloscopes 15
2.1 The need for digital storage 15
2.2 DSO functions and controls 22
2.3 Automatic measurements and processing 26
2.4 Interfacing 29
Chapter 3 Analog, Digital or Both? 31
3.1 Introduction 31
3.2 The simple repetitive signal 31
3.3 A more complex repetitive signal 31
3.4 Non-repetitive signal and transients 32
3.5 Analog, digital, or both 32
Chapter 4 Accessories and Software 35
4.1 How probes work 35
4.2 Probe types 38
4.3 Other accessories 39
4.4 Software 40
Chapter 5 Newer Developments 41
Measurements and Exercises 43
6.1 How to perform measurements 43
6.2 Basic exercises 44
6.3 Measurement exercises 48
6.4 Cursors and automatic measurements 50
6.5 Digital Storage Oscilloscope exercises 54
Index 56
List of National Sales Organizations 57 JNTRODUCTION

Oscilloscopes are vital tools in every area of electronics. And the increasing penetration of [electronic equipment means that more and more engineers and technicians are faced, at some [tlme or other, with the need to use these instruments for development, troubleshooting and

I maintenance tasks.

To maximize its practical value to users, this booklet is much more than just a theoretical descrlpfton. A complete chapter deals with a variety of standard measurements, including exercises showing how to set up the instrument and how to reach the desired result.

j This need arises both in 'traditional' electronic industry, which large numbers of young lengineers enter each year, and in new application areas where experience with the use of (electronic test tools may well be limited.

l This booklet aims to describe the oscilloscope and its applications to users of all levels. It

l! starts with a description of the basic analog oscilloscope, explaining all the controls and how [these influence the measurement capabilities of the instrument.

! Then the digital storage oscilloscope and its controls are described including the specific

! benefits and application areas of these instruments. A comparison between analog and digital I storage oscilloscopes is given, with the aim of helping users to decide whether one of these hypes - or a combination of both - is the best choice for their application.

:0

j The latest developments in oscilloscope technology are also described, including new ijcapabflities with which the users of earlier generations of oscilloscopes may not be familiar. ::';

I Since oscilloscopes are used to measure a wide variety of signals, the interconnections to the ; system under test can play an important role in the measurement. Most often, probes are used i for this purpose, and their operation and influence on measurements are explained.

.. Asa guide to oscilloscope control panels, examples are given based on a modern analog instrument, the Fluke PM 3094, and a combined analog and digital storage oscilloscope, the Fluke PM 3394A Autoranging CombiScope™. Photos of these control panels can be found on f¢ld-but pages at the front and back of the book. You may find it useful to keep these pages gp~Jlwhile reading.

iAspartof its continuing program of user support and training, Fluke makes available a selectibnof other publications on specific aspects of oscilloscope theory and applications. In aqqition, the complete Fluke range of oscilloscopes and other test and measurement tools is qg§qribedin the Fluke catalog. To get copies of any of these publications, please contact your n~~r~stFluke sales office. The addresses are qiven at the back of this book.

:.::-'-,.,':'-:: ... :.::.:,-".' .... '., ',." " .. . : : .. :.:::._'.:.'": '- ....

iyveihOpe you will find 'The ABCs of oscilloscopes' to be interesting and useful, and that it will helpV61.ltb get maximum benefit from these versatile instruments in your work!

Oscilloscope Basics

~1.1 Description and

, function

An oscilloscope can simply be .thought of as a voltmeter with ipictures.

.: A normal voltmeter has either a !pointer moving over a scale or a humeric display to give a read-out lof the signal voltage. An oscilloslcope on the other hand has a lscreen, displaying in graphical form !the variations in signal voltage over itime, i.e. the waveform.

'i The major differences between ~oscjlloscOpes and voltmeters are: ~1. Voltmeters give a numeric readlout, usually as a Root Mean

Squared or RMS value. It is not ! possible to get information about signal shape. Some voltmeters can also measure

:f peak voltages and frequency.

s. Oscilloscopes, however, show

<

4 the history of the signal as it

j changes over time, displayed in a graphical format.

!2. Voltmeters usually give mea-

I~~~~~~~~p~~ ~~~ ~:~~~y~~~'or

more signals at the same time.

I; Th~~isPfay system

.~ The display device of any

~ostilloscope is a Cathode Ray Tube, i abbreviated as CRT, see Fig.i. This ItubeiS based on a system that ~generates electrons, the gun, and hmitsthem towards the screen. hheseelectrons are focused to iforrnabeam that produces a spot

' •. ~ at-the-center of the screen. The

~ phosph6(on the inner surface of ithescreenemits light from the

j points at which electrons strike it. 1 Onthelr way to the screen, the

·1 electronspass through a system of I. deflection plates. The spot can be

,; shiUedacross the screen by means i of voltages applied to the deflection •• ~. systerl1,(;onsisting of horizontal (X) .j alidvei1ical (Y) deflection plates. I This is called electrostatic

i~ deflection.

On the inner surface of the screen, horizontal and vertical lines are drawn or etched to form a grid known as the gral/cule. Graticules usually have 8 vertical and 10 horizontal l cm squares called divisions. Some graticule lines are divided further into sub-divisions, and there are special lines marked 0% and 100%. These special lines are used with the graticule lines marked 10% and 90% to perform a measurement called risetime. We will look at this later.

As stated above, the phosphor used in the CRT emits light when struck by electrons. The phosphor continues to glow for a short time after the beam is removed. This time is called the persistence and varies with different phosphors. The most commonly used phosphors are P31, with a persistence of less than a millisecond (ms) and P7

with a longer persistence of around 300 ms, which is very useful for viewing slower signals. The P31 emits green light while the color of the P7 is more yellowish-green.

The input signal is applied to the Y-axls deflection plates and the oscilloscope itself sweeps the beam across the X-axis. This makes the spot of light draw a copy of the input signal on the screen. This

copy otthe signal is called the trace.

Controls which influence the screen are:

-Intensitv

This adjusts the display brightness. The oscilloscopes we refer to in this book have a circuit which adjusts intensity automatically for different sweep speeds. When the beam moves very fast the phosphor is excited for a shorter time, so the intensity must be increased to see the trace. When the beam moves slowly the spot becomes very bright, so the intensity must be reduced to avoid' burning-in' the phosphor. This ensures a longer tube life.

There is a separate intensity control for the on-screen text.

- Focus

It controls the spot size to give a sharp trace. On some oscilloscopes, like the ones used as an example, focus is also optimized by the oscilloscope itself to keep the trace sharp at different intensities and sweep speeds. A focus control is also provided for manual adjustment.

Fig. 1 Diagram of a Cathode Ray Tube

electron gun

[simplified)

vertical deflection plates

electron' beam

horizontal deflection plates

screen"

)

trace

)

- Trace rotation

This allows the x-axis sweep to be aligned with the horizontal qraticule lines. The earth's magnetic field changes from place to place, and this can affect the displayed sweep. Trace rotation compensates for this. The control is a preset, and is usually adlusted only after the oscilloscope has been moved.

- Graticule illumination

Graticule brightness can be controlled independently, useful for screen photography or when working in low light conditions.

- Z-modulation

The trace intensity can be varied electrically by means of an external signal. This is useful when horizontal deflection is generated externally and for finding frequency relations, using X-V-display.

The input for this signal is usually a BNC socket on the rear panel.

Fig. 2 Block diagram of an analog oscilloscope

1.2 Analog oscilloscope block diagram

The CRT is the basis of any oscilloscope. Now that we know what it is, we can look at how it is used as the heart of an oscilloscope.

As we have seen, an oscilloscope has two vertical deflection plates, two horizontal deflection plates, andan electron gun by means of which the intensity of the emitted beam can be controlled

electrically.

Around this, the following circuitry is added to make a complete oscilloscope (see Fig. 2). The vertical deflection system comprises of:

- input attenuator (one per channel)

- pre-amplifiers (one per channel)

- an electronic switch to select

which input channel is used - deflection amplifier

The horizontal deflection system is built around a timebase, a trigger circuit and the horizontal deflection amplifier.

The intensity is controlled electronically to activate or blank the trace at the right moment in time.

To make all this work, an oscilloscope has a power supply that can operate from the mains supply or from an internal or external battery.

The basic performance of any oscilloscope is determined by the characteristics of the vertical deflection system, so let's first look at these in more detail.

1.3 Vertical deflection

Sensitivity

The vertical system scales the input signal so that it can be viewed on the screen. Oscilloscopes show signals with peak-to-peak ('p-p') voltages from a few millivolts to tens of volts. All these must be made to fit on the screen so that they can be measured against the graticule. This calls for attenuation

[>

Chop [>
&
Alternate
y
[>
Trigger

Timebase

of the large signals and amplification of the small 'signals. The sensitivity or attenuatorcontrol does

Sensitivity is measured in volts per division. If we look at Fig. 3 the sensitivity is set to 1 V/div. The 6 volt pop signal therefore deflects the trace through 6 vertical divlslons, knowing the sensitivity setting and the number of vertical divisions the beam moves through, we can measure an unknown pop signal

On most oscilloscopes the steps on the sensitivity control are in a 1- 2-5 sequence, i.e. 10 mV/div.,

20 mV/div., 50mV/div., 100mV/div., and so on. The sensitivity is set by pressing the amplitude up/down buttons or, on other instruments, turning the vertical sensitivity knob.

If these steps do not scale the signal exactly as required on the screen then the variable (VAR.) trol can be used. Making a risetime CI measurement using the graticule is p a good example, as we will see in w chapter 6. The variable control continuous adlustrnent between the SI

1-2-5 steps. Usually when using al

VAR the exact sensitivity is not CI

known; only that it lies somewhere L,

between two of the steps in the 1 al

5 sequence. The V-deflection for T

channel is now said to be uncali- is

brated or 'uncal'. This condition is is

usually indicated on the front panel al or screen of the oscilloscope. o

On more modern oscilloscopes, o

like our example, the sensitivity is ty

continuously variable between mini-

mum and maximum but remains CI

calibrated because of the modern T

techniques used for control and a1

calibration. a1

On older instruments the channel 9 sensitivity setting is found by SI looking at a scale around the sensi- Zf tivity control. On newer oscillo- Ie scopes it is displayed clearly on the ti, screen as shown in Fig. 3, or is

indicated on a separate LCD. II

Coupling in

The coupling control determines p

the way the input signal is passed al

from the front panel BNC input 31

connector to the rest of the vertical Ic deflection system for that channel.

I

Iplifica-

1e sensi- .. , does thls.i in volts

ig. 3 the! The 6 j tlects the·~ isions, Bym

~:~~sa~dej

:an mea- .~ 11 voltaqe.] the steps'; re in a t-] liv., }mVldiv., s set by 'down ments, b lity knob. 31e the

In the AR.) conrisetime aticule is

I see in nrol give ween the using

i not newhere

1 the 1-2 on for th 'lncaliiltlon is xrt panel pe. iscopes, tivity is

There are two settings, DC coupling and AC coupling.

DC coupling provides a direct connection for the signal. All components (AC and DC) of the signal willtherefore affect the display.

ACcoupling inserts a capacitor in series between the BNC connector and thsattenuator. Any DC component in the signal is blocked. Low-frequency AC components are alsO blocked or greatly attenuated. The-low-frequency cut-off frequency isthefrequency at which the signal is displayed at only 71 % of its true amplitude. The low-frequency cutoffctepends primarily on the value of the input coupling capacitor. A typically value is 10Hz, see Fig. 4. \ASsociated with the coupling control is the input Groundfunction. TlliS,disconnects the signal from the atte'nuator and connects the attepuator input to the oscilloscope gr6~ndleveL When ground is seJ~ttedYou will see a line at the zefOV6ltieVeL This line at reference leVelorbase line can now be positibnect\Using the position control.

mains nodern Ii and

.:.: ,:,·':·:::·.-:'i\,::,·:.::\:,:,::·:·,::::::,::;:,:':::,::::._: . '-,.:.'.:.'.'.

..lnPH!impedance iMqstoscHiosGopes have an input imp§dMceof1 Mohm (1 Mn) in

termine R~r~II,~,rvvitharbund 25 pF. This is

passed\~88~ptablefb(most general-purpose nput~ppncations as it causes little

) vertical16~dirJgori most circuits. ~hannel.,\$9Il1esignals come from sources

or is . D.

.c h r+

with a 50n output impedance. To measure these signals accurately and to avoid distortion they must be carried and terminated correctly. Cables with son characteristic impedance are used and these must be terminated with a son load. Some oscilloscopes, like the PM 3092 and PM 3394A, have this son load built in as a user-selectable function. To avoid accidental operation, its selection may have to be confirmed. For the same reason, the son input impedance cannot be used with certain probes.

Position

The verticel position control or pas positions the trace in the Yaxis on the screen. The ground level can be detected by selecting ground for

Input

•

Fig. 3 A signal with 6 volts p~p deflects the beam vertically through 6 divisions at a sensitivity of 1 Vldiv

Fig. 4 Simplified input circuit showing AC and DC coupling, input grounding and 50Q input impedance selection

50Q

the input coupling, thus disabling any input signals. In more modern instruments, a separate ground level indicator is found which continuously allows the operator to find the reference level with the waveform.

Dynamic range

This is the maximum amplitude of a signal that can be handled without distortion, while still being able to see all parts of it by changing the vertical position. For Fluke oscilloscopes this is typically 24 divisions (3 screens).

Addition and inversion

Simply adding two signals might appear to have little practical value. However, inverting one of two related signals and then adding them in effect subtracts the signals. This is very useful for removing common mode interference (e.g. hum), or for differential measurements.

Subtracting the input signal from the output signal of a system, after proper scaling, shows the distortion caused by the system under test.

Since many electronic systems have an inverting nature, the desired subtraction can indeed be achieved by adding the two oscilloscope input signals!

Ground

to input attenuator [1MQ)

. •

Fig. 5 Frequency Response of a typical 100 MHz Oscilloscope (simplified and actual)

Alternate and Chop

The oscilloscope CRT in itself is able to display only one trace at a time. In many oscilloscope applications, however, signals are often compared, for example to study input/output relationships or the signal delay through a system. This requires the instrument to display more than one signal, virtually at the same time.

To do this, the electron beam can be controlled in two ways:

1. It can alternately draw one trace completely, then the other; this is called alternate mode or simply ALT.

2. It can draw the traces in parts by switching or chopping between them very quickly. This is called chopped mode or CHOP. It results in two traces being drawn during one sweep, piece by piece. The chopped mode is better for

displaying low-frequency signals at low timebase speeds since the chopper can switch fast.

Alternate mode is better for higher frequencies needing faster timebase settings. The oscilloscopes we use as examples select ALT or CHOP automatically at different sweep speeds to give the best display. The user can also choose ALT or CHOP manually to suit any particular signals.

Bandwidth

The most important specification

of any oscilloscope is the bandwidth. Bandwidth describes the frequency response of the oscilloscope's vertical system, and is defined as the maximum frequency that can be displayed with an amplitude on screen which is no more than 3 dB lower than the true signal amplitude.

The -3 dB point is the frequency at which the displayed signal amplitude 'Vdisp' is displayed at 71 % of the true value of the signal at the input, ' Vinput', as shown below.

Given that:

dB (volts) ~ 20 log (voltage ratio)

-3dB ~ 20 log (VdiSp 1 Vinput )

-0.15 = log (Vdispl Vinput)

10-0.15 = VdiSP 1 Vinpul

V disp = 0.71 Vinpul

Fig. 5 shows a typical frequency response curve for a 100 MHz oscilloscope.

For practical reasons, bandwidth is often considered as if the frequency response is flat up to the cut-off frequency and then declines at 6 dB/octave (-20 dB/decade) from that frequency. Of course this is a simplification. In reality the amplifier's sensitivity declines slowly at lower frequency and reaches -3 dB at the cut-off frequency. The figure gives the simplified as well as the realistic frequency response.

Bandwidth limiter High-bandwidth oscilloscopes,

Amplitude %

71

Actual Frequency ------Response Curve

-3dB point

-3dB Bandwidth

High Frequency Roll-off -20dB(decade

Frequency ~----------------------~------------------~ MHz

100

usually types with bandwidths of t 100 MHz and above, can be reduced r:

in bandwidth to typically 20 MHz, ~!

thus reducing noise levels and t

interference, especially for high-

sensitivity measurements. i:

Risetime

Risetime and bandwidth are

directly related. Risetime is usually l' specified as the transition time for a signal to go from the 10% to the

90% level of the steady maximum value.

For an oscilloscope this is the fastest transition which can theoretically be displayed. The high-fre- z: quency response of an oscilloscope C has a carefully determined curve. ~ This ensures that signals with a high t harmonic content, such as square- r

waves, are faithfully reproduced on t

screen. If the roll-off is too fast,

ringing can result on fast rising

edges. If the roll-off is too slow, i.e. starts too early on the response

curve, overall high-frequency

response is adversely affected and squarewaves lose their 'squareness'.

This response curve is similar for all general-purpose oscHloscopes, so from it a simple formula relating bandwidth and risetime can be derived.

The rule is:

tds) = 0,351 BW (Hz)

This can be recalculated for highfrequency oscilloscopes:

t, (ns) = 3501 BW (MHz)

For a 100 MHz oscilloscope, the rlsetlrne is 3.5 ns (ns = nanoseconds ( =10-9s).

There are special lines on the graticule marked 0% and 100%. These are used for risetime measurements. We use the VAR sensitivity control to align the top and bottom of the signal under test with I the 0% and 100% lines.

The risetime is then read along

the X-axis as the time between the intersection of the signal with the 10% and 90% graticule lines. (See chapter 6).

To measure the risetime of an oscilloscope we follow the same procedure except that the test signal must have a risetime much faster

- . _ - - - @Hapter 1 Oscilloscope Basi~~

iths of

e reduced D MHz, and

high-

than that of the oscilloscope, at least 5 times for a 2% error. The displayed risetime is a combined function of the oscilloscope risetime and the signal risetime. The relationship

is: _., _

- . '(t 2 + t 2)

tr displayed - 'I rsignal rscope

Remember this formula, you will find it useful!

1 are

i usually time for a to the edrnum

1.4 Horizontal deflection

is the

1 theorejh-freilloscope curve. lith a

Timebase

To draw a graph we need horizontal as well as vertical information. Oscilloscopes draw traces to show how signals behave over time, so

the horizontal deflection must be proportional to time. The system that controls the horizontal deflection, the X-axis, is called the timebase.

Inside the oscilloscope we find

an accu rate sweep generator. It sweeps the electron beam across the screen at precise, user selectable speeds. The output of the timebase generator is shown in Fig. 6.

Sweep speed is measured in seconds per division (s/div.). The sweep speed range on a typical oscilloscope is from 20 ns/div. (ns = nanosecond = 1 xi 0.9 seconds) to 0.5 s/div. These sweep speeds also follow a 1-2-5 series like the sensitivity control. Once we know the time represented by each graticule division we can measure the time between any two points on a trace.

Ising slow, i.e. onse

cy

ted and

scopes,

I relating 1 be

n the 00%.

Example

Fig. 7 and Fig. 8 show a 1 kHz sinewave, (period equals 1 ms), displayed at 1 ms/div. and at 200ws/div. (~s ::: microseconds ::: seconds).

l along een the ith the

~s. (See

:Df an same

est signal \ faster

'.:'.,::".-:-.'.,.

Horizontal position control

The horizontal or X-axis position control X-POS moves the trace horizontally across the screen. This Illeans that a specific point on the trace can be aligned with a vertical

. graliCllle line to provide a starting pqintfora time measurement.

,

Fly- ,! Hold-off J SwBBp time

back; .-'

I Waiting for a new !;rlgger

Sweep time

Ryback

Input. Signaj

Trigger

Tlmebeee

.. k.v ..... .v;V ..... V ...... U.: ...... \J:~ ... ,v..: ... ·V

.. ..... : ."

......... ~- -.

.................... [: .....

MTBl.QQm:;;

MTB 2001.(5

.c h I-}

Fig. 6 Timebase generator output showing Sweep Time. Flyback and Ho/d-offtime

Fig. 71kHz sinewsve, TB set to 1 ms/div

Fig. 81kHz sinewave, TB set to 200 /ls/div

Ttrnebaee ~agnificatiDn

................... --_-_------:------10-----1 __ -- __ BP--

X-posit1on control

Fig. 9 Timebase magnification and x-position control

Fig. 10 Dual Iimebese operation (500 jls/div. and 50 us/d!«, 4 divisions delay)

Variable timebase

We can select sweep speeds which are different from the settings in the standard 1-2-5 range. This lets us place for example one period of any waveform across the entire screen. As with VAR for the Y-axts, most oscilloscopes give an indication that variable timebase is in use and the X-axis is not calibrated. More modern oscilloscopes, like our examples, can operate in a calibrated continuously variable mode. This allows for better time resolution measurements because the entire screen can be used to display the signal part of interest. It also dramatically reduces the possibility of operator error.

Timebase magnification

Timebase magnification magnifies the X-deflection sweep, usually by a factor of ten. The effective timebase speed as seen on the screen is therefore 10 times faster. So the typical oscilloscope with

20 ns/div. unmagnified can now sweep at 2 ns/div. The screen is now a movable viewing window on the signal. This gives an advantage compared with simply selecting a faster timebase speed - it allows the original signa! to be retained while examining it in much greater detail.

Fig. 9 shows how the x-pas control is used to scroll through the signal.

.... : .,,, .

CHI

MT8 OOus

2 v- ALT DTB5 .Ous 2.049ms

chl+

Omil Timebase

In a number of applications dealing with complex signals, a small part of a waveform has to b displayed so that it fills the entire e screen. One example is the in which a selectable video line of composite video signal needs to be studied. In these situations, triggering of the standard timebase will not do the job. This is why timebase operation is found on modern oscilloscopes.

The main timebase (MTB) can triggered on a main trigger event in the waveform, e.g. the vertical synchronization signal of the video system. Part of the MTB trace is displayed brighter, the intensified part. A second timebase, called the delayed timebase or DTB, is at the beginning of this intensified part and can be set to its own speed, faster than the deflection setting of the main timebase. The delay between the start of the MT8 _ and the beginning of the intensified part is adjustable.

It is even possible not to start DTB when the selected delay has expired, but to arm a trigger circuit

for the DTB at that moment. If a trigger event occurs later, the DTB sweep will then start.

So with a dual timebase the electron beam is swept alternately across the screen at two different speeds by the two timebases.

Let's look at Fig 10. First the timebase runs at 500 ~s/div. This draws a waveform on the screen. During this sweep after a time of 2 ms, equivalent to 4 divisions, the trace is intensified. This time is set by the delay1 control. The duration of the intensified area is set by the DTB sweep time control, 50 ~s/div. in our example.

When the delayed timebase runs after the 2 ms delay, it displays jus!. a tenth of the original main timebas1. - A trace but it does so across the entire It i~ wh

the

1) On older instruments DHA Y is referred to i the the Delay Time Multiplier. It consists of a .• bas scale on a multi-tum potentiometer. TheDe' pro is calculated by multiplying the D T8 sweep Th i speed by the number on the scale when lliB po:

intensified part of the trace is located as . '

desired on the MTB. Hence the name. slg

screen.

I

Changing the delay time changes the starting point of the delayed timebase sweep on the main timebase. Changing the delayed timebase sweep speed changes the length of the displayed part of the

main timebase.

The main timebase can be switched off when the delayed timebase has been set up to show the part of the signal under examination. This gives more brightness to the delayed trace.

The timebase operation modes of a typical dual timebase oscilloscope are:

- MTB::: Main timebase only

The MTB only is operational and the oscilloscope performs like a single timebase instrument.

'-- MTBI ::: Main timebase intensified Shows only the MTB sweep, but part of the trace is intensified to show the DTB start position and sweep.

c,_'. MTB Intensified and DTB

As MTBI but also shows the DTB ······.··sweep.

-' DTB = Delayed timebase

....... Shows only the DTB sweep.

..' There is more about the dual timebase in the trigger section of this book.

Timebase modes

The tlmebase circuit has several modes of operation. For normal al1alog oscilloscopes these are: automatic, normal or triggered and single or single shot.

-Normal

The timebase must be triggered to produce a trace. The rule here is easy: 'no signal, no trace'. Theoscil'. loscope must have an input signal at . the selected trigger source, large

..•... enouqh to trigger the timebase, in its absence there is no trace on screen.

- Single

The timebase will sweep on receipt of a trigger, and will do this only once. The trigger circuit must be armed i.e. prepared, for each trigger event. If it is not armed then subsequent trigger events will not start the timebase. The trigger circuit is re-armed by pressing a button marked single or reset, depending on the oscilloscope. To take the guesswork out of single-shot work, modern oscilloscopes can display their trigger levels in volts on the screen, or as horizontal lines.

1.5 Triggering

We have seen how the input signal is used for vertical deflection and how the timebase gives horizontal deflection. But how does the electron beam follow exactly the same path each time it crosses the screen? The answer lies in the trigger circuit. Without the trigger all you would see is a jumble of waveforms with random starting points. Each time the timebase sweeps the screen the trigger circuit ensures that the timebase starts at a precisely defined point on the input signal. This precise starting point is determined by the following controls.

Trigger source

This determines where the trigger signal comes from. In the majority of cases it comes from the input signal itself. So if Only one channel

is in use then the trigger source is also set to that channel. If more channels are in use then the source can be chosen from any of them.

Composite triggering is used to trigger alternately from different channels in turn as they are displayed. This is useful to display signals that are not frequencyrelated.

If the oscilloscope has an external trigger input (Ext.), it can derive the trigger from a signal connected to it.

For work on systems at mains frequency or a frequency derived from it, line triggering provides a trigger from the mains. This is a good way to find mains- related interference.

Trigger Level

The trigger level control sets the voltage level that the signal on the selected trigger source must cross to have the trigger circuit start the timebase.

Trigger Slope

The slope control determines if triggering will occur on a rising ('positive') or falling ('negative') edge of the source signal.

m"ln<"I'~~ •• i·.' .; Automatic

. It is useful to see the trace even

.• When there is no signal present. In the absence of a signal to trigger on, the automatic mode allows the time-

base to free-run at a low frequency producing a trace on the screen .

.• This allows you to set the vertical

'. POSition of the trace, e.g. if the

name. '.' Signa! is a DC potential only.

v-

h +

-----.-~. ~,

Fig. 11 An un triggered signal

G~l(1ter 1 Oscilloscope Basics' '. , .

CHI:

em :

CHI : ~ V~

Fig. 12 Influence of trigger level setting

Trigger Coupling

This is the way in which the selected source signal is passed to the trigger circuit.

- HF-Rej.

Passes the source signal through a lowpass filter to reject high frequencies. This means that we can trigger on a low-frequency signal even if it contains a lot of highfrequency noise.

so the oscilloscope will always be triggered when a signal is present.

- DC Coupling

The source is connected directly to the trigger circuit

-AC Coupling

The source is connected to the trigger circuit through a series capacitor

-LF-Rej.

Passes the source signal through a highpass filter to reject low frequencies. This is useful for instance for displaying signals which contain large amounts of mains hum.

- Level p-p

The range of the trigger level control is set to be slightly smaller than the peak-to-peak value of the source signal. In this mode it is impossible to set a trigger level outside the value of the input signal,

- TV Triggering

In this mode the level control is inoperative and the oscilloscope

Fig. 13 Video line signal

uses the synchronization pulses within a video signal. There are two modes for TV triggering: frame triggering TVF and line triggering TVL.

- TVF

Each TV picture frame is made up of two fields. Each field contains halt of the lines required to make the complete frame. The two fields are then interlaced on the TV screen to produce the frame. This technique reduces the bandwidth required for the transmission channel and reduces flickering of the picture. There is a specific sequence of synchronization pulses at the start of each field, called the frame sync., on which the oscilloscope triggers. Modern oscilloscopes can distinguish between field 1 and field 2.

-TVL

Each field contains a number of lines. Each line starts with a line synchronization pulse, or line sync. The oscilloscope triggers on each of these pulses and draws all lines on top of each other. Individual lines can be viewed by using the dual timebase and frame triggering or with the help of a special accessory called a Video Line Selector such as the Fluke PM 8917. With the CombiScope™ used in the example, the built-in line counter can be used to simply dial up the line number required (PM 3394A series only).

Trigger Hold-off

Some signals have many possible trigger points. A good example I the digital signal in Fig 14. Although it is repetitive over a longer time period, the short time situation is different. To look at fewer pulses in more detail the time base must run faster but now the displayed part of the signal changes from sweep to sweep. To overcome this the trigger hold-off delays the time between sweeps, enabling the trigger to come from the same edge all the time.

Delayed Timebase Triggering

We saw in the time base section that the OTB started, i.e. was triggered, after a delay on the MTS sweep. This delay is measured from

Input Signal

Timebase

Hold-off

Waiting for

new trigger ~--

I
2
I-- I· -- .... --. --
1 2
- 1
I Hold-off too short

trigger point, and after this

. e the DTB is actually

. the delay system. This

.. called DTB starts.

can also operate in a mode, similar to the MTB. .. controls to set DTB Trig-

. level, slope and coupling

rate independently of

.. By selecting this set-up the red for triggering

. . at the moment the delay is started by a new detected in an input I mode of operation is DTB.

see in the exercises how ggered DTB.

Scope semen

1.6 Additional functions

x- Y Deflection

X- Y Deflection 0 r X- Y mode is a method where the timebase is disabled and an input signal, different from the one used for the vertical deflection, is used to deflect the electron beam in the horizontal direction. This means that two input signals can be plotted against each other to see their relationship .

The most common use of this method is to see the phase relationships of signals. Fig. 15 shows the pictures called Lissajous Figures which are produced when sinewaves of harmonically related frequencies are used. If signals of unrelated frequencies are applied, no stable display can be obtained. For signals with a fixed frequency relationship,

Hold-off (extended)

--_ ..... _-_.

,-- I

r--- c- --- r----- .. --
1 2
- ~ '-- Correct hold-off setting

the phase relationship can be derived from the display. As an example, Fig. 16 shows the display for two sinewaves of the same frequency, with phase differences of 0 degrees, 45 degrees and 90 degrees .

There are many other applications in which X-V mode is used. Many of them are in electromechanical environments. The relationship between two physical parameters can be displayed using appropriate transducers to convert the physical quantity into signals the oscilloscope can display, for example displacement against pressure. The X-V mode can also be used in the electronics laboratory for component testing, such as plotting the characteristics of a diode, or in fact any situation in which two interdependent quantities are found.

3

4

5

Scope screen

Fig. 14 Trigger holdoff on complex pulse

Fig. 15 Lissajous figures; vertical defiecljon ~equency is a full multiple of the horizontal deflection frequency

Fig 16 Lissajous figures, same frequency applied to horizontal and vertical deflection

Fig. 17 The effect of a delay line on a fast rising edge

CHI

x~ CH2

0" Pha seshift

45° Phaseshift

with delayline

without delayline

CHI

20niV~

: MTB 20.0ns

chl+

The delay line

Although the delay line is part of the vertical deflection system, we will discuss it here because the trigger and horizontal systems are influencing factors.

The trigger and timebase circuits, however fast, take time to respond to a valid input trig-ger condition. The timebase also has a small nonlinear period at the start of the . sweep until full sweep speed is reached. For lower-bandwidth oscilloscopes, these times, in the order of nanoseconds, are negligible compared with the fastest signals

the oscilloscope can display. For higher-bandwidth instruments with sweep speeds up to 2 ns/div., these times are significant. To enable us to display events in the order of a few nanoseconds, the timebase needs to be triggered before the trigger event in the waveform reaches the screen.

This means that the beam is already sweeping when the trigger information of the signal arrives at the deflection plates. The full rising or falling edge can now beseen.as well as a few nandsecondsofsignal before the trigger. These few nano. seconds are called pre-triggerinfor~

matlon This is achieved by a signal delay line in the vertical system after the trigger circuit pickoff point and before the final amplifier. This delay line stores the Signal for a period of time which is proportional to its length. By the time the signal reaches the end of the delay line, the timebase has started and the sweep is running.

2,1

fJ Cha; P31 asci In s· phor per! sho: is pi pho the mor

\J

Summary

We have seen how a basic oscilloscope functions, using the system blocks described.

All elementary controls have been described. We will see later that

. most of these controls are found on analog oscilloscopes as well as on digital storage instruments.

Digital Storage Oscilloscopes

Digital storage

.. ···As you may remember from Chapter 1, the persistence of the P31 phosphor in a normal analog oscilloscope CRT is less than a ms. Insome cases, CRTs with P7 phosphor are used, giving about 300 ms persistence. The CRT continues to show a signal as long as the signal is present to illuminate the phos.phor. When the signal is removed the trace decays quickly with P31, or . mOre slowly as with P7 phosphor. iWhat happens if the signal

Occurs only a few times per second, ()rifthe duration of the signal is several seconds, or worse still, if

. th~isignal occurs only once? In these cases it is almost or com pletelyTrnpossible to view the signal with the analog oscilloscope that we disclissed in the previous chapter. What is needed is some method of retaining the swept path of the sigllal over the phosphor. The old way of accomplishing this was to useaspecial kind of CRT called a Storage Tube which used an

. electrically charged mesh behind the phosphor to store the path of the electron beam. These tubes were very expensive and delicate, and could only hold the trace for a lihlited time.

Digital Storage overcomes all of these disadvantages, and brings a great number of additional features as well. Some of these are:

II large amounts of pre-trigger information can be shown

III fully automatic measurements are possible, either with or without cursors

II waveforms can be stored indefinitely

III waveforms can be transferred to a computer for storage or further analysis

II hard copies can be made on a printer or plotter for documentation purposes

II newly acquired waveforms can be compared with reference waveforms, either by an operator or fully automatically

II decisions can be taken on a pass / fail basis ("Go / No Go" tests)

III waveform information can be processed mathematically

Digital Storage - What is it?

As the name suggests, digital storage is the storage of a signal in digitally coded form in an oscilloscope.

When the signal enters the digital storage oscilloscope, or 080, and before it is passed to the CRT

deflection circuitry (see Fig. 18), the voltage is sampled at regular time intervals. These instantaneous values or samples are used by an Analog-to-Digital Convenor (AD C) to generate binary words representing each sample Voltage. This process is called Digftization.

The binary values are stored in digital memory. The speed at which the samples are taken is called the Sampling Rate, and is controlled by the sample clock. Sampling rates for general use range from 20 Mega-Samples per second (MS/s) to 200 MS/s.

The data in the memory is used to reconstruct a waveform on the oscilloscope screen.

I n a DSO there is more than just analog circuitry between the input connectors and the oscilloscope CRT. A waveform is first stored in memory before it can be displayed. What is seen on screen is always a reconstruction of theacquired data, rather than an immediate and continuous display of the signal applied to the input connectors.

Fig. 18 Block Diagram of a Digital Storage Oscilloscope

ADC

Memory

ADC

Trigger

X-tal Timebase

DAC

[>

y

DAC

Fig. 19 Basic Sample and Hold circuitry

Input

Control

I· Hold capacitor

+ Ref. Voltage

Digital Output

Fig. 20 Analog-toDigital Can vetter, basic circuitry

Fig. 21 The effect of vertical resolution on the displayed waveform

Sampling and Digitizalion

Digital storage is achieved in two steps. First, samples are taken of the input voltage. This is done by a Sample-and-Hold circuit, see Fig.19.

The input amplifier A1 charges and discharges the hold capacitor via switch S, when S is closed. Once S is opened, the voltage on the capacitor can no longer change. Buffer A2 feeds this sample to the Analog-fa-Digital Converter (ADC). The ADC measures the amplitude of the sample and expresses it as a digital word.

The converter is built around an array of comparators, see Fig. 20. Each comparator checks if the sample is higher or lower than its reference voltage; if higher the output is active, if lower its output is inactive.

All comparators have slightly different reference voltages, all derived from one reference voltage source using a chain of resistors. For any sample value, a number of comparators will be active and all others will be inactive. The code converter in the ADC converts the sample value into a digital word, and sends this to the digital memory.

This type of ADC is called a converter, because it converts an analog input voltage 'in a flash' into a digital word. Other types, where the conversion is done in a few steps, are also used but these have the drawback of taking more time convert a sample.

Analog-Io-Digital Converter and Vertical Resolution

The ADC has to determine the amplitude of a sample by comparing it against a number of reference voltages. The more comparators are built into the ADC to do this, and the longer the resistor chain is, the more different voltage levels can be recognized. This is called the vertical The higher the vertical resolution,

ADC Comparator Reference Levels

Sample Mornents+e-

Reconstructed Waveform

I T

T Input Signal

t f

ADC Comparator f Reference Levels T

I

j

i

~

I

Reconstructed Waveform

Low resolution. hence large voltage steps

~--I- 1··-1--+-+-1···1 .. I---+--+-+-t-I-I I I··

Sample Moments-

smaller the signal detail that can be seen in the waveform (see Fig. 21).

The vertical resolution is expres-

sed in bits. This is the total number of bits (i.e. the size of the digital output word) that make up an output word.

The number of voltage levels that can thus be recognized and coded can be calculated as:

number of levels =: 2 numberofbils

Most oscilloscopes use 8-bit converters and are therefore able to .. describe a signal in 28 = 256 differ-

• ... ent voltage levels. This provides

.. sufficient detail to study signals and .• to make measurements. This way,

... the smallest signal steps displayed

....•• are of about the same size as the .. diameter of the spot on the CRT

.. screen. One ADC output word, rep.•..•. resenting the value of one sample,

• .•.. comprises 8 bits and is referred to

• .•.•. as a Byte.

A practical limitation to increasing the resolution is one of cost. In the construction of the ADC itself, each additional bit in the output word calls for double the number of comparators and a larger code converter. This makes the ADC then occupy double the chip space and

... take double the power (which in

..•..•. turn affects surrounding circuitry).

..As a result, additional resolution

. comes with a price tag.

• Timebase and horizontal ....... resolution

The task of the horizontal system •.•..... in a digital storage oscilloscope is to ..• • ensure that enough samples are acquired, each at the right moment. ·.Just as in an analog oscilloscope,

. the speed of the horizontal deflec-

. n depends on the timebase

. setting (s/div.).

The total group of samples that make up one waveform is called a record. A record can be used to

.e: reconstruct one or more screens. ',The number of samples that can be F'·'''''''''' is called the record length or acquisition length, expressed in Byte or in kByte, where 1 kByte

kB} equals 1024 samples. Usually, oscilloscopes display

512 samples across the horizontal axis. For ease of use, this number of samples is displayed at a horizontal resolution of 50 samples per division. This means that the horizontal axis has a length of 512/50 = 10.24divisions.

From this, the time interval between samples can be calculated:

sampling interval x: timebase setting (s/div.) number of samples

For a timebase setting of

1 ms/div., and given the 50 samples per division, the sampling interval can be calculated as:

sampling interval ~ 1 ms / 50 ~ 20 us .

The sampling rate is the reciprocal of this sampling interval:

sampling rate =

sampling interval

Useful tisetime .

Many oscilloscope applications deal with measuring the switching characteristics of Signals, e.g. the rise- and falltimes.

As we have seen in Chapter 1, the risetime of an instrument gives the fastest transition which the instrument is able to handle accurately. For an analog oscilloscope, this depends entirely on the analog circuitry of the oscilloscope.

If a DSO is used, the fastest transition that can be acquired depends on the analog circuitry as well as on the time resolution. For a correct risetime measurement, sufficient detail must be acquired on the edge concerned, which means a number of samples have to be acquired during the transition. This risetime is called the useful rise time of a DSO and is a function of the timebase setting.

We'll look at rlsetlrne measurements in detail in the exercises section of this book (Chapter 6).

Usually the number of samples that can be displayed is fixed, and a change in the timebase setting is achieved by changing the sampling rate. The sampling rate specified for a particular instrument is therefore valid only at a specific timebase setting. At lower timebase settings, a lower sampling rate is used.

Given an oscilloscope that has a maximum sampling rate of 100 MS/s, the tirnebase setting at which this sampling speed is actually used is:

timebase setting

= 50 samples * sampling interval ~ 50/ sampling rate

= 50/ (100 x W)

= 500 ns/div.

It is important to know this timebase setting because this is the timebase setting for the fastest acquisition of non-repetitive signals; it gives the best possible time resolution.

This timebase setting is known as 'the maximum single-shot timebase setting' at which 'the maximum real-time sampling rate' is used. This is the sampling rate which is stated in the instrument's specification.

Maximum captured frequency, Shannon's sampling criterion

When the first attempts were made to digitize signals, a study revealed that proper recovery of the Signa! required the sampling clock to have a frequency of at least double the highest frequency contained in the signal itself. This is generally known as Shannon's sampling theory.

However, this study was performed with communications applications in mind, and not oscilloscopes .

Looking at Fig. 22 (top) it appears that a signal's frequency can indeed be recovered when a sampling clock of double the signal frequency is used. With proper reconstruction facilities, this may result in a waveform that closely resembles the original signal. But is it really that simple?

Let's assume the samples are taken at a slightly different moment in time with the same sampling clock, but not necessarily at the peaks of the signal. Any amplitude information will now be seriously in error, or may even be completely

c[l1apter 2 Digital Storage Oscilloscopes

Fig. 22 Sinewave sampled at twice the signal frequency, near the peaks and near the zero crossings

Input Signal, sampled in

the peaks .

... and the reconstructed waveform.

Input Signal, sampled near the zero crossings .....

... and the reconstructed waveform.

lost. In fact, if the samples are taken exactly at the zero crossings (see Fig. 22, bottom), no signal will be found at all since all samples represent the same signal value: zero.

Oscilloscopes are used to study signals. A good study of a signal requires good frequency representation, but moreover it requires accurate waveform representation at the correct amplitude. As Fig. 23 shows, three samples per cycle degrades the fidelity with which the signal is reproduced.

As a rule of thumb, ten samples per cycle is generally regarded as the minimum to give sufficient signal detail. In some cases less detail

is required and five samples per cycle might be enough to get some idea of the signal's characteristics (see Fig. 24 ). For an oscilloscope with a maximum sampling speed of 200 MS/s, this gives a maximum signal frequency to be acquired accurately of 20 to 40 MHz. In those cases, the display fidelity can be enhanced by using special display systems that draw a best-fit sinewave through the samples. This is known as sine-interpofation.

AliaSing

As we have seen, a minimum number of samples is required to reconstruct a waveform. In all cases

the sampling clock must be five to ten times higher than the signal frequency.

If the sampling clock is lower than the Signal frequency, you will get unexpected results.

Let's look at a situation as in

Fig. 25. As the drawing shows, successive samples are taken from different cycles of the waveform. However, it just happens that each new sample is acquired with a slightly longer time interval with respect to the zero crossing. If we now display the samples and E reconstruct a waveform from them, f

the display will show a sinewave t

again. But the reconstructed t

Fig. 23 Signal sampled at about 3 samples per cycle

Input Signal, sampled at

about three samples/cycle .....

... and the reconstructed waveform.

( t I

Input Signal. sampled at

about five samples/cycle .....

... and the reconstructed waveform.

Input Signal and sample moments; sampling frequency too low .....

... and the aliased reconstruction.

Fig. 24 Signal sampled at 5 samples per cycle

Fig.25 An aliased sine wa ve

W~veform has a completely different from that of the input

! This is known as an alias, or . ost signal' of an incorrect fre-

. It may however represent rrect wave shape, and often

... right amplitude!

\i Most modern oscilloscopes are equipped with a so-called Autoset feature that automatically selects the proper deflection factor and

tilllebase setting once input signals are connected. This automatic set"'-----'I lJpalso helps to avoid aliasing!

: In cases where the signal tre-

.quency varies so much that a given timebase setting is correct at one

•. ment but shows aliasing at

er moment (or for another

of the signal), peak detection paragraph 2.2) can be used to

... the true amplitude of the signal moment

et the genuine picture of complex signals, the analog of a CombiScope is to be preferred. After all, aliasing is ~-:'1) impossible in the analog mode!

Real-lime sampling and equivalent time sampling

Digitization as described so far is called real-time acquisition or realtime sampling. All samples are acquired in a fixed order, the same order in which they appear on the scope screen. Only one trigger

event initiates a total acquisition.

In many applications, the time resolution that is available with realtime sampling is not sufficient. Often the signals in these applications are repetitive, Le. the same signal pattern is repeated at regular intervals.

Fig. 26 Real time sampling

Input Signal

'2 3 4 5 6 7 8 9 10 .

Sampling Moments

I I I I I I I I I I I I I I J I I I I I J I

Fig. 27 Sequential sampling display build-up

For these signals, oscilloscopes can build up a waveform from groups of samples, acquired in successive cycles of the signal. Each new group of samples is acquired from a new trigger event. This is called equivalent time sampling. In this mode, after one trigger event the oscilloscope acquires a small portion of the signal, for example five samples, and puts these in memory. Another trigger event is used to acquire another five samples which are written to different locations in the same memory, and so on. After a number of trigger events, enough samples are held in memory to reconstruct a complete waveform on-screen. Time equivalent sampling gives a high time resolution at high timebase settings. This makes it appear as if the instrument has a virtual sampling speed or equivalent sampling rate that is much higher than the actual sampling speed of the digitizer.

Equivalent time sample rste

Equivalent time sampling enhances the time resolution of an oscilloscope by reconstructing a repetitive waveform from different cycles.

As an example, consider a DSO with the timebase set to 5 ns/div., displaying 50 samples per division. The equivalent sampling rate can be found by:

equivalent sampling rate = 50 = 50

~- = 10,000 MS/s = 10 GS/s

5 ns 5'10.9

The result is that the display builds up with samples that appear in a fixed sequence, the first one at the far left side of the screen, from where new samples build out the waveform to the right.

The number of acquisition cycles, and thus the number of trigger events, equals the record Sequential sampling allows for post-trigger delay, but cannot provide pre-trigger information. Filling a record at fast timebase settings takes very limited time, and is faster than when random sampling

is used.

9

This equivalent sampling rate is an indirect way of indicating the horizontal resolution at high timebase settings. It also indicates the sampling speed that would be needed to get the same time resolution if real-time sampling was used. This equivalent sampling rate is far higher than the real-time sampling rates that can be achieved at present.

Two different techniques are used for time equivalent sampling: sequential sampling and random sampling.

Random sampling

In instruments using random sampling, a group of samples is acquired at a random moment, irrespective of a trigger event. These samples have a known time given by the sampling clock. While samples are taken and stored continuously, the instrument waits for trigger event to occur. If one a timing system makes a time measurement until the nextsampling moment. Because the sampling interval is fixed, this time measurement enables the scope to calculate the location in memory for all acquired samples (see Fig. 28).

Once all samples of the first Chi

acquisition are stored, a new group sig of samples is acquired and a new sig trigger event is awaited. As soon as Chi one occurs, the new time measure- fixi ment gives the positions of these on new samples. Hopefully they will fall

in empty positions in memory, in • hiG between the positions filled during in i

the first sequence. This way, the are

trace is built up from clusters of 510

samples that appear at random . pal

positions along the horizontal axis. to-

For the fastest timebase settings, car filling a complete record using • the random sampling takes much more on time than when using sequential.· cia sampling, because it is a matter of

statistics to get all the empty tin

memory locations filled. A great tri~

benefit of the random sampling .: wil

technique is that it allows for pre-act trigger as well as post-trigger delayThl to be used.filh Thi )Cal

Sequential sampling

With sequential sampling, samples are acquired in a fixed order, from left to right across the screen. Each sample is acquired as the result of a new trigger event. To fm a complete record, as many trigger events as memory positions are required (see Fig. 27).

The first sample is acquired right after the first trigger event, and is stored immediately. The second trigger is used to start a timing system that causes a small time delay ,i'l.t before the second sample is acquired. The time resolution in the trace memory equals the small delay ,i'l.t, and may be smaller than 50 pico-seconds. After the third trigger, the timing system causes a time delay equal to 2,/1.t before the third sample is acquired; and soon.

Each new sample 'n' istaken.;. after a slightly longer delay (n-1 *~t with respect to a similartrigger . event.

lnput , Signal

... ~t~1 ~,

Sampling Interval

Reconstructed

waveform

after successive timebase sweeps

I Third weep

J_t3

Trig. moment

4

t4

Charge Coupled Devices

Some oscilloscopes use a Coupled Device or CCD as analog storage medium, an anashift register. These can be seen an array of small cells, each with capability to store an amount of charge. Such a packet of represents a sample of the At the command of a clock the cells can transfer their from one to the next, in a direction, like firemen passing

buckets of water.

The CCD can be used with a sampling clock to shift analog information. Once all cells

filled, the fast clock stops and a clock is used to shift the out into a standard analogconverter. This converter

now work at much lower speed; acquisition speed is dependent on the speed of the CCD input

If the sampling clock runs conand stops as soon as a comes, all cells of the CCD contain samples that were

before the trigger moment. entire CCD will therefore be with pre-trigger information.

is very valuable to study

of system behavior.

Single-shot applications

The prime difference between analog oscilloscopes and DSOs is in the capability of the latter to store waveform information. This makes the DSO especially valuable to study phenomena that have a low repetition rate, or that are not repeated at all, the so-called single-shot signals. Examples of these are in measuring the in-rush current of an electrical system, or measurements during destructive tests which can typically be made only once. In fact, nonrepetitive or single-shot signals occur in many systems. Although analog scopes often have a singleshot capability that allows one timebase sweep to be made, digital storage oscilloscopes are secondto-none in acquiring detailed waveform data. With single-shot acquisitions, the instrument first has to be prepared or armed for triggering. A dedicated control is usually provided for this, labelled 'single' or 'single - reset'.

Display types, raster scan vs. vector scan

At the beginning of Chapter 1, we saw that the CRT is the heart of any oscilloscope. In the CRT as it is described there, deflection of the

electron beam is achieved by applying a voltage between two deflection plates. This is called electrostatic deflection. This deflection system can work over a very wide frequency range, from DC upwards. This is used in analog oscilloscopes where the input signal is applied, after attenuation or amplification, continuously and directly to the deflection system. Analog oscilloscopes are therefore often regarded as the most 'trustworthy' instruments to measure signals; what we see on the CRT screen is what is actually going on in the system under test.

Deflection is done by the input signal and the timebase; the two together guide the electron beam to the position on screen that needs to be highlighted. This type of display is called a vector scan display.

In digital storage oscilloscopes, waveforms are acquired and written in memory before being displayed. In some DSOs, another type of CRT is found, similar to the one found in PC monitors and television sets. In these CRTs, the beam is deflected by means of a magnetic field, imposed by coils mounted on the outside of the CRT. This is called magnetic deflection. It can work

Fig. 28 Random sampling trace buildup.

with a limited range of deflection Pre-trigger, post-trigger high speed to display the jitter. At fr
frequencies only. These tubes are In the analog oscilloscope as we the detection of a trigger, a post-
therefore driven in exactly the same saw in Chapter 1, each time base trigger delay timer is started. This
way as TV screens: lines are written sweep was initiated by a trigger can be set to count for the duration
at a fixed frequency from left to event. As a result, signal behavior of about a full cycle. After the preset
right, close to each other. Writing can be studied from the trigger delay time has expired, the oscillo-
one complete screen (a full field) moment onwards. scope starts acquiring the signal. In
may take 500 lines or more. The In many applications, the inter- this example that is just before the
DSO calculates which points on the esting part of a waveform is not next riSing edge of the squarewave.
screen need to be highlighted, and found immediately after the Signal Since the delay timer is a very
when the writing system reaches detail that allows for stable trigger- stable crystal-controlled digital
such a point on-screen the beam is ing. This may occur later, or even clock that works independently of
intensified. This type of display is earlier. the signal under test, any jitter in
referred to as raster scan display. It For example, if a semiconductor the signal will show up in the
can work only in a digital storage is turned on, the output signal may acquired edge as instability. This
oscilloscope, not in an analog be large in amplitude and can be means that in successive acquisl-
scope. What we see is not the input used to trigger the scope. However, tions the edge will be found at dif-
Signal, but is a reconstruction on- if you want to study the small input ferent moments (different positions
screen of data describing the input Signal that causes the semiconduc- on-screen) with respect to the trig-
signal, acquired at an earlier tor to start conducting, this signal gerevent.
moment in time. may be too small to trigger the
Lately, DSOs equipped with a oscilloscope accurately. This calls Trigger pOSitioning
Liquid Crystal Display (LCD) have for an oscilloscope with a so-called Oscilloscopes that offer pre-
also been introduced on the market- pre-trigger view capability: the trigger and post-trigger delay 't
place. These displays take less capability to trigger on a signal (in capabilities must have some means
power than CRTs and are therefore this case the large output signal) to control these. This can be in the
ideal for portable instruments. We'll while displaying the signal before form of a trigger position control,
see a very good example later in the that trigger moment. This allows for Which allows the trigger position to
Fluke ScopeMeter, which can work a detailed multi-channel display of a be shifted across the screen or
for several hours on one small set system with input and output sig- across the acquisition record. 0
of batteries due to the low power nals to see what causes the system On some instruments the trigger tc
drawn by the display. to respond. position can be set to a limited is
In other situations, you may want number of pre-programmed values 9
to study in great detail a part of a only, for example at the beginning,
2.2 DSO functions and Signal that takes place a while after the middle and the end of the acqui- s:
controls the trigger event. For example, to sition record. It is very convenient ci
find the amount of iitter in a to have wider-ranging control over •• s:
The typical features of a DSO squarewave, an oscilloscope with the trigger position. The PM 3394A •.. S!
come with some controls that are post-trigger delay or post-trigger therefore allows the trigger moment ni
not found on an analog oscillo- view capability can be used. The to be set at any position in the total re
scope. We wi1llook at the most oscilloscope is then triggered on an acquisition record. The trigger tr
common of these controls. edge, and the timebase is set to a moment is also continuously tc
variable. 01
fl. Fig. 29 A sinewave with superimposed glitch

Glitch capture

Fig. 29 shows a sinewave that is S( distorted by a fast glitch or spike. IN This can be the result of cross-talk fa from other circuits or wiring close hr

to the system under test. These In

glitches are often the cause for a tir

system to malfunction. Can we then GC

find them with an oscilloscope? ·.··.11E

With an analog oscilloscope, . ar

these glitches will only be seen if •• pl. they are repetitive and synchronous t.h with the main signal (in this exam- .•. 11 pie: the sinewave). Or if we're lucky 1h and many of these glitches occur, ·11E we may be able to see them as a .ot

At t-

of in

is sldlt-

Cltifiier 2 Digital Storage Oscilloscopes

~" "

::",:" .. ' around the main signal.

Iy these glitches happen

now and then, and are not nous since they originate ',"., -e- ",..., another system.

fk<.·:~)O can we find them with a

Not necessarily, we first have sure that the oscilloscope

nn"";'''>'11 to catch these fast · hes.

we have seen, a DSO takes

· pies of the input signal at spa. moments. The time between pies depends on the timebase

· ng, as described at the begin-

.. of this chapter (see 'horizontal . If a glitch is narrower c,, : .' the time resolution, it is therea matter of luck if it is captured Peak detection or glitch capis the answer to this.

peak detect, the osclllo-

}.,·.: .• 'vV,J\> monitors the amplitude of the

W>lIIIJT""m continuously and tempostores the peak amplitudes, positive and negative, by

s of peak detectors. At the

a sample is to be displayed, the ,."n'''n'~ of either the positive or peak detector are digitized the detector is cleared. Dissamples thus alternately give positive or the negative peak as "'"'.' . ."."'"" detected in the signal since digitization. Peak detection to find signals that could

ise be missed due to using

too Iowa sampling speed, or that would be distorted as a result of aliasing. Peak detection is also very useful for capturing modulated signals like the AM waveform in

Fig. 30. For this kind of signal, the timebase has to be set to match the modulation frequency, which is typically in the audio range while the carrier frequency is typically

455 kHz or beyond. Without glitch cap-ture the signal cannot be acquired correctly, while with glitch capture a picture similar to that on an analog scope is seen.

Peak detection is performed in the scope by means of hardware (or analog) peak detectors, or by means of fast sampling. An analog peak detector is a dedicated piece of hardware that stores the peaks of the signal as voltages in a capacitor. It has the drawback of being relatively slow; it is usually able to store only glitches wider than several microseconds at reasonable amplitude.

Digital peak detectors are built around the ADC, in which sampling has to take place continuously at the highest possible sampling speed. The peak values are then stored in a dedicated piece of memory, and treated as 'the' sample value at the moment a sample is to be displayed. The digital peak detector has the advantage of being as fast as the

digitization. The instrument we use as an example in this booklet, the PM 3394A, captures glitches as narrow as 5 ns at the correct amplitude, even at very low timebase settings like 1 second per division!

Roll mode

The DSO as we have seen it so far displays a waveform in a similar way as it is seen on an analog oscilloscope: starting from a trigger event, the instrument acquires samples and stores them at successive positions in the acquisition memory. Once the last position in the memory is filled with new data, the acquisition process stops to enable the instrument to copy the waveform data to the display memory. During this time no new data is acquired, just like the analog oscilloscope is not able to display a trace during the time its timebase is being reset.

For low-frequency applications, where changes might take minutes rather than microseconds, DSOs can be used in a way that allows for a fully continuous display: the Roll mode. In this mode, samples are acquired and are immediately copied to the display. However, these new samples appear at the right side of the screen and the existing waveform scrolls to the left (see Fig. 31). The oldest samples disappear once they reach the left

Fig. 30 An AMmodulated signal captured without and with peak detect mode.

Fig. 31 Roll mode

Roll Mode.

.................. : .. -

CHI O. 5 :v~

MTB 2.000; ROLL

.chl+

side of the CRT. This way the display always shows a continuous overview of the latest signal behavior over time.

With this roll mode, the oscilloscope can be used to replace a paper chart recorder for displaying slowly changing phenomena, e.g. chemical processes, charging and discharging cycles of batteries, or the impact of temperature on a system's behavior.

to zoom in on small details.

In DSOs the displayed waveform can be expanded in more and smaller steps. Usually the time base magnifier operates in powers of 2: magnifying by *2, *4, *8, * 16 and so on.

Vertical magnification is convenient as a replacement for increased vertical sensitivity once a waveform is stored in memory, e.g. for a Single-shot acquisition.

Display magnification

In analog oscilloscopes, the timebase can be magnified ten times

Special triggering

The storage power of a DSO makes it an extremely useful tool for

Fig. 32 Glitch triggering

I TRIGGER
ch4: puls~ 2.01us MAIN TB
r--
edge tv J
logio
r- state J
pattern
glitoh
r ]
n u
T - £1
) U >tl
t- U <tZ
range ~]
'il
RANGE j
This glitch triggered the oscilloscope. O.50us
ather edges had no effect. 4.50us
r--
CH4 2.00 v~ MTB50 .Ous - ~.' 709v ch4 capturing signals that happen very rarely, or once only, like single-shot events or a system locking up. Capturing these signal calls for a versatile triggering system to detect that special condition to initiate an acquisition. Often edge triggering is not enough for this, so some additional triggering capabilities have been developed for this. We'll discuss some of these here.

- Pattern triggering

In loaic hardware, signals are transported by means of a large number of parallel lines. The instantaneous status of the total hardware is described by the status of anum" ber of these lines at any given moment. Recognition of the status of the hardware requires an instrument to sense a number of those lines. With pattern triggering anum" ber of lines, e.g. four lines, can be monitored. At the moment a user-

described pattern (e.g. HHLH) or c

word is detected, the oscilloscope tr

triggered. Since pattern triggering s

designed for use with digital logic, d

the individual lines can be rnoni- tl

tared for high (H) or low (L) state, it

or can be ignored (don't care or x). 0

- State triggering ir

Logic hardware is often built c

around a central clock system. All S'

hardware in it stores its input sig- ir

nals at the command of this clock d

system. Test instruments should do 0

the same. When using state trigger- tE

ing, input signals are treated as ir

pattern triggering, but one of the fE

input Signals is now seen as the c;

clock signal. The oscilloscope trig- tl

gers if the input word on three

inputs, stored in the scope on the rising or the falling edge of the clock, matches the user-programmed trigger word.

- Glitch triggering

With glitch triggering the scope can trigger on 'short pulses, like glitches and spikes that may cause a system to malfunction.

If a system was designed to work with signals ranging from DC to a specific frequency, signals of higher frequencies may be induced in the wiring, e.q, as the result of

Trigger settings; Ch 1 , positive edge Event delay on Ch2, count ~ 4

channel 1

_j

1rigger

(~ Event

Counting

lve

channel 2

This edge initiates the acquisition ....

. .. in addition, pretrigger view brings the edge on screen

I

CHl :0. 5V':-

E-4

: MTB 2.00u~ -2.'14dv : .cn t+

............... .... _ ..

-talk or pick-up of high-power nts. The oscilloscope can be . to trigger on pulses that have a

I n shorter than half a cycle of

. highest allowed frequency, since be assumed that these do not r during normal operation. Another area for applications is gic hardware, where all state

ges are synchronous with a

m clock. As a result, all pulses hardware should have a

on that is an integral multiple ri"""~"'tit'l system clock cycle. In such sys-

s, errors can sometimes occur bination with pulses of a difduration. The oscilloscope now be set to trigger on pulses have a duration shorter than

lilt

: All

one clock cycle,

- Time qualified triggering

This allows an oscilloscope to trigger in any of the ways given above, if the condition is met for a certain amount of time. This can be for a minimum length of time ('if valid longer than .. .'), for a maximum length of time, or for a period of time ranging from a minimum to a maximum value, Time qualified triggering is very useful to trigger on events not matching the normal behavior of a system. It can also be used to find interruptions in a signal that is expected to be active continuously.

- Event delay

Event delay can be used to trigger on a combination of input signals, where one is used to delay the acquisition. The trigger cycle is initiated by a main trigger signal, usually one of the input channels. Once this is received, the oscilloscope starts checking a second input signal (this may also be the same as the main trigger signal, at a different level) and counts the number of 'event triggers' on this input signal. Once a preset number of events has occurred, the acquisition starts,

See Fig. 33.

Typical applications are in serial data lines, control systems and mechanical environments,

Trigger input signal

Trigger signal after n~cycle counting using n~5

\

/

Every n-th pulse is selected and applied to the rest of the trigger system

Fig. 33 Event delay

Fig. 34 N-cycle triggering

- N-cycle samples are acquired before and known mathematical techniques. Sl
This can be used to select every after the transition. If the display is These will be described in Chapter 6. o~
nth occurrence of an input signal expanded horizontally, making the With an analog oscilloscope, the bi
and apply this to the normal trig- distance between the acquired sam- user had to perform measurements til
gering system. It is useful if a signal pies larger, we would lose signal manually, for example by interpre- is
is distorted by a sub-harmonic, that brightness. Oscilloscopes therefore ting the displayed curve, counting H
is if the signal is periodic but if not calculate interpolated or display divisions to find elementary ampli- til
all cycles are exactly the same, for samplesto keep the number of tudes or time intervals, applying Fe
example in a system in which a fixed displayed samples high. With larger mathematical descriptions and b,
frequency is used, but every 12th expansions, it becomes more calculating the measurement result. rh
pulse is wider than the others. The important to display a continuous These processes are possible for (s
scope can now be set to respond curve through the samples rather basic waveforms, although only at a 5\
only to one of these wider pulses by than straight lines between them. moderate accuracy, while for more
selecting 'N-cycle '" 12'. For this, the oscilloscope can use complex signals they are much a~
sine interpolation, in which a best-fit more difficult and may require some eli
Waveform storage sinewave of changing amplitude and guesswork. PI
Once a waveform is in the frequency is assumed through the Once a digital storage osculo- a~
scope's memory, it can be copied to acquired samples. Sine interpolation scope has acquired a waveform, all
a back-up memory, also known as a allows for a reconstruction that has the information is available to allow D
registeror memory location, for the natural smoothness of an analog these measurements to be calcu-
analysis at a later stage or for refer- oscilloscope, even if the number of lated automatically, giving a much th
ence or comparison purposes. samples per division is low. more accurate and reliable result, 01
DSOs are equipped with a number To check which samples are truly which is obtained much more d,
of such memories which allow from acquired, a dots mode is usually quickly and easily. in
2 to 200 traces to be stored. Memo- provided that disables any inter- Most DSOs have the capability to tir
rles can be organized as trace mem- polation. Once selected, we only see perform two or more measurements di
cries, in which each trace of a multi- the samples as separate bright dots on input signals simultaneously, on
channel acquisition is stored lndlvi- on-screen, without any intercon- one or more channels. This allows w
dually, or as record memories, in nections. for comparisons between signals, cl
which complete acquisitions are e.u. a comparison between the ee
stored simultaneously, regardless of Windows mode amplitudes of the input and output zc
the number of channels. The latter When signals are compared, e.q. signals of an amplifier or attenuator. of
has the benefit of keeping all rele- a newly acquired waveform is com- It is very convenient if the scope p(
vant timing information together. pared with signals that have previ- also allows measurements to be 01
A large number of memories is ously been stored, it can be useful performed on both stored wave- III
also convenient for engineers wor- to show the traces on seperate parts forms and newly acquired signals, er
king in the field. These can be used of the scope screen. To do this, a allowing actual behavior to be com- til
to store all relevant waveforms windows mode may be provided pared with standard Signals, or of
during field measurements, enabling that splits the screen in two or more showing the influence of time or IS
hard copies to be made later, or parts for display of different traces. modifications made to the system. st
waveforms to be transferred to a By reducing the vertical display The most complete set of mea- IJi
computer for further analysis. amplitude, windows mode enables surements also provides measure- c!
the use of the full dynamic range of ment results in a statistical format. s~
Display algorithm, interpolation, the analog front end of a OSO. This means that the minimum value,
dot join Measurement accuracy can thus be the maximum value and the mean In
What we see on the screen of a optimized, while reducing the (or average) value of a specific mea- th
OSO is a reconstruction of the sig- displayed waveform amplitude. surement can be presented at any JC
nal, assembled from a number of moment, relating to a longer acqui- In
samples in the scope's memory. sition period. This capability can lI~
These samples are displayed and a 2.3 Automatic measure- reveal trends in a system's Ih
line is drawn through them. This can ments and processing without the need for continuous Ilc
be done in several ways. The easiest monitoring of the display. pi
way to draw a line between points is Measurements It is important to remember that si
by using straight lines. This is Oscilloscopes are used to display any oscilloscope measurement is w
known as linear interpolation. This waveforms and measure waveform basically an analysis of the 01
gives sufficient waveform recon- parameters, such as peak-peak am- data. The measurement result w
struction as long as the samples are plitude, RMS amplitude, DC level, fore applies to the acquired wave- sc
very close together, e.c. 50 samples frequency, pulsewidth, risetime etc. form which is stored in the scope's
per division. This makes it possible For any waveform, these measure- memory. This means that the 8)
to see the edges of a Signal, when ments can be obtained using well of the scope can influence the mea- til
I
,'\/

s. " "",., surement result. For example: if the ,oscilloscope is set to a slow time-

., •• ,,' base setting like 1 ms/div., a risetime measurement on an edge that

ne- • '.," expected to be between 50 and

ing • ','". ·"."."' •• ,'00 ns wi!! be dominated by the

lpli- resolution of the acquisition.

g such a measurement, the time-

should be set to display the ng edge with sufficient detail ',"'., I~,~" useful risetime), for example

, ns/div.

, .. ",., In Chapter 6, we'll look in detail different measurements, their initions and how they are ,""rTnrmed, both manually and matically.

, " processing

~u- ' " , The waveform data acquired by

uch oscilloscope contains a wealth

ult, ' information. The usual format for

e- ... ~ display is as a waveform show-

voltage on the vertical axis and eon the ho rizontal axis - the V-ti lay.

\ Another way of displaying the ""HIOTnrrn data is by plotting one

als, nel against another, so that for

<'·"rii]I'h displayed data point the horiposition represents the value channel, while the vertical ltion shows the value of the

er. This is known as X versus V e, or simply X-V mode. It

the user to see the phase or ing relationship between signals related frequency. The X-V mode

.'T".,iC',,'" xtrernely useful for testing phase > itters and filters, but also in comination with motion sensors to

ck for vibrations in moving

, ms.

-.': , The benefit of a DSO used in X-V e over an analog scope is that rn"''_'L .rne bandwidth of the DSO is the full uisition bandwidth of the instruwhile an analog scope can be in X-V mode only over a limfrequency range. On the other the X-V mode of the OSO disthe samples as contained in a , gle record, which describes the

rm over only a limited period e (the duration of the record), ile the X-V display of an analog pe is a continuous, 'live' display.

There are many other ways to Irr"l."tlvt·~~·valuable information from

chI pkpk ch1freq"

1-

v:

J!i<:

" Maximum" values

1-

the acquired waveform is fully described by 256 different voltage levels. The resolution can be increased by averaging successive acquisitions. The more acquisitions are used to calculate an averaged waveform, the higher the vertical resolution becomes. Each time the number of acquisitions is doubled, an additional bit is added to the resolution.

- Envelope mode

With signals that are changing over time, for example when changes in amplitude are expected or when jitter occurs, it is useful to see not the instantaneous waveform, but instead the behavior of the waveform over a number of acquisitions. When the envelope mode is active, the oscilloscope builds up the display by storing the minimum and maximum values for each record position over successive acquisitions. The resulting display shows the cumulative effect of the long-term changes. It allows for measurement of long-term jitter or long-term amplitude changes.

- Digital filtering

Filtering a waveform is a processing capability which, by processing the acquired waveform data, reduces the bandwidth. The term 'filtering' refers to the fact that this processing function has the

Fig. 35 Screen with peak-peak and frequency measurements in a statistical format, giving minimum, mean and maximum values over time

the data in such a way that the information contained in it is presented in a more useful format. Such manipulations are generally known as mathematical processing of the waveform. This is sometimes done by a computer, after the waveforms have been transferred to the computer for further analysis.

More versatile DSOs have this mathematical power on-board, using either the main processor or an additional Digital Signal Processor (OSP).

- Averaging increases the resolution

Averaging is used to reduce noise superimposed on a signal, by combining successive acquisitions. Each sample of the resultant waveform is calculated from the samples at the same position in successive acquisitions by means of averaging the sample value.

Since noise, by nature, is different with every new acquisition, samples in successive acquisitions will have slightly different values. These differences are reduced by means of averaging, resulting in a smoother waveform, while leaving the bandwidth unaffected. However, when averaging is applied, the oscilloscope takes more time to respond to changes in the signal.

Most DSOs work with an 8-bit vertical resolution. This means that

cl1i~ter 2 Digital Storage Oscilloscopes <

?w!"

Fig. 36 A jittering pulse, with and without use of the envelope mode

,CJ!1 :

CHl·. 1 :V~ :

: MTB 11 .• onus -Z;Z6dv:

same effect as a low-pass filter applied to the oscilloscope input.

Digital filtering is achieved by averaging each sample in a record with the adjacent samples in the same record. As a result, noise in a waveform is reduced, but so is the bandwidth. Unlike averaging, filtering reduces the bandwidth in order to reduce noise, but it can also be used on single-shot signals where averaging would require multiple acquisitions on a repetitive signal.

- Waveform comparison, template testing

Stored waveforms can be dis-

played together with newly acquired waveforms, for example to compare the behavior of a known good unit with that of a faulty one. In many situations, these signal comparisons are made to see if a system complies with its specification, for example during manufacturing testing. This can be done easily and fully automatic by oscilloscopes with a 'pass/fail' test capability. A standard signal, including margins, is stored in one of the registers. This is known as the template. The oscilloscope now acquires the signal from individual parts or components and compares each

new acquisition to the tem-plate. If the signal falls within the template, the scope responds with a 'pass' indication; if any part of the signal falls outside the template, a 'fail' is reported.

- FFT or Fast Fourier Transformation.

The Fast Fourier Transformation is a mathematical operation that extracts the individual frequencies contained in a signal, and displays them with their individual amplitudes.

FFT is a useful tool to find the amount of distortion in a signal, to

Fig. 37 Waveform and its frequency transformation

dentify frequencies contained in a mplex waveform or to look for k between systems.

". Multiplication of wavetorms

•...•••.. The most important application this is in measuring the electrical r, defined as the mathematical uct of the current and voltage.

. allow power to be measured, an loscope must be able to

ire both the voltage and the rrent waveforms and multiply The resultant waveform

ows the instantaneous value of liF"'''''',''''r against time. This measurent is important in tests on power pplies, power amplifiers and volt-

regulators where currents can

i significant, and the power han.. by components can be critical.

ration

ii ntegration gives the area under a "rve, which allows the total effect

. individual signal contributions .f time to be seen. For example < . total effect of successive chargcycles on the amount of charge . by a capacitor, or the cumula...•. effect of power cissipated by a ponent.

> r tests on mechanical systems,

icles .'. ..•. sducers that generate a voltage

)Iays . ortional to the acceleration are

)Ii- "r<'."N,rj. Integration of these signals

the speed.

ss'

ntiation

ifferentiation shows the rate of e of a signal, for example the rate of an electrical signal,

is the differentiated curve of against time.

measurement, dissipation actual example of matnemat'. rocessing capabilities is the ing dissipation measurement. r ,.,., .. ,." ... - .... component handles a certain of power, defined as the

ct of the voltage across the ponent and the current passing

gh it. In the design of a coma certain limit is set on its handling capability, usually on the temperature rise that

ower level will produce over a IW"''''· .... _,._ period of time. The amount

(= energy) dissipated by a

component is the product of the instantaneous power and time.

The heat produced can be determined by measuring the voltage and the current and multiplying them, which shows the instantaneous value of the power in watts at any moment. A second mathematical process can now be used to integrate the power curve to give the total power dissipation in wattseconds. The equivalent of 3,600,000 watt-seconds is 1 kWh, the unit for which we are charged

by the energy company!

2.4 Interfacing

InterfaCing

I n many situations the information, contained in the scope needs to be transferred to a PC. In other situations, the oscilloscope may need to be controlled from a PC. These functions require both instruments to have the ability to communicate. This means the oscilloscope has to be equipped with communication hardware and supporting software. This is referred to as an 'interface'. Two types of interface are commonly used: RS-232 and the General Purpose I nterface Bus or GPIB, also known as IEEE-488 bus. These communication interfaces are optional with most oscilloscopes.

RS-232 is a serial interface, commonly found on PCs for communication using modems, and for the connection of devices like a mouse, printer etc. Each device connected to the PC needs its own RS-232 interface on the PC. This means that only a limited number of instruments - often only one - is connected to a PC.

Many software packages use serial communication since this requires minimal changes to the PC and can be used with relatively simple cabling. It is therefore easy to provide this software on-board the scope, allowing the use of any available PC to store waveforms for back-up purposes etc.

The GPIB is a parallel bus, designed for use in instrumentation systems. It allows a large number of

instruments tobe connected to the same bus, all at the same time. The bus allows instruments to call for the attention of the controller at any moment during the test protocol, for example if a measurement error is found. The PC that controls this bus can be a dedicated GPIB controller, although today a standard PC with a GPIB card is most commonly used. A GPIB card is not standard in the PC, and therefore needs to be added.

Software available for GPIB systems typically creates a complete test environment, integrating a variety of instruments into a single test system.

Printers, plotters, hard copy

In many field tests, engineers require some form of hard copy of what was measured. This can be used for reference at a later stage, for example when new adjustments need to be made or when malfunctions are reported.

Most DSOs can make hard copies using a digital plotter or a printer like those used with a PC, and some can also drive a paper chart recorder since these are still found in many labs. The hard copy capability usually comes with one of the interfaces which are available either standard or optionally.

If no printer or plotter is available at the time or location where the measurements are made, it will be useful to have a large number of memories built into the scope, allowing all the required information to be stored so that hard copies can be made later.

Making single-shot phenomena repetitive

I n many applications, signals other than clean sinewaves or squarewaves are found, like heartbeat signals in medical electronics or the 'bouncing' of switch contacts.

Testing the operation of such systems should preferably be done using signals closely resembling those encountered in actual practice.

For this purpose, arbitrary waveform generators have been developed. With an ARB generator, out-

2 Digital Storage Oscilloscopes

put signals are described as a function of time by means of an array of data words. The output voltage is generated by feeding the successive data words to a Digitalto-Analog Converter (DAC).

An ARB generator makes an excellent combination with a digital storage oscilloscope. The DSO has the unique capability to capture a section of the actual signal and

store it as an array of digital words. This record can then be transferred to an arbitrary waveform generator, so that the captured signal can be reproduced whenever and as often as required. The signal can even be scaled in amplitude, the frequency can be modified, or the data can be sent to a computer and modified, so that the ARB generator provides a modified 'recording' of the original signal.

To make it easy to use the combination of an ARB generator and a OSO, a number of Fluke CombiScopes are equipped with a 'direct dump' capability. One cable between the oscilloscope and the ARB is all that is needed to transfer a captured signal to the generator so that it can be reproduced.

(,{'Analog, Digital or Both?

1 Introduction

When the purchase of a new oscilloscope is considered, users

i usually ask what kind of instrument should choose. What bandwidth .. is needed, can an analog scope do the job or should it be a DSO?

..... An equally relevant question that ·is sometimes overlooked is: 'Is a SO capable of doing the job?' In

. 'other words, will the 050 be able to make all the measurements which are used to, using an analog

Unfortunately, there is no 'ghtforward answer to all these .. uestions. The answer depends ntirely on the measurements to be and on the signals to be

red. In some Situations, the

. of the DSO are obvious.

> n others, an analog oscilloscope .••..•.. utperforms any DSO available.

s look at some of the main

. .2 The simple repetitive

signal

..... The simple repetitive signal can Iy be viewed equally well on analog and digital storage Iloscopes. There are advantages limitations to both. For the

i alog scope, display of very low.. uency signals is difficult

use the persistence of the CRT too short. As the light output is

rsely proportional to the sweep d, low-repetition rate signals

•• ••···• •••• ;::~L fast rise and fall times can be

to see. Digital storage

Iloscopes have brightness

h is independent of the sweep d and signal repetition rate.

can be either an advantage or a ntage, depending on the

content.

•.•.•..•. For displaying fast edges of

. waveforms with sufficient on rate, there is very little rence between the performance

of a digital storage oscilloscope or an analog scope. Both will visualize the waveform equally well.

If measurements of signal parameters are to be made, the DSO has the benefit of automatic measurement capabilities. With an analog oscilloscope, the user has to set up cursors or interpret the display to get the measurement result.

If adjustments have to be made, the analog oscilloscope is often preferred. This is because the rea/time display of the scope shows the voltage at the input at each moment. The update rate (the number of times per second a new trace is written) is high - at high sweep speeds it can we!! be beyond 100,000 sweeps/s. Any change in the signa! is therefore displayed immediately. Some signal changes are recognized in the display as intensity changes of the waveform.

The display of a D80, in contrast, shows a reconstruction of the acquired waveform. The number of acquisitions per second is well below 100. On some scopes, no more than 5 acquisitions are made each second! As a result, there is a clear time lag between changes in the signal and the moment they are shown on the screen. This is an important drawback of a DSO when making adjustments to a system.

3.3 A more complex repetitive signal

The simple repetitive signal as discussed above is seen in many areas of electronics. However, it is very common for this signal to M used as a carrier of information. For this purpose, the sigpal can be modulated in amplitude, in frequency or in phase. The modulation can take many forms, e.g. a sinewave, a pulse, a ramp or a 'staircase'. Several types of modulating signal and more than

one type of modulation are often used simultaneously.

A good example of a complex, everyday analog waveform is the composite video signal. This is made up of many components of different frequency and amplitude.

It includes pulses as well as sinewaves, plus more sinewaves which are phase-shifted for color information. In a situation such as this, both analog and digital storage oscilloscopes have their own specific benefits, each allowing different parts of the signal to be viewed optimally.

For example, the infinite resolution and large number of sweeps per second of the analog scope show the time distribution of the waveform. The intensity variations in the trace are proportional to the amount of time spent at a particular level. This can give a good display of the color modulation. The high update rate (the large number of sweeps per second) immediately shows the effect of any adjustment made to the system.

With a DSO, the limited number of samples and the lack of intensity variations leave a lot of detail undisplayed. Some DSOs may have two or more intensity levels, but these are only relative and, together with the limited display resolution, still cannot replicate the analog display.

Fig. 38 shows an analog oscilloscope display of a composite video signal. Note the large amount of detail and the intensity variations. In Fig. 39, a similar signal is displayed on a typical raster scan CRT of a digital storage oscilloscope.

If only a small part of a video line is to be displayed, e.g. TV transmission test signals included in a specific line, teletext data or the color burst in a specific line, the DSO is preferable. With an analog scope, the relatively low repetition rate, together with the short

Fig, 38 Video signal as seen on an analog oscilloscope

Fig, 39 The same video signal as seen on a digital storage oscilloscope (see text)

Fig, 40 These kind of glitches usually happen very infrequently. They can thus only be captured using digital storage

20niV~

:PKD :

MTB 500u~ -O_Y6dv

CHI

duration of the event itself, easily results in a picture which is too dim to be seen. The DSO has a uniform intensity, reqardless of the repetition rate of the signal, and therefore shows this signal at high brightness.

If a repetitive waveform also contains short duration spikes or glitches, these are impossible to view on an analog scope when examining the complete waveform, When a DSO is used, these spikes are shown by the peak detect mode (see Fig. 40).

3.4 Non repetitive signals and transients

There was a time when many of these signals were ignored as viewing was too difficult and expensive, It required the use of very expensive analog storage oscilloscopes, cameras, longpersistence tubes etc.

If an event happens just once, the analog scope usually cannot cope. This is where the DSQ shows its main capability: capturing that one rare event and displaying it for as long as you like,

That rare event may even be the resultant of a disturbance: the DSO has the capability to show pretrigger information, including the cause of the disturbance, by triggering on the disturbance itself.

Fig. 41 shows the microphonics of a cable, when an open connector was bounced on the table. The trigger moment is indicated by the A-symbol, and the waveform information at the left of it is pretrigger information.

3.5 Analog, digital or both

As we have seen, for everyday use engineers require both analog and digital storage oscilloscopes to get the best view of their signals. For some applications the analog scope is preferable, while for others a DSO is the better choice.

A number at DSQs available on the market are claimed to have 'analog look and feel'. These are

DSOs, although some may

the controls positioned in the way as on many analog

while others have a high rate to mimic true analog capabilities. Or they may have sampling rate together with acquisition record of very limited thus lacking the time

needed to avoid aliasing more complex signals. Of

these instruments not only the advantages of the DSO, also have all its shortcomings. Fortunately, there is one

that really combines the and benefits of both the scope and the DSO in a instrument the CombiScope.

is a combination of a highanalog oscilloscope and a featured digital storage

with easy switching one mode of operation to the

the instrument is set up as a store acquired waveforms, hard copies from it, yet have infinite resolution and the trusted display of the scope when needed. And the CombiScope they can have brightest display, regardless of signal repetition rate. In other the CombiScope combines best of both worlds in just one

n

.... T.

MTB I.OOms -+.16d~

CHI

50mV~

chl-

Fig. 41 Micropllonics of a cable wilen the connector is bounced on a table. It takes digital storage to capture such rare signals

Accessories and Software

4.1 How probes work

An oscilloscope probe is more than just a piece of wire to apply the test signal to the oscilloscope input " it is an integral part of the mea" surement system. Many different types of probes are available, all with characteristics that make them more suitable for specific tasks. Some probes are called active probes because they include active electronic components to provide amplification. Probes without such active components are called pas" sive probes, incorporating only pas" sive components like resistors and capacitors. These usually provide attenuation of the input signal.

We will first concentrate on passive probes,

main specifications, their

on the circuit under test and signal under test. Next we will briefly at several specialist

of probes and other

An important task of the probe is ensure that only the wanted appears at the oscilloscope

If we just used a piece of

it would act as an antenna.

of unwanted interference

be picked up from radio fluorescent lights, electric 50 or 60 Hz mains hum and

the local CB enthusiast! Some noise could even be injected into the circuit under test. So

... first thing we need is screened

Ie. The screened cable on a probe is grounded to the

by the BNC connector and by und lead at the probe tip to

rcuit under test, ensuring good ing.

bandwidth

like oscilloscopes, probes

. a finite bandwidth which has to for. If we take a 100 MHz ''''V'>vUIJIJ and a 100 MHz probe,

the combined response is less than 100 MHz. The capacitance of the probe adds to the input capacitance of the oscilloscope, lowering the system bandwidth and the maxi" mum displayed risetime, t, See under risetime in Chapter 1.3.

Using the formula from Chapter 1.3:

t, (ns) ~ 350/ BW (MHz)

An oscilloscope and probe, each with a 100 MHz bandwidth, each have tr = 3.5 ns. The effective system risetime is given by:

= 4.95 ns.

With a 4.95 ns system risetime, the system bandwidth is 350 14.95 MHz = 70.7 MHz.

Fluke supplies all oscilloscopes with probes that also give the specified oscilloscope bandwidth at the probe tip. As we can see from the above calculation, this calls for a probe that has a much higher probe-only bandwidth than that of the oscilloscope.

Loading effect

When making a measurement, we often consider the measured voltage to be exactly the same as when no connections are made.

Each probe has an input impedance, consisting of resistive, capacitive and inductive components. Connecting such a probe will affect the circuit under test because of the extra load imposed by-the probe. Therefore the probe characteristics must be considered when looking at the measurement results, as well as the test circuit impedances.

Some probes have no series resistance. Basically, these consists

of a length of cable and a test tip. Within their operating frequency range or useful bandwidth, there is therefore no attenuation of the signal. These probes are referred to as 1:1 or xi probes. They have a loading effect on the circuit because they connect the oscilloscope input impedance and their own capacitance (including the cable capacitance) straight across the test point See Fig.42.

The capacitive loading effect becomes more significant as the signal frequency increases. A 1:1 probe typically has an input capacltance from around 35 pF to over 100 pF, depending on the cable type and length, probe body construction etc. This will load the circuit under test with a low impedance. A 1:1 probe with an input capacitance of 47 pF has a reactance of only 169il at 20 MHz, making this probe unsuitable for work at these frequencies.

Attenuating probes reduce loading effect

We can reduce the loading effect by adding impedance in the probe in series with the scope input impedance. However, this means that not all the signal voltage appears at the oscilloscope input as we now have a voltage divider.

Fig.43 shows a simplified probe equivalent circuit. Rp and Rs form a 10:1 divider, where Rs is the scope input impedance. The compensation capacitor, Ccomp, is adjusted to match the probe to the oscilloscope. This ensures that the correct trequency response curve is main" tained at the probe tip. The frequency response of this type of probe is now much wider than that of a 1:1 probe.

The standard input resistance of oscilloscopes is 1 Mil. This calls for a series resistance in the probe of

9 Mil, giving a probe tip input impedance of 10 Me!. at low frequencies.

Probe

... ()

Probe cable incl. cable capacitance

Input voltage

Oscilloscope Input

15 ... 25 pF

Rs = 1 MQ

Fig. 42 Probe equivalent circuit

Compensating the probe

A practical 1 0:1 probe has several adjustable capacitors and resistors to give the correct response over a wide frequency range. Most of these are set in the factory when the probe is manufactured. There is only one trimmer capacitor for the user to adjust. This is called the low frequency compensation capacitor, and it should always be adjusted to match the oscilloscope input with which the probe is used. Adjustment is easy, using the frontpanel signal output of most oscilloscopes. This is marked 'Probe Adjust', 'Calibrator', 'CAL', or 'Probe Cal.', and provides a squarewave voltage output. A squarewave contains many frequency components. When all these components are passed to the oscilloscope at the correct amplitude, the squarewave

Fig.43 10:1 probe cirCUit diagram

signal is reproduced correctly on the screen. Figs. 44 shows the effects of under-compensation, correct compensation and overcompensation.

The effects of over- and undercompensation are very noticeable at higher frequencies. The displayed amplitude of the 1 MHz sinewave is now very inaccurate.

So remember always to check the probe compensation before using attenuator probes. There can be slight differences in input capacitance between channels, so you should always compensate the probe on the channel you want to use it with.

Probe

Rp ~ 9 M!1

Probe cable incl. cable capacitance

Input voltage

Oscilloscope 'Input

Input voltage to the oscilloscope equals O.1x probe tip voltage

Maximum input voltage

The construction of the most commonly used 10:1 probes makes them suitable for a maximum input voltage at the probe tip of 400 or 500 V peak. These probes can thus be used for a wide range of applications at all general signal levels up to hundreds of volts. For higher voltage levels, 100:1 probes with wider voltage rating are recommended.

Probe read-out

Modern oscilloscope probes are equipped with a coding system that enables the oscilloscope to recognize the kind of probe which is connected. Consequently, the osc1110- scope can compensate the vertical deflection indication and all amplitude measurements to avoid confusion. When probes without such a recognition system are used, the user has to rescale all waveforms and measurements to compensate for the probe attenuation.

Ground lead inductance

Fig 45. shows how the ground lead inductance forms a series tuned circuit with the combined probe and scope input capacitances. The tuned circuit is damped by the input resistance.

LIke any other tuned circuit, this one will also ring if a step voltage is applied to it. Excessive ground lead inductance can also reduce the displayed risetime. Fig. 46 show s a fast-risetime pulse connected to the oscilloscope using ground leads of various lengths.

The effect of ground lead inductance can be seen clearly. So keep your ground lead as short as possible, especially with high frequencies and fast-risetime signals.

Safety ground

Because of electrical safety, most oscilloscopes are connected to the safety ground by the power cord. The signal under test may be at the same reference potential, but this is not necessarily true. Before conneetingthe probe's ground contact, make sure it is not going to shortcircuitany part of the system under test! .

II

s t

On the other hand, if the system under test is at the same reference potential as the scope ground contact, this does not imply the safety ground can be used as a signal

return path. The long wiring of the safety ground interconnections have far too much inductance for this purpose. Use the probe's ground lead as the signal's reference connection.

1-

Incorrect squarewave representation ....

Probe undercompensated

1-

CHI: IOQmV~

MTB IOOU:s ·1.:25dv:

e

. ... and 1 MHz sinewave amplitude too low.

: c h i.+

CHi: 100mV~

MTB 50D~s -1.~26dv:

:chl+.

Incorrect squarewave representation ....

Probe overcompensated

: f"-..i-,: __

....

·:······v

: MTS i oou» -1.:26dv:

CHI 10QmV~

.... and 1 MHz sinewave amplitude too high.

. ..:... . , .. ..... .. ~ ...

:chl+

MTB 500ns ·1.:26dv:

CHI· loa,"v~

chl+

Correct squarewave representation ....

Probe correctly compensated

.... and correct 1 MHz sinewave amplitUde: 500 mV pp.

...... : ..... ,',

CHI: lOOmV~

. .

MTI::! IOOuS -1 .:'26dv :

.'----

.c h t+

MTB 500n.:s -1 .:26dv :

,chl+

loa~nV""_'

Figs.44 . The effect of probe compensation on a 2 kHz squarewave and a 1 MHz sinewave

Probe input

Pr-obe input capacitance Ci~10 .... 20 pF

intruducas inductance

Probe input impedance .....

Fig.45 Probe equivalent circuit with ground lead inductance

4.2 Probe types

We have already seen the 10:1 and 1: 1 probes. Here are some more types for general-purpose use.

Switchable probes

These make life very easy as they combine both a 10:1 and a 1:1 probe in a single unit. The 10:1 setting is preferred for its low loading and wide bandwidth, while the 1:1 setting can be used for lowlevel signals at low frequencies.

Attenuator probes

The other attenuator probe commonly used is the 100:1 probe. It has a low capacitance, typically

Ri

Input voltage

Current probe.

As the name implies this probe shows the current waveform in a conductor rather than the voltage. It uses a current-sensing transformer in the probe head. This clips around a cable, eliminating the need to break into the circuit. The signal from the probe head is converted into a voltage, scaled and passed on to the scope input. The scope now shows A/div. or mA/div. The frequency range is over 70 MHz.

Oscilloscopes with mathematical processing capabilities can now, using current probes, handle power measurements by multiplying voltage and current waveforms. More information is given in paragraph 2.3.

. ... and equivalent circuit.

2.5 pF, and 20 MQ input resistance. The voltage rating is much higher at typically 4 kV, making the probe suitable for measurements like on high-voltage converters.

FETprobe

This is an active probe for higher-frequency use at up to

650 MHz. It has a very low input capacitance down to 1.4 pF, making it very suitable for measuring fast transients in circuits with high source impedances, or where minimum loading effect is required. Due to its active desiqn, the FET probe can also be used as a 1:1 probe with extremely low input capacitance.

Isolation amplifier

Although not a probe in the common sense, we can regard the isolation amplifier as a special kind of 'probe' that isolates the oscillo-

. scope measurement point from ground potential. This is necessary because ground on the scope input is at earth potential, unless the scope is powered by a mains isolating transformer or from batteries. The isolation amplifier also enables measurement of small signals superimposed on a large common mode voltage (See Fig. 47). The input unit is made entirely of plastic, and is battery-powered for safety. Most of the applications are in the field of power and control systems.

Fig. 46 The influence of the ground lead on pulse response

Probe with command switch

One of the latest improvements

in probes is targeted at users who do a lot of their work with a probe. In the PM 3094 and PM 3394A series of oscilloscopes, Fluke has included a new feature called a probe command switch. This uses a small switch in the body of the probe, which can activate a pre-selected function such as Autoset, or the selection of another setting from the settings memory. In the CombiScopes, it can also be used to start 'Touch, Hold and Measure': an acquisition can be performed and frozen, and multiple measurements are then made on

the stored waveform.

:chl

MTB 20.0 s -2~72dV

4.3 Other accessories

A wide variety of additional accessories is available. Some of these help to fit instruments in a measurement environment, like interfaces, rackmounts and software packages, while others make carrying an oscilloscopes and probes easier or tidier, or help in specific types of measurements.

Video line selector

This accessory is ideal for the video engineer. It produces line, field and frame sync. pulses to trigger a scope, allowing the engineer to study individual parts of specific video lines. The required line number is set, e.g. by means of switches, and when that line is reached the scope is triggered,

Some instruments like the

PM 3394A have a video line selector built in, Control is then an integral part of the instrument control.

Terminations

When working with a low-impedance system it must be terminated correctly to avoid signal distortion, Some scopes have inputs which can be switched to 50Q . For those which do not, these terminations can be fitted to the scope input. They are available with attenuation factors of 1:1, 10:1 and 100:1, A 75Q version is also available for a.o. video systems.

Interfaces

The most COmmonly found interon oscilloscopes are RS-232 GPIB (see paragraph 2.4 for more information). These allow data

.to be transferred out of and into the lloscope. Although their use for rm transfer is possible only

digital storage oscilloscopes, interfaces are also available analog scopes to enable remote of the instrument settings.

packs and inverters allow bench scopes to in the field where mains

is not available, such as in a or aircraft. They also enable

oscilloscope to make 'floating' rements with respect to

Gommon mode voltage source

+

<«>

ground, This means they can be used to avoid ground loops if a system is at earth potential.

Miscellaneous

Carrying cases protect the scope during transport, and are thus especially valuable for the frequent

traveller.

Accessory pouches can hold probes and other small accessories together with the scope. This is useful since probes should be adjusted for the scope they are used with,

A protective cover protects the scope's front panel from damage

Fig. 47 Circuit with common mode voltage

Fig. 48 Video line selector and selected video line

and dirt.

Scope carts help when a scope is frequently used at different locations around the lab, and can often also store accessories and a hard copy device.

For some scopes, an additional memory pack is available to store instrument settings, so that on power-up the same settings can be used that were active before powerdown.

4.4 Software

Various packages are available, some for simple waveform transfer and storage in a computer, others for sophisticated data analysis. Some examples, typically designed towards test instruments, are given below.

Communication Software

Packages are available that transfer waveforms from the scope to the PC and store them on the PC's harddisk. Captured waveforms can be displayed on the PC screen, together with the instrument settings and can be stored on disk for instant recall at any time.

Some packages also allow sending the waveform back to the scope, to be used for reference purposes. If the transfer time is not critical, often an RS-232 link is used because it is found on every PC.

Some packages, also enable transfer and storage of the waveform as an HPGL -tile that can directly be incorporated as a graphical image into a document, using one of the many wordprocessors available on the market. Many of the screencopies included in this book have been treated that way.

Other packages, like AnyWave, add valuable automated test functions to the oscilloscope. These programs can be used, for instance, to create a test mask (template) defined by an upper and a lower limit waveform. This test mask is then used as a reference for all captured signals allowing instant, automatic pass/fail testing on

waveforms. Waveforms can be transferred to the PC using either RS-232 or IEEE communications. Failing waveforms can be stored automatically.

For oscilloscopes equipped with template testing, AnyWave also makes creation of test envelopes easy. Envelopes can be edited using the PC's keyboard and mouse. If preferred, an acquired waveform can be used as the starting point.

Furthermore, AnyWave can be used to create waveform descriptions for arbitrary waveform generators.

Stored waveforms can be processed with a spread-sheet package or with dedicated analysis software.

Programming environments

These packages create an environment for writing test application programs using standard programming languages like QuickBasic, 'C' or other. Using this environment, application program development is much easier, more efficient and faster. ~ith packages like TestTeam and LabWindows, instrument drivers are delivered to further enhance the simplicity of programming. The packages not only support oscilloscopes, but help in writing test procedures using a variety of instruments like multimeters, signal generators, programmable power supplies, frequency counters and signal switches. The written programs can be executed from the Test-Team environment or can be converted into stand-alone executable files controlling the test system.

Newer Developments

In this chapter we will look at some new developments in the latest generation of oscilloscopes.

ScopeMeter

ScopeMeter® is a breakthrough in oscilloscope technology.

Through larue-scale integration and the use of a Liquid Crystal Display, ScopeMeter combines a fully featured 2-channel 50 MHz DSO and a Digital Multimeter in a single handheld unit weighing only 1.8 kg.

ScopeMeter is the instrument for service engineers requiring a truly portable oscilloscope. It is also a full-function oscilloscope for floating measurements, designed to be used at voltage levels up to 600 VRMS with respect to ground. It even drives a printer over an optically isolated RS-232 interface to make hard copies of measurement results.

The low power consumption allows operation on a single battery charge for a number of hours, long enough for all measurements on an average working day.

Autoset

Autoset sets the sensitivity, the timebase and triggering for optimum signal viewing. The most complete and versatile implementation scans all input channels, activates all channels carrying a signal and sets triggering and timebase in accordance with the lowest frequency found on any of the input Signals. On some scopes, a limited implementation is found of this feature, where only channel 1 is analyzed and settings are based on the signal found on this one channel only.

In the most modern oscilloscopes like those used as examples in this book, Autoset is programmable so users can customize its functionality their own applications.

Autoranging oscilloscope

The controls which are operated most frequently on any oscilloscope

are the ones that control the timebase and the attenuator. To make working with oscilloscopes even easier, one of the latest improvements in oscilloscopes has been the introduction of fully autoranging attenuators and an autoranging timebase. These new features enable an engineer to troubleshoot a system, checking signals of various amplitudes and frequencies, without the need to reach for the control panel of the scope. Handsfree operation of the scope is now just as easy as using an autoranging digital multimeter!

Timebsse in degrees

Now that electronic circuitry is heavily used in application areas other than the electronics industry itself, the need to measure other types of signals is increasing. Today's cars and their engines are equipped with numerous electronic circuits, for example fuel injectors are controlled by an engine management system, brakes are under electronic control etc.

Fluke has designed an oscilloscope that is ideal for measurements on automotive engines and systems.

Its timebase can be set in degrees rather than in seconds to facilitate analysis of engine functions. The built-in automatic variable timebase ensures a correct display of the selected part of the operating cycle, independently of the engine speed. Measurement results can be scaled for 2-stroke or 4-stroke engines, and include an RPM (rev./min.) read-out. For work with pressure transducers, the read-out can be scaled in pressure units.

Autocal and closed-case calibration

Autocal optimizes the performance of the oscilloscope by compensating for changes in environmental and instrument temperature. Calibration data is stored in non-

Fig.49 ScopeMeter

volatile memory, and this data is used together with the internal voltage and time references to ensure that the oscilloscope always operates at maximum performance.

The technology employed in the most modern instruments like

PM 3092 and PM 3394A has completely eliminated conventional trimmer potentiometers and trimmer capacitors. I nternal adjustments are made using stored calibration data, adjustable gain amplifiers, varactors, comparators and digital-to-analog converters. Stored calibration data can be changed (after passing two security levels) without opening the case of the instrument. This means calibration can be carried out at lower cost under computer control, while the instrument is calibrated at the real internal operating temperature.

Another big money saver is that this oscilloscope can be calibrated every two years, achieving the same accuracy specification as older instruments that have to be calibrated each year.

Variable timebase (eliminating 1-2-5-range fimebase settings)

Many signals have a changing frequency. If the timebase setting information is used to calculate the signal frequency, timebase settings in a fixed 1-2-5 range make these calculations easier. With automatic measurements becoming more and more standard, especially in DSOs, this need for fixed timebase settings has practically disappeared.

What engineers need in many applications is a display that shows a limited number of cycles, e.g. two or three. This can be achieved by using a variable timebase. It allows the user to set up the scope to display exactly the number of cycles required, independently of the signal frequency. If the variable tirnebase also has automatic adjustment, it can track the signal frequency to give a constant number of cycles at changing signal frequencies. Applications are in filter testing, audio testing, variable-speed drives, phase locked loop testing etc.

User Interface

The classical oscilloscope was an analog oscilloscope. Over a number of years, oscilloscopes all looked pretty much the same: from left to right there was the CRT with its controls, next to it there were two blocks of controls for the two input channels, while further to the right there was a large rotary control to set the timebase. This control layout depended greatly on the construction of the oscilloscope, since all controls acted directly on the associated circuitry.

With rising signal frequencies and the desire to control instruments remotely, the oscilloscope's internal circuitry was changed to enable electronic switching, in turn controlled by a built-in microprocessor. This microprocessor now also takes care of automatically setting up the instrument, reading the front panel controls and making the settings accordingly, positioning the cursors and calculating the CUfsor read-outs, as well as handling communication over the IEEE-488 or RS-232 bus in remote control operation.

Now that the actual setting of the instrument is done by a microprocessor, manufacturers have more freedom in the design of the front panel. As a result, control panels differ with the kind of application area the products are designed for.

Instruments that are used in automatic tests, and that are therefore mainly controlled from an instrument controller or PC, usually have a front panel with only a limited number of controls. Control over these instruments requires extensive use of software menus, selecting the assignment of the control keys to the various functions depending on previous key actions.

Instruments that are mainly used under manual control have control panels with a larger number of keys, offering better access to all elementary functions. The benefit of this kind of user interface is that only a limited number of actions are required to change basic settings like the vertical deflection or the timebase speed, because these functions are always directly accessible.

This latter control panel is especially useful for engineers working with a wide variety of signals, for example in applications in the electronics bench, in R&D, during start-up of prototypes and in service.

Chapter 5 Newer Developments

~:::':

Measurements and Exercises

6.1 How to perform measurements

In Chapters 1 and 2 we looked at the controls with which we can influence the signal display. In this section we will take a look at important waveform parameters, and you will see how to use the oscilloscope to measure them.

Oscilloscopes measure two basic quantities: voltage and time. From these two quantities all measurements are made or derived, either manually using cursors, or automatically.

When we make measurements it is very important to know our oscilloscope's capability. Don't try to look at a 10 MHz squarewave on a 20 MHz scope, because you will not be able to see the true shape of the squarewave. The 10 MHz squarewave consists of a fundamental 10 MHz sinewave and its harmonics at 30 MHz, 50 MHz,

70 MHz etc. You may also see part of the effect of the 30 MHz harmonic (although not at the correct amplitude!), but the next frequency component is 2.5 times greater than the scope's bandwidth! So what you wiJJ see will look more like a sinewave than a squarewave (see Fig. 50).

-'~.:

The same is true for risetime measurements. If you use a scope with a rlsetime 10 times faster than that of the signal, the effect of the

:'.:{,:.:-.-:'.

scope's own risetime on the

measurement will be almost negligible. However, if the scope and signal have the same risetime, then the error is a massive 41 %.

·-'f':;·

.. ·····Y.

':-:r':

.:.,::.:"-':'.:'::

:·-'··-·:':·,,:j_::'i-::·':

h1; freq- 10.DMHz

SeD e bandwidth 200 MHz :

Fig. 50 A 10 MHz squarewave displayed on 20 MHz and 200 MHz oscilloscopes

Some standard waveforms

The three most common waveforms are the sinewave, triangular waveform and the squarewave (see Fig. 51). All are found on almost any function generator, and are very common in practice.

The sinewave consists of a single frequency component; the squarewave and triangular waveform are made up of many different harmonically related sinewaves. Squarewaves are made up of odd harmonics of the fundamental sinewave, and triangular waves are made up of even-numbered harmonics. The waves are symmetrical in time and amplitude.

There are derivatives of these waves, usually symmetry changes. As a result, the triangle becomes a sawtooth (so-called for obvious reasons) and the squarewave

~' Chapter 6 Measurements and Exercises

Ratio of oscilloscope risetime to signal risetime

Error of risetime measurement in %

1:1 2:1 3:1 5:1 10:1

41

12

5

2 0.5

becomes a rectangular wave.

One complete cycle of a wave is called a period. This is the time taken to go from a specific point on one cycle to the equivalent point on the next (see Fig. 52).

The frequency is the number of these cycles the waveform makes in one second.

So if we divide the time taken for a period into 1 second, we get the frequency, which is expressed in Hertz (Hz).

e.g. Period Frequency

= 1 ms therefore 1

lx10-J

= 1000Hz

= 1 kHz

Waveforms which are repeated are called repetitive or periodic waveforms. These are the easiest waveforms to measure.

The other most commonly measured parameter on repetitive or periodic waveforms is the amplitude. This is the voltage from the highest to the lowest point on

Fig. 51 Most common signal types

sinewave

n

n

positive pulse

sawtooth

Some standard waveforms


- '--- squarewave

o

u

negative pulse

Fig. 52 Sinewave showing frequency and amplitude

CHl 1 V=

MTB

50us

chl+

the wave. This is also referred to as the peak-to-peak amplitude, or Vp-p (see Fig. 52).

6.2 BASIC EXERCISES ,. or how to interpret sinewaves

This section contains exercises to familiarize the user with the major control groups on the oscilloscope. It has two parts. The first is for controls on analog oscilloscopes, also covering controls which are also found on digital instruments. Specific exercises relating to digital

storage oscilloscopes are given later.

Required equipment

- Oscilloscope: an analog oscilloscope such as PM 3092 (analog exercises only) or a Comblscope'" from the PM 3382A ..... PM 3394A series, or similar.

- Two 10:1 PM 9010/091 or

PM 90201091 probes or equivalent with trimmer tool for probe compensation.

- Function generator capable of producing sinewaves, squarewaves and triangular waveforms up to

2 MHz or more. E.g. PM 5135 or PM 5136 or equivalent.

- Two 500. BNC cables.

triangular waveform

Gelling started

Disconnect the probes from the oscilloscope if connected.

Connect the oscilloscope to the mains power supply using the correct mains cable and switch it on.

Connect a probe to channel 1

and connect the probe to the probe adlust connector on the oscilloscope front panel, using the grabber hook.

If using a CombiScope, select analog mode.

Press AUTOSET (green button at the top of the front panel).

On older scopes without this function, set the following:

Intensity - midway

Vertical position controls - midway Horizontal position control - midway

Timebase 0.2 rns/div.

CHi sensitivity 0.2 V/div. Take care! Allow for probe attenuation if the oscilloscope does not do so automatically.

Trigger - Source CHi, mode AUTO, Peak-Peak

Ensure the probe is compensated. You might have to adjust the LF compensation trimmer. Refer to 'compensating the probe' in Chapter 4 if necessary. Now we are ready to explore the controls!

Refer to the fold-out sections at the front and back of the book for location of the controls.

Screen controls offering this calibrated variable Reconnect the probe.
Adjust the intensity and focus attenuator. Note the changed sensitivity
controls and watch the effect on the readout, now 10 V/div., and the Lz
screen. Coupling symbol has disappeared. The
Set the focus control to give a Press the AC/DC/GND button. oscilloscope has recognized the
sharp trace and clear text. Watch the coupling symbol after the probe as a 10:1 probe with a high
Note how you need a higher sensitivity readout change, as we impedance, and does not allow it to
intensity level to see the rising and cycle through the selections. Note: be used with a 50Q oscilloscope
falling edges of the squarewave on some instruments there are input impedance!
from the probe adjust output separate controls to select AC or DC
because the beam is moving faster coupling, and another to connect Horizontal controls
on these edges. the input channel or switch over to
Adjust the text intensity control ground. With these instruments, Timebase
for an acceptable text brightness operate both to see the effect. Connect the probe to CHi and
level. When changing from AC to DC the Probe Adjust signal.
Disconnect the probe from CHi coupling, the waveform appears to Press AUTOSET.
input and adjust trace rotation so jump up the screen. Press the ns and s buttons under
that the trace is parallel to the The lower part of the squarewave TIME/DIV on the control panel. You
horizontal graticule lines. is at zero volts, so it must be on the will see more periods of the Probe
Turn the graticu{e illumination centre graticule line when DC Adjust signal at slow timebase
control and watch the graticule light coupling is selected. The channel speeds and fewer at higher speeds.
up. identifier '1-' gives us the channel Notice the timebase readout
Connect the probe to CHi again. number and ground position. We changing on the screen in 1-2-5
can see that ground is indeed the steps.
Vertical controls centre graticule line when GND is Press both ns and s on the
selected. The attenuator input is TIME/DIV controls together. This
Position now connected to ground. gives a calibrated VARiable
Disconnect the probe from the time base, similar to what we saw
Probe Adjust output. Multichannel operation for amplitude.
Turn the POS (vertical position) Connect the second probe to Press ns and watch the first
control for CHi and position the CH2, connect the probe tip to the period of the signal expand to fill the
trace on the centre graticule line. probe adjust bus and press screen.
Note the channel identifier and AUTOSET. Turn the channel 2 POS
ground level indicator '1-'. control and note there are now 2 Horizontal magnification and
Reconnect the probe to the traces on the screen. position
Probe Adjust output. Position the traces with the CHi Connect the probe to CHi and
and CH2 POS controls. the Probe Adjust signal.
Sensitivity The traces can be switched on Press AUTOSET.
Press the AMPL (amplitude) and off using the ON button for Adjust the timebase to display
up/down buttons and see how the each channel. approximately 10 periods of the
displayed amplitude changes with Perform probe compensation on Probe Adjust signal on the screen.
the different sensitivity selections. CH2 if necessary. Look at the timebase readout on
Look at the bottom of the screen Add both channels together screen.
to see the sensitivity readout using the 'CHi +CH2' button. On PM 3094 press 10X MAGN.
changing. Now invert CH2 using the 'INV' On PM 3394A press MAGNIFY 7.
To select VAR (variable) button and note that the square- Note that the readout is now a
sensitivity, press the up and down wave disappears. factor of ten faster than before. On
keys simultaneously. Release both the PM 3394A, a so-called 'memory
buttons. Now use the up and down Probe recognition bar indicator' is also displayed,
buttons to smoothly change the Disconnect the probes from CHi indicating what part of the wave-
displayed amplitude. and CH2 inputs. form memory is on the scope
Note the changing sensitivity Using the AMPL or sensitivity screen. Refer to Fig. 53 and Fig. 9.
readout. control, select 1 V/div. for CHi. We can now scan through the
Press both buttons simuI- Now set CHi input impedance to magnified trace using the horizontal
taneously to reselect the 1-2-5 50n if available. The oscilloscope X-POS control.
steps. may have a special key to activate
this low-impedance input or it may Dual timebese
Note: Fluke analog oscilloscopes be in a 'vertical menu'. Note the Lz Connect the probe to CHi and
like the PM 3094 and CombiScopes symbol after the CHi sensitivity read- the Probe Adjust signal. Press
like the PM 3394A are unique in out showing low input impedance. AUTOSET. Chapter 6 Measurements and Exercises .

I I 1

non-magnified
.-------

L--Jv

C~ V
L
LJ U
/
/
(
J/ /
1\ /
r~ V

t;:H;L O. v~ maT:
1/ -CHI 0 .. 2 v~ MTB 100 s Fig. 53 Normal and magnified display

Refer to the front panel fold-outs at the front and rear of this book as necessary.

Press DTB in the delayed timebase control section.

On PM 3394A select DTB ON using the top softkey. On PM 3092 step through the DTB menu using the second softkey to select MTBI+DTB

Use the Position and Trace Separation controls to place the main timebase in the upper part of the screen and the delayed inthe lower.

Use the DELAY and DTB timebase controls to select and magnify one of the rising edges of the Probe Adjust signal,and adjust the trace intensity as necessary.

Note the on-screen readout of delay time and delayed timebase speed. The delay time increases as the bright section moves to the right, away from the MTB trigger.

input and see the trace disappear: no signal, no trace!

Select AUTO again and the trace reappears.

Connect the probe again to see the signal.

Now select SINGLE from the same menu. The trace again disappears. Turn the trace intensity control clockwise.

On PM 3092 press SINGLE RESET. On PM 3394A press SINGLE in the horizontal control section.

Watch the screen carefully. As you release the button one trace will flash across the screen. Each time you press the button you will get one sweep.

We will do another exercise on this when we have learned some more about triggering.

Trigger controls

Set the function generator as follows: 1 kHz sinewave, sweep off, DC offset off, output voltage 1 Vp-p.

Connect the function generator output to CH1 using a BNC cable.

Press AUTOSET.

Use X-POS to move the trace to the right so you can see the start of the sweep.

Peak-to-peak triggering

Adjust the trigger level control. Notice how the starting point moves up and down the slope.

Change the output amplitude of the generator and adjust the trigger level again. Note that the oscilloscope always triggers on the waveform .

The trigger level is displayed as a percentage of the p-p signal amplitude while adjustments are in progress.

Automalic, triggered and singleshot timebase

Connect the probe to CH1 and the Probe Adjust signal. Press AUTOSET.

Press HOR MODE on PM 3092 or TB MODE on PM 3394A.

Select TRIG (triggered) from the menu using the softkeys. The oscilloscope now needs a trigger signal to start the timebase sweep.

Remove the probe from CH1

Trigger slope

Press the TRIG 1 button found in the CH1 control section. The oscilloscope now triggers on the negative-going slope.

Note how the slope symbol in the bottom right corner of the screen changes as you press the TRIG 1 button.

Triggering from another channel Connect CH2 input to the TTL output on the rear panel of the function generator.

Press Autoset and position the traces on screen so they do not overlap.

Watch the trigger source indication at the bottom right of the screen and press the TRIG 2 button. The scope now triggers from Channel 2. Remove the BNC cable from CH2 input to prove it!

Triggering at a specific voltage level

Set the function generator as follows: 1 kHz, triangle wave, sweep off, DC offset off, amplitude 1 Vp-p.

Connect the function generator

output to CH1 using a BNC cable.

Press AUTOSET.

Set CH1 to DC Coupling.

Press Trigger MTB on PM 3092 or Trigger on PM 3394A, and select Level-pp OFF

Using the bottom softkey of the trigger softkey menu, set trigger coupling to DC.

You will see the symbol 'T -' or 'M-' appear on the left of the screen. This indicates the trigger level on the signal.

Now adjust the Trigger Level control. The T- will move up and down the signal while the on-screen readout gives the actual voltage the scope is triggering on.

Also note that you can now set a trigger level outside the p-p signal amplitude, thus losing triggering.

Chapter 6 Measurements and Exercises~

Single-shot triggering

Now that we know how to set a precise level to trigger on, we can do another single-shot exercise.

Connect the probe to CH1 and the Probe Adjust signal.

If you are working with a CombiScope, make sure it is in the analog mode.

Press AUTOSET.

Disconnect the probe from the Probe Adjust output.

Turn the screen intensity control fully clockwise for maximum trace brightness.

Press HOR MODE on PM 3092, TB MODE on PM 3394A and select SINGLE from the menu using the softkeys.

Set CH1 to DC Coupling.

Press Trigger MTB on PM 3092 or Trigger on PM 3394A and select Level-pp OFF.

Using the bottom softkey select trigger coupling to DC.

You will see the trigger level symbol 'M-' or 'T-' appear on the left of the screen.

Now adjust the Trigger Level control. The trigger level symbol will move up and down the screen while the on-screen readout shows the voltage the scope will trigger on. Set this to 200 mV.

On PM 3092 press SINGLE RESET, on PM 3394A press SINGLE in the horizontal control section.

The red LED beside the Trigger Level control llqhts to show that the oscilloscope is waiting for a signal on which to trigger.

Watch the screen closely and touch the probe tip to the Probe Adjust output.

The trace will flash across the screen once. Touching Probe Adjust again with the probe will have no effect until the trigger circuit is rearmed.

If you have a PM 3394A, repeat the above procedure in the DSO mode. You may use about one division pre-trigger view to have the rising edge on-screen.

- Triggered dual timebase

Set the function generator to sinewave, start frequency 120 kHz, stop frequency 121 kHz, sweep time 50 ms, amplitude 1 Vp-p, with a

continuous sweep.

Connect the generator to CH1 input using a BNC cable.

On the scope set the following:

CH1 0.2 V/div., AC coupling, Main Timebase 20 ~s/div., Trigger CH1, positive slope, level pop ON, level 50%.

Switch on the Delayed Timebase, select STARTS after delay mode.

Set DTB sweep speed to 1 us/div. and delay to 120 us.

Use the trace separation and position controls to get a picture like Fig. 54a.

The waveform cycles on the right

of the screen appear to be moving in the X-axis, It is very difficult to examine one of these cycles in detail.

Now select DTB to Triggered 'TRIG'D' mode from Channel 1. On the PM 3092 this may be indicated as 'Ch1',

Select the DTB triggering to DC coupled, and set the DTB trigger level to 500 mV using the LEVEL DTB control on PM 3092 or the DTB menu on PM 3394A. Notice the 'D-' symbol on the left of the screen. This is the DTB trigger level indicator.

...

f\

.. ~ 1\ !\ I~ .. "','" .~ 1/ ... 1/

,,'[\ l\ t\ • ~ II !I

.......... ~ ....

. . .

. ~ .

1'1 '1\

'\

.... ~,. \

,~I\

,

: MTB 20.0'us 'chl+

-C(-I;1 o. Z v~ A~ T p;rB 11, .,QO I,!~ 1;'::0. o_:_~s

Fig.54a MTB and OrB. DTB in STARTS mode

\

Fig. 54b and DTB. DTB in triggered mode

IJ I

MTB trigger lev indicator

v

OlE gro""~ , ' indicator

CHI, 0.2 V-

... \ .." ...

• II

~ ,.

J : .

, Identifies OTB ,

A L T ~~: ~o ~~~:i:::gLS:~

DTB trigg~~el

indicator ~

r-, ..,_,.

.. P.>:-: <: ... ",-

Channel 1 gro;(und \J' ' ~

level indicator 11,/ I ...

1\' 1 if

F::

~ Chapter 6 Measurements and Exercises ~ ~

Fig. 55 Calibrator output

(200 mV/div., 200 !1sldiv.)

Fig. 56 The oscilloscope is used to set a generator to a specific amplitude and frequency

Press the CH1 trigger button to select DTB triggering from Channel 1 on a positive slope.

You should now have a picture like Fig. 54b.

Note the change in the delay readout. It now shows the '>' or 'greater than' symbol in front of the delay time. This means the trigger circuit for the DTB is waiting for the first occurrence of 500 mV, on the first positive-going edge, greater than 1 00 us (in the figure) after the MTB is triggered.

Now change the delay time using the DELAY control. Note how the intensified area jumps along the

MTB sweep between qualifying trigger points, while the DTB remains stable. Change the DTB trigger level and watch the start of the intensified area move up and down the sinewave as the trigger level changes.

Change the generator output Signal to a squarewave. Select STARTS mode again on the DTB. Increase the MTB sweep speed to 2 us/div. Increase the DTB sweep speed to 50 ns/div. Set the delay to

8 us so that the intensified area is on a rising edge in the centre of the screen. Try to examine the edge using the DTB. A little difficult?

Select DTB to Triggered 'TR IG'D' mode. The rising edge is now stably displayed by the DTB. Changing the DTB trigger level enables you to select what part of this rising edge is to be displayed.

6.3 Measurement exercises

Now that we have learned to use the controls of the oscilloscope, we are ready to do some actual measurement exercises.

Measuring the frequency and vollage of the probe calibrator output

Connect the probe to CH1 and to the Probe Adjust output, and press AUTOSET to get an optimized display. On instruments not having AUTOSET, adjust the sensitivity, timebase, horizontal and vertical position controls to get a display similar to Fig. 55.

Count the duration of a cycle and the amplitude in divisions of this signal.

Multiply these by the volts/div. and time/div. settings. This gives the amplitude and period of the waveform.

The frequency is a derived measurement, and is obtained by taking the reciprocal of the period.

The measurement results you get may differ from our picture because of tolerances in the specifation.

Using the oscilloscope to set a function generator 10 a specific amplitude and frequency

Using the oscilloscope we will set the function generator to give the sinewave in Fig. 56, frequency 100 kHz, amplitude 3 Volts peak-to-peak.

Set the trigger to 'level p-p', AC coupling, source CH1 and the timebase to AUTO.

Set the sensitivity to 500 mV/div.

A deflection of 6 vertical divisions now equals 3 V.

A 100 kHz wave has a period of 10 microseconds (us): 1/(1 00x1 03) = 1 Ox1 0-6. So if we set the timebase to 2 us/dlv., one period of the wave will occupy 5 divisions.

Adjust the generator output level and frequency controls as well as

lc H 1 F,00mV-

• MTB 200~!?

, chl-;t-

.:'~

HI O. 5, v~

MTB 2.008s

chl+

the oscilloscope trigger level and slope controls to get the picture shown in Fig. 56.

In the real world, signals are never quite as well shaped as those we can produce with the function generator. Let's look at a typical pulse waveform.

Fig. 57 shows the time and amplitude characteristics of pulse waveforms.

The 'duty factor' or 'duty cycle' is important in pulse waveforms as it is used to calculate the average power in pulsed systems such as radar, motor speed control, Jlghting control and many more.

Duty Factor = T wi Trep

where Tw is the time the pulse is active

Trep is the repetition time or period of the pulse.

The duty cycle can be expressed as a percentage or as a number between 0 and 1.

AVerage power = peak power x duty factor

Jitter is a measure of the stability of the pulse Signal. It is expressed in time or as a percentage of the repetition time.

The amplitude characteristics are mainly distortions of the ideal rectangular pulse. Special attention should be paid to the levels where parameters begin and end (0%, 10%,50%,90%,100%). Now you can see why the extra lines marked 0% and 100% are on the scope's uratlcule. When we use variable sensitivity to place the signal on these lines, we can make measurements at the other levels. All the amplitude measurements shown in Fig. 57 are referenced to the 100% level.

To measure the risetime of a squarewave using the Delayed Timebase, first method

Set up the generator to give a squarewave of 3 Vp-p. about 120 kHz, sweep off, and apply the Signal to the oscilloscope's Channel 1. Press AUTOSET, change the MTB sweep speed to display more than a

10%

Fig. 57 Pulse parameters

-_ 50%

J---

I ~I -

-1

- .--..j

0%

Fig. 58 Risetime measurement using the DTB

100%

-. 90%--

CHI; 63ImV~

: mtb I. OOtlS

• DTB 20.0hs 8,406us

'chl+

full cycle on-screen.

Switch on the delayed timebase and position the intensified area on a rising edge. Choose an edge on the left side of the screen to minimize the effects of signal jitter.

Set the delayed timebase sweep speed so that the rising edge is displayed over several horizontal divisions. Adjust the DTB delay and trace intensity if necessary.

Switch off the main timebase. Switch CHi coupling to Ground and position the trace on the centre line of the display.

Switch CH1 to AC coupling. Using the variable sensitivity, place the top and bottom of the Signal on the 100% and 0% graticule lines respectively. You may

have to make minor adjustments to the vertical position control.

Use the X-POS control to intersect the rising edge with a major vertical graticule line and the 10% horizontal line.

Now measure the time to the trace intersection with the 90% horizontal graticule line. See Fig. 58.

The time the signal takes to switch from the 10% amplitude point to 90% of its final amplitude is called the risetime.

To measure the riselime of a squarewave using the Delayed Timebase, second me/hod

Use the set-up of the previous exercise.

Using the variable sensitivity,

· - - ---- --------

- --

, '"' ~ ~ ,

Fig. 59 Pulse width measurement

Pulse Width =

4.8 x 50 us = 240 us

-

120mV~

: mtb 200 us

: OTB 50. 08s - O. 02dv:

CHI

place the top and bottom of the signal on the 100% and 0% graticule lines respectively. You may have to make minor adjustments to the vertical position control.

Use the Delay control or the Delay Time Multiplier to intersect the trace with the graticule where the centre vertical and 10% horizontal lines meet.

Make a note of the delay time. Use the Delay control or the Delay Time Multiplier to intersect the trace with the graticule where the centre vertical and 90% horizontal lines meet.

Again note the delay time. Subtract the first reading from the second to get the time difference between the 10% and 90% levels. This again is the risetlrne.

Greater resolution can be obtained this way when the Delayed Timebase sweep speed is increased.

This method can also be used to accurately measure pulsewidths and pulse repetition rates.

Measure the pu!sewidth of the Probe Adjust output using methods 1 and 2 as above

Connect the probe to CH1 and the Probe Adjust output.

Press AUTO SET to get an optimized display.

Make sure the probe is compensated.

Switch on the delayed timebase

:chl+

and position the intensified area on a positive half-cycle. Choose a cycle on the left side of the screen to minimize the effects of signal jitter.

Set the Delayed Timebase sweep speed so that the selected half-cycle is displayed over several horizontal divtsions.

Switch off the Main Timebase. Using the variable sensitivity, place the top and bottom of the signal on the 100% and 0% graticule lines, respectively. You may have to make minor adjustments to the vertical position control.

First method: use X-POS to intersect the rising edge of the trace with a vertical graticule line and the central horizontal graticule line. The horizontal graticule line should be used since this is at 50% of the amplitude of the signal (see Fig. 59).

Now measure the time to the intersection of the falling edge and the central graticule line. This gives us the pulsewidth.

Second method: Use the delay control to position the rising edge of the trace on the intersection of the central hortzontal graticule line and the vertical graticule line. It doesn't matter if the falling edge goes off the screen. To get better measurement resolution you can increase the Delayed Timebase sweep speed.

Adjust the trace intensity if necessary.

Make a note of the delay time. Adjust the delay control again so that the next falling edge passes through the same point on the graticule. Again note the delay time.

Subtract the first reading from the second to get the time difference betweeri the 50% points on the rising and falling edges. This is the positive pulsewidth.

Now using the delay control select the next rising edge and note the delay time.

Subtract the first reading from the second reading to calculate the width of the negative half-cycle.

Can you now calculate the frequency and duty cycle?

6.4 Cursors and automatic measurements

Cursor types

In all the measurements made so far, we have used the graticule and the setting information of the attenuator and timebase. More modern analog instruments have cursors to make measurements easier and quicker.

Cursors are lines drawn on the screen by the electron beam. Cursors can be both vertical and horizontal. Their positions on the screen are relative to voltage and time. These positions are used as a basis for the oscilloscope to measure voltage and time, and to derive other measurements such as frequency, risetime etc.

The cursor timing and voltage read-outs are automatically adjusted when the sensitivity or timebase settings are changed.

The readout can be in absolute volts, i.e. volts referenced to ground, relative volts, i.e. the voltage difference between the cursors, or percentage. Percentage is particularly useful for measurements on pulses because, as we have already seen, measurements such as duty cycle are expressed as a percentage of the period.

There are two types of cursor system. The first is used on analog

instruments and some digital instruments, and is called screen mapped cursors? There is no link between the cursors and the input signal, which means the operator has to align the cursors with the waveform manually to make measurements (see Fig. 60). This manual alignment leaves room for errors since the operator has to make the alignment based on the visualized waveform and cursor . The waveform and cursor may be affected differently by any small display inaccuracies, resulting in measurement errors.

The second type of cursors are based on digitized waveform data stored in the oscilloscope, These are called memory based cursors, They do not have the errors that can be introduced by the deflection system, The cursors track the waveform on the screen, Since all the waveform data is held in the oscilloscope, other measurements such as risetime, frequency and period can be calculated on designated parts of the waveform. This type of cursor is found on, amongst others, the PM 3365A

(see Fig, 61).

On some instruments like

PM 3394A, the cursors can be assigned to different traces, allowing measurements like propagation delay, switching time etc.

Amplitude qualified cursors

A third and not so widely found type of cursor is the amplitude qualified cursor, These cursors are especially useful in defining application-specific time measurements which differ from the 'standard' measurements like risetime. Examples are found in component testing (e.q, diode reverse recovery time tests), control loop settling time tests, PLL lock-in time testing and many others.

The name comes from the fact that the time measurements are

2) Note that for certain measurements in more advanced instruments like

PM 3094, the screen mapped cursors are linked to the signal. These instruments use the peak to peak trigger circuit to measure input voltage and can derive other measurements from this.

chI: dT~g~,On dV~3.52 V

.,

" ,

"'0 ,.

:'

--\----------

.. ..'

:chI+

CHI:

,

'~ time cursors ::

... :., -: : ~

•••••• 1

,

,

MTB 20.Qns

A:

t~2.1EJ-IS

: t r=: : 465ris

. . . . ~ ....

(- CUR 1 -)

.. ; .

. '. . . .. . .... ~ .

«: CUR.\' ->

RETURN

made by placing the cursors at defined amplitude crossings in the signal. For example, one cursor can be positioned at the point where the signal reaches 20% of its final amplitude, regardless of the actual amplitude; the other cursor at the point where the signal reaches 80% of its amplitude, The cursor readout gives the time between the two cursors, which is now the time needed for the signal to switch from 20% to 80% of its amplitude.

Using amplitude qualified cursors, time measurements are independent of the actual signal amplitude, Measurement flexibillty

is very great; the cursors can be positioned at any level with respect to defined references, reference levels can be chosen from a list of actual amplitude related values (e.q. minimum value, maximum value, an absolute level, ground or a statistical high or low level). The cursors need not be placed at the first time the specified level is crossed; any other crossing such as the second, third or last crossing can also be used.

To show the power of this system, let's look at the following example (see Fig, 62 from the PM 3394A).

Fig. 60 Screen mapped cursors

Fig, 61 Memory based cursors of PM 3365A

: Chapter 6 Measurements and Exercises '

~,

c h I : dT-L 49us V~+2. 00/+5.00%

. . . . . .. . 5 % of step amplitude

: :J

5 % settling time l

Fig. 62 Amplitude qualified cursors used to measure settling time.

We see here the output voltage of a control loop, where the input signal made a sudden change (step voltage). For this system, the settling time was specified as the time the loop takes to reach and stay within 5% of the correct output voltage again.

With most oscilloscopes, the time measurement must be done by an operator, using the cursors under manual control; with amplitude qualified cursors it can be done automatically.

The measurement starts at the moment the input signal produces the step. If the input signal cannot be accessed, the beginning of the increase of the output signal can be used as well.

The scope is set up to find the moment at which the waveform is close to the beginning of the rising edge, let's say at a level of 2% of its amplitude. For practical measurements, this is good enough to define the beginning of the step of the input voltage. For the first cursor, the initial stable level is referred to as 0%, the final value is taken as 100%. The first cursor is placed at 2%; in between these 0% and 100% reference levels.

The second cursor may have different reference levels; it goes to the point at which the signal crosses the 5% level of the waveform, having defined 0% as the final

+

value and 100% as the initial value of the step voltage! The final value is the value reached after all ringing is faded out. This is selected as the 'statistical high' level.

To find the moment in time after which the waveform stays within 5% of its final voltage, we select the cursor to be placed at the point at which the signal crossed the 5% amplitude for the last time.

The resulting display, including cursors and reference lines, is shown in Fig. 62. The settling time can be read from the top text line as 1.49 us,

This measurement automatically gives the loop's settling time, regardless of the amplitude of the input step. This is very useful for tests which are performed repeatedly, e.g. in production testing, and allows the instrument to perform these tests continuously without an operator.

Cursor measurements

Measuring the period and the frequency using cursors Connect a probe to CH1 and to the Probe Adjust output. Press AUTOSET to get a display.

Activate the cursors, select time mode or vertical cursors in order to make time measurements.

Position one cursor over the

beginning of a signal cycle, and the other over the end of the same cycle. The cursors should now identify the period of the signal. The read-out can be selected as L'I. T (the time between the cursors), which gives the period of the signal, or as 1/L'1. T which gives the frequency.

Meas~ring the duty cycle using the cursors

Using the same set-up as the previous exercise, it is easy to find the duty cycle of the calibrator signal.

Change the read-out to give L'I. T as a ratio. Make sure the cursors enclose exactly one full cycle. Now we have to tell the scope that this time span equals 100% of a cycle by pressing 'L'I. T =100%'.

Leave the first cursor at the beginning of the cycle, reposition the second cursor to the edge in the middle of the cycle (see Fig. 63). The cursor read-out now indicates the duty cycle of the enclosed part of the pulse, which is approximately 50%.

, Phase measurement using the cursors

Connect the function generator output to CH1, the TIL output (on the rear of the generator) to CH2. Set the frequency of the generator to the highest possible value, using a triangular output signal.

Activate the scope's AUTOSET function. Make sure CH1 is used as the trigger source. Adjust the vertical deflection and vertical position of the traces to see the two traces clearly separated. Change the tirnebase to show slightly more than one cycle on-screen, using the variable mode if necessary. Adjust the trigger level to show sufficient information on the rising edge of the CH1 waveform (see Fig. 64).

Position each waveform symmetrically around a horizontal graticule line.

Activate the cursors and select vertical cursors for time measurement, select phase read-out.

Now position the cursors to identify one cycle of the CHi waveform, starting from the first rising edge. Use the crossing with the

horizontal graticule line as a reference to find the middle of the edges.

Press the '& T =360° key to make the scope recognize this time span as a full cycle. On the PM 3394A, the cycle is recognized automatically from the frequency of the trigger source.

Leave the first cursor position unchanged, while positioning the second cursor at the middle of the rising edge of the CH2 waveform. With the PM 3394A, the small cross which identifies the crossing of the waveform and the vertical cursor line can be placed on either waveform. Selection is by the 'select cursor trace' in the cursor menu.

Read out the phase difference between the two waveforms in degrees.

When using the PM 3394A, note that this instrument automatically recognizes the frequency of the signal on the selected trigger source. From this frequency, the 360 degrees reference for the phase measurement is automatically derived. If the signal frequency changes, the PM 3394A therefore adjusts the cursor reference automatically. With other instruments, including the PM 3092, the new cycle has to be selected under manual control.

Measuring the amplitude of a signal by means of the cursors Apply an input signal to the scope and press AUTOSET.

Switch on the cursors, and select horizontal or amplitude measurement cursors.

Position one cursor to intersect with the lowest signal values, and the other to identify the highest signal values.

The cursor read-out now gives the difference between the two voltage levels, which is the peak-topeak amplitude of the waveform. On some scopes, like the

PM 3092 we are using as an example, the read-out can be changed to show the absolute level of each cursor individually. This is especially useful when a signal is superimposed on a DC component, . and for measurements on logiC

. signals.

T -':.l-

,

CHI 200mV~:

•• J_ •• : ••

,

dV+ 497fT V

r : .:

... I;_c t •

r ' ,.

,

,.

,:

•• ~. • • •• j ~ ••••.

, r :

r-' ~......y'.

chl+

r •

::

•.• J .

r : , ,: s : ,.

,:

, I ~ .

,. ,. ,. ,

:: MTB 200us

, ,

. ,

, ,

,

c~i: ph+: 90_:?

, . , :

., .,

, . ,

::

. ~: .. + ii ..

c . '.

,

: ,

.' ,

. ,

, . , . ,

2~"~":" .

, : , .

,

....... -r-c

, : , :

, ,

c. • :. '..... : ~

CH~3.00 :ll~

CH8: 2;~~

,

MTB lOOns

.7mV:

,. .. ~ ..

chl+

Measurement of the modulation depth of an AM signal 3

The previous amplitude measurement may seem rather simple, and possibly does not show much of the measurement power of cursors. So let's look at a more complex measurement in which we'll use the cursors to find the ratio of two amplitudes.

With amplitude modulated signals, the modulation depth is defined as the ratio of the amplitude of the modulation to the amplitude of the carrier. In the literature, it is often referred to as 'rn'.

To measure the modulation

depth, set up the generator to give a signal with amplitude modulation. If possible, set the modulation depth to a known level. Apply the modulated signal to CHi, apply the modulating signal to CH2 (sometimes found at the rear of the generator).

Press AUTOSET. With Fluke oscilloscopes, AUTOSET automatically selects the lowest frequency signal as the trigger source. Switch off CH2, because we want to look at

3) This measurement requires a signal source with amplitude modulation capability. e.g. PM 5134, PM 5138A or PM 5139.

Fig. 63 Use of cursors to measure the duty cycle

Fig. 64 Phase measurement using the cursors

the modulated signal only. With PM 3394A, select the analog mode or, when in digital mode, select 'peak detect' mode.

Change the timebase and vertical sensitivity to get a display similar to that of Fig. 65.

Switch on the cursors, select horizontal cursors and switch over the read-out to the t1V ratio.

Regardless of the amount of modulation, the amplitude of the carrier can be measured by placing the cursors over the waveform at the levels a and c (see Fig. 65).

Press the t1V=100% key to identify this as the amplitude of the carrier, which is regarded as 100%.

Next, shift the upper cursor down until it is at level b(see Fig. 65).

The cursor read-out now gives the amplitude modulation depth in percent.

How does it compare to the level set on the generator?

Fig. 65 AM signal

6.5 Digital Storage Oscilloscope exercises

Most of the measurement exercises we have done so far can all be made with an analog oscilloscope as well as with a digital storage oscilloscope. In this section we'll practice some of the specific DSO capabilities, like single-shot capture and automatic measurements.

A digital storage oscilloscope alone has the capability to store a waveform. This enables us to capture signals in situations in which we might not even think any Signa! was available.

Capturing a Single-shot signal

Connect a BNC cable to the CH1 input of the PM 3394A, and leave the other end of the cable unconnected. Make sure the oscilloscope is in the digital storage mode.

Press AUTOSET and note that the instrument does not find any input signal. It therefore selects the default settings 50 mV/div.,

1 ms/div., Channel 1 on, triggering from the positive edges of

Channel 1 , using AUTO triggering in the absence of a signal.

Select 2 divisions of pre-trigger view. Switch over to 10 mV/div. Select DC triggering from the trigger menu, switch off the level peak-peak triggering. Switch to the triggered mode from the TB mode menu, thus disabling the autotriggering. Set the trigger level indicator'T - ' about half a division above the baseline.

Take the open end of the BNC cable and bounce it on the table.

Note that this causes a small electrical signal to be generated by the cable, and that this signal is acquired by the oscilloscope (see Fig. 66).

CBhapter 6 Measurements and Exercise'~

Storage of a waveform

Now that we have captured a true single-shot event, let's store tt in a back-up memory for later use.

Select the SAVE menu. This allows the acquisition to be saved in any of the memories, to be selected by the TRACK rotary control.

Save the waveform in memory 3 (m3). The circle in front of m3 will now be filled, telling you that memory 3 now contains waveform information. If the circle was already filled, the oscilloscope will first ask you if it is OK to overwrite this memory.

Automatic measurements

Most modern digital storage oscilloscopes like the Fluke

PM 3394A can make measurements completely automatically without the use of cursors. The list of available measurements is very comprehensive, and covers all the usually required day-to-day measurements. The instrument uses the stored waveform data and follows standard algorithms defined by the IEEE4 to arrive at the selected measurement result.

As an example, we'll measure the amplitude of the acquired singleshot signal.

Open the MEASURE menu.

Select 'MEASurement 1 '. By default, it is set up to make a peak-to-peak amplitude measurement, which suits our application nicely. Switch on this measurement, and the result is shown on-screen immediately.

Note that a wide range of such automatic measurements can be performed, combined in the menu as amplitude-related (,volt'), timerelated (including period and frequency) and delay measurements.

Cursor limited measurements

Consider the signal of Fig. 67. If the ringing frequency is to be found, the scope will respond by giving the frequency of the squarewave.

To measure the frequency of the ringing automatically, the set-up of

4) IEEE = Institute of Efectricaf and Electronics Engineers, Inc.

scope has to be changed in

· ch a way that only this ringing is uired. Alternatively, the DTB can

used to acquire only this specific of the signal, but by doing so, overview of the total signal is

Some oscilloscopes have the ity to perform automatic

rements within a specified

· rtion of the waveform, which is d by means of the cursors. · is way, the overview of the total I is maintained while the urement can be controlled so

it is performed on a defined I of the signal, like the

or the amplitude of the

, or the frequency within a rst of pulses.

' •. Recall the waveform you stored mory 3 to get it back on the Iloscope screen by selecting the menu. Scroll through the

u listing until m3 is highlighted. on the display of this

rm. Display of the input

can now be disabled by

ng the Channel 1 key labelled Since there is now another

rm on screen, the acquisition ry wi!! be disabled.

the measurement menu, set up measurement 1 to calcu-

· the peak-to-peak amplitude. the 'Cursor limit' sub-menu switch on the cursor limiting. cursors will automatically

Place the cursors over a detail of signal and notice the amplitude ication.

· Shift the two cursors to isolate er part of the signal, and see

the peak-peak reading differs different parts of the waveform.

.•. Make sure the instrument is in

· ital storage mode.

Press the purple AUTORANGE

in the timebase menu as well as

T-'-

. . . . . . . . ; . .

:ch1+

ch1~ c~1.: .. _fri:!q~ 58. 6MHz

1-

CH 1 • 25. Q mV~: : MT

2 0 0 r) s - 1 :. 1 8 d v:

'chl

for the input channel carrying the signal.

Now change the amplitude and the frequency settings of the signal generator and note that the scope continuously adjusts automatically to changing signals.

As you can see, the autoranging CombiScope follows any changes in the signal. This makes the instrument the perfect tool for applications in which a number of different signals have to be

scanned, for example in repair work or when starting up a prototype.

If you have any electronic system at hand, try for yourself by probing

test points in that system. Note how signals of different amplitudes and different frequencies are all recognized by the instrument, and all lead to a stable display of the correct amplitude at the correct sweep speed.

As a comparison, check the same test paints again without Autoranging. Notice how much ease of use the hands-free Autoranging brings you!

Fig.66 The electrical signal generated by bouncing an open cable on the table

Fig. 67 Squarewave with ringing

Index
AC coupling 7 Jitter 49 Storage Tube 15
Acquisition length 17 Line triggering 11 Subtracts 7
Adjustments 31 Linear interpolation 26 Sweep generator 9
Alias 19 Liquid Crystal Display 22 Sweep speed 9
Alternate mode 8 Lissajous Figu res 13 Text intensity 45
Amplitude 6,43 Load 35 Time display 31
Amplitude qualified cursor 51 Loading effect 35 Timebase 6,9
Analog-to-Digital Convertor 15,16 Low frequency compensation Timebase magnification 10
Arbitrary waveform generators 29 capacitor 36 Trace rotation 45
Armed 21 Low-repitition rate 31 Transmission test signals 31
Atten uator 6 Magnetic deflection 21 Trigger circuit 6
Automatic 11 Main tirnebase 10 Trigger hold-off 12
Autoranging attenuators 41 Mapped cursors 51 Trigger level 11
Autoranging time base 41 Mathematical processing 27 Trigger position 22
Autoset 19 Maximum real-time sampling Triggered 11
Back-up memory 26 rate 17 Uncalibrated 6
Bandwidth 8,35 Maximum single-shot timebase Update rate 31
Base line 7 setting 17 Useful risetime 17,27
Bits 17 Memory location 26 Variable 6
Charge Coupled Device 21 Modulation 31 Variable timebase 10
Chopped mode 8 Modulation depth 53 Vector scan display 21
Chopping 8 Normal 11 Vertical (V) deflection plates 5
Color burst 31 Pattern triggering 24 Vertical position 7
Composite triggering 11 Peak detection 19,23 Vertical resolution 16
Coupling 6 Periodic waveforms 43 Video signal 31
DC coupling 7 Persistence 5,31 Volts per division 6
Deflection amplifier 6 Phosphors 5 Windows mode 26
Delay 10 Post-trigger delay 22 x-axls position 9
Delayed timebase 10 Post-trigger view 22 X-V Deflection 13
Digitization 15 Pre-amplifiers 6 X-V mode 13
Display samples 26 Pre-trigger 14
Divisions 5 Pre-trigger view capability 22
Dots mode 26 Probe 35
DTB starts 13 Probe adjust 45
Electrostatic deflection 5,21 Random sampling 20
Equivalent sampling rate 20 Raster scan display 22
Equivalent time sampling 20 Real-time acquisition 19
External trigger input 11 Real-time sampling 19
Flash converter 16 Record 17
Focus 45 Record length 17
Glitch 22,32 Register location 26
Glitch capture 23 Repetitive waveforms 43
Graticule 5 Roll mode 23
Graticule illumination 45 Sample clock 15
Ground 7,45 Sample-and-Hold 16
Horizontal (X) deflection plates 5 Sampled 15
Horizontal deflection amplifier 6 Sampling Rate 15
Horizontal position 9 Sensitivity 6
Horizontal resolution 17 Sequential sampling 20
I nfinite resolution 31 Signal delay line 14
Input attenuator 6 Sine interpolation 18,26
Intensified 10 Single 11
Intensity 45 Single shot 11,21
Interference 35 Slope 11
Interpolated 26 Spike 22,32 , , , Index

ational Sales Organizations
Dhabi Bahrain Canada China,People's C.S.F.R.
AI Afaq Mohammed Fluke Electronics Republic of ELSa
bu Dhabi Branch Fakhroo & Bros. Canada Inc. Fluke International Corp. Na Berance 2
O. Box 26386 P.O. Box 439 400 Brittania Road Fluke Representative 16200 PRAHA 6
U DHABI Tel. 973253529 East Unit #1 Office Tel. 42 2 3164810
I. 971 2 771370 Fax 973 275996 Mississauga, Ontario P.O. Box 9085 Fax 42 2 364986
971 2315542 TeL 905 890-7600 Beijing 100004
Bangladesh Fax 905 890-6866 Tel. (86) (1) 512 3435 Data Elektronik Brno
Motherland Corporation Tlx (716) 222529 Jugoslavska 113
24 Hatkhola Road Fluke Electronics FLUKE CN 61300 BRNO
(1 st floor) Canada, Inc. Fax (86) (1) 512 3437 Tel. 42 5 574002
(Tikatuli) Dhaka - 1203 1690 Woodward Drive Fax 425574002
700 Drancy Tel. (880) (2) 233240 Suite 216 Fluke International Corp.
I. 1.48 96 63 63 Tlx (950) 642022 PCO BJ Ottawa, Ontario Shanghai Cyprus
1.48966360 Fax (880) (2)833983 K2C 3R8 Representative Office D. Ouzounian, M.
Tel. (613) 723 9453 No. 301, Hua Shan Rd. Soultanian & Co.
Belgium Fax (613) 723 9458 Room 405 P.O. Box 1775
Antillana N. V N. V Fluke Belgium S.A. Shanghai 200040 NICOSIA
inda Michigan 4 Langeveldpark - Unit 5 & 7 Fluke Electronics Tel. (86) (21) 248 8999 Tel. 357 2 442220
O. Box 3523/3051 P. Basteleusstraat 2-4-6 Canada, Inc. Fax (86) (21) 248 3789 Fax 357 2 459885
Ilemstad, Curacao 1600 Sint-Pieters- 1255 Trans
. (599) (9) 615277 Leeuw Canada Highway C.I.S. Denmark
(599) (9) 612772 Tel. 02.33 12777 Suite 120 Infomedia Fluke Danmark AlS
Fax 02.33 11 489 Dorval, Quebec UI. Petrovsko- Ejby Industrivej 40
H9P 2V4 Razumovsky proezd. 29 2600 Glostrup
Bolivia Tel. (514) 685 0022 103287 MOSCOW Tel. 43.44.1900
Coasin Bolivia S.R.L. Fax (514) 685 0039 Tel. 7 095 2123833 Fax. 43.43.9192
uenos Aires Casilla 7295 Tlx 411670 dnepr su
P 1407 La Paz, Bolivia Fluke Electronics Fax 7 095 2123838 Dubai
I. (54) (1) 636 1200 Tel. (591) (2) 340962 Canada, Inc. Haris AI-Afaq Ltd.
(54) (1) 6362185 Tlx (336) 3233 '101, 1144 - 29th Ave. N.E. Colombia P.O. Box 8141
COALAP BV Calgary, Alberta Sistemas E Instrumen- DUBAI U.A.E.
uslralia Fax (591) (2) 359268 T2E 7P1 tation Ltda. (5.£,1.) Tel. 971 4283623/4/5
Scientific & Tel. (403) 291 5215 Carrera 21, Tlx 48168 AFAQ EM
Industrial Pty Ltd. Brazil Fax(403)2915219 No. 39A-21, 01.101 Fax 971 4281285
P.O. Box 119 Philips Medical Ap. Aereo 29583
34 Wataloo Road Systems Ltda. Chile Bogota Egypt
North Ryde Av. Interlagos, 3493 - Intronica Tel. (57) (1) 2875424 EEMCO
N.S.W.2113 Campo Grande Instrumentacion Fax (57) (1) 287 2248 Electronic Equipment
Tel. (61) (2) 888 8222 04661 - 200 Electronica, SAC. I. Marketing Co.
(790) AA20165 Sao Paulo S.P. Guardia Vieja 181 Costa Rica 9 Hassan Mazher
AUSNRSI+PHILIND Tel. (55) (11) 523 4811 Of. 503 Electrocom St. Heliopolis, CAIRO
Fax (61) (2) 888 0440 Tlx (391) 55824 Casilla 16500 Apartado 7742-1000 Tel. 2022913112
Fax (55) (11) 524 4873 Santiago 9 San Jose Fax 2022913112
Tel. (56) (2) 232 3888 Tel. (506) 53 0083
Bulgaria Fax (56) (2) 231 6700 Tlx (303) 3050 England
Sophilco Fax (506) 2257887 See United Kingdom
P.O. Box 42
1309 SOFIA Croatia
Tel. 3592200785 KAL TlM- ZAGREB
Fax 359 2 220910 41425 Suo Jana, DRAGA 8
Tel.38541837115
Fax 38541 837237 ,~~ . National Sales Organizations

Ethiopia Hong Kong Ireland IL Myoung, Inc. Mexico
Fluke Europe B. V Export Schmidt Electronics Circuit Specialists Ltd. (For distribution Mexicana de Electronica
Science Park 18th Floor Plassey Technological products) Industrial S.A. (Mexel)
Eindhoven 5110 Great Eagle Centre Park, Castletroy 780-46,yeogsam-Dong Diagonal No. 27
5602 BD Eindhoven 23 Harbour Road Limerick Kangnam-ku Entre Calle de Euginia Y
The Netherlands Wanchai Tel. (353) (61) 330333 Seoul Ave. Colonial del Valle
Tel. (31) 40 644200 Tel. (852) 507 0222 Fax (353) (61) 330452 Tel. (82) (2) 55285824 C.P. 03100, Mexico OF
Fax (31) 40 644260 Tlx (780) 74766 Fax (82) (2) 5530388 Tel. (52) (5) 682 8040
SCHMC HX Israel Fax (52) (5) 687 8695
Fiji Fax(852) 827 5656 R.D. T. Equipment & Kuwait
Communications Systems Yusuf A. Alghanim & Morocco
Pacific Ltd. Hungary P.O. Box 58072 Sons WL.L. Fluke France S.A.
37 Freeston Road MTA-MMSZ kft. Tel Aviv 61580, Israel P.O. Box 223 SAFAT
Walu Bay Etele ut. 59-61 Tel. (972) (3) 6450745 Aighanim Industries Nepal
G.P.D. Box 858 P.O. Box 58 Fax (972) (3) 647 8908 Airport Rd Associated Enterprises
I
Suva H-1502 BUDAPEST Shuwaikh 13003 G.P.O. Box No. 790 I ..
Tel (679) 312744 Tel. 36 1 1662366 Italy Tel. 9654842988 ext. Ga 3-30, Pyaphal Tole
Fax (679) 300379 Fax 36 1 1611021 Fluke Ilalia S. r.l. 2079 Katmandu 3
Viale delle Industrie, 11 (Fax 965 4319070 Tel. (977) (1) 213868
Finland Iceland 20090 Vimodrone MI ALAMANA) Tlx (947) 2568
Fluke Finland Oy TAEKNIVAL hf Tel. 2.268434.1 Fax 965 48472441 ASOENT NP
Sinikalliontie 3 P.O. Box 8294 Fax 2.501 645 4846819 Fax (977)(1) 221 744
02631 Espoo Skeifunni 17
Tel. 90-5026600 128 REYKJAVIK Japan Lebanon Netherlands
Fax 90-5026 414 Tel. 354 1 681665 Fluke Corporation Electronic Supplies Fluke Nederland B. V.
Fax 3541 680664 Sumitomo Higashi S.A.R.L. Postbus 1337
France Shinbashi Bldg. Autostrada Dora, 5602 BH Eindhoven
Fluke France S.A. India 1-1-11 Hamamatsueho Kassargian Bldg. Tel. 040.644 100
37, Rue Voltaire Hinditron Services Minato-ku, Tokyo 105 P.O. Box 90 Fax 040.644.111
BP 112 Pvt., Ltd Tel. (81) (3) 34340181 1388 BEYRDUTH
93700 Draney Industry House Fax (81) (3) 34340170 Tel. 961 1 884271/ New Zealand
Tel. 1.48966363 23-B Mahallndustrial 894243 Philips Scientific &
Fax 1.48 96 63 60 Estate Fluke Corporation Fax. 961 1 898842 Industrial (Pty) Ltd.
Mhakali Caves Road Katsushige Building Test & Measurement
Germany Andheri (E), Bombay 2-3-10 Knhraibashi Luxembourg Division
Fluke Deutschland 400093, India Chuo-ku, Osaka 541, N. \I. Fluke Belgium S.A. Private Bag 41904
G.m.b.H. Tel. (91) (22) 836 4560 Tel. (81) (6) 2290871 Langeveldpark - Unit 5 & 7 St Lukes, 2 Wagener
Heinrich-Hertz StraBe 11 Tlx (953) 11 79286 Fax (81) (6) 229 1098 P. Basteleusstraat 2-4-6 Place
0-34123 Kassel HIPL IN 1600 Sint-Pieters- Mt. Albert, Auckland 3
Tel. (0561) 9594 0 Fax (91) (22)836 4682 Jordan Leeuw Tel. (64) (9) 849 4160
Fax (0561) 9594 119 Jordan Medical Tel. 02.33 12777 Fax (64) (9) 849 7814
Philips India Ltd Supplies and Services Fax 02.33 11 489
Greece Hindustan Times House P.O. Box 140415 Norway
Philips S.A. Hellenique 18/20 Kasturba Gandhi AMMAN Macedonia Fluke Norge AlS
P.O. Box3153 Marg Tel. 9626699353/4/5/6 AD TEHNOKOM Ole Deviks vei 2
15, 25th March Street New Delhi 110001 Fax 9626663556 Gradski Trgovski Centar 0666 Oslo
177 78 TAVROSI Tel. (91) (11) 331 8370 3 Sprat Tel.: 22.65 34 00
ATHENS Fax (91) (11) 331 6839 Korea 91000 SKOPJE Fax: 22.65 34 07
Tel. 30 1 4894911 B&P International Tel. 389 9 91 115780
4894262 (direct) Indonesia Co., Ltd. Fax 389991 225353 Oman
Fax 30 1 48185941 P. T. Daeng Brothers Geopyung Town A"1809 Mustafa & Jawad
4894434 Philips House 203-1 Nonhyun-dong Malta Trading Co.
J/n H.R. Rasuna Kangnam-Ku CAM Services Ltd, P.O. Box'1918
Said Kav. 3-4 Seoul, 135-010 "CAM CENTRE" P.C.112
Jakarta 12950, Tel. (82) (2) 546 1457 Triq 1-lndustrija RUWI-MUSCAT
Tel. (62) (21) 520 1122 Fax (82) (2) 546 1458 QORMI QRM 09 Suit. of Oman
Tlx (796) 62189 Tel. 356 4846401 Tel. 968 709955
PHDC IA 484650 Fax 968 564005
Fax (62) (21) 520 5189 Fax 356447174 · Fluke Corporalion

PO Box 9090, Everett WA USA 98206

Fluke Europe B V

P.O. Box 1186, 5602 BO Eindhoven, the Netherlands

For more informalion call

In lhe U.sA (800) 44H853 Dr lax (206) 3565116 In Europe (3HO) 644200 or lax (31-40) 644222

In Canada (905) 890-7600 or lax (905) 890 6866

F rom other cOl!ntries + 1 (206) 356-5500

© Copyrighl1994 Fluke Corporation

Ali rights reserved

The wordmark Philips amllhe Philips shieldemblem are used under license from Philips Electronics NV

Data subject to change without notice.

Printed ill the Netllerlands 94-36

G0340A-14U94051NL EN

LUK

[.t<)