SERVICE INFORMATION FROM HEWLETT-PACKARD
RI MARCH-APRIL 1980
Basic Techniques of Waveform Measurement
Using an Oscilloscope
I i d i t d s N t : This article is based Once these concepts are mastered, CRT, an electron beam draws the
on a three-prt uideotape series (HP the only remaining hurdle for you is waveform on a phosphor-coated
pln 907410) originally developed to to locate the controls on the scope’s screen. T i screen presents three
train customer technicians on the use front panel. Most manufacturers try types of information: voltage infor-
and operation of a n oscilloscope. to help you by grouping similar con- mation on the vertical or Y axis,
While most of the references EO eon- trols together and separating the dif- time infarmation on the horizontal
trols are based on an HP 174OA ferent gmups by color or lines on the or X axia, and intensity information
Dual-Trace Owilloscop, the infor- front panel. on the Z d s . All oscilloecopes have
mationpresented is bagic and applies controb to @ust the voltage, time
to the opernhn of other munufactur- GETTING BACK TO THE and intensity idormation in order to
em’ oscilloscopes as well. BASICS present a meaningful picture on the
I The oscilloecope presents a voltage CRT. Figure 1 shows a block &a-
Generally speaking, a technician be- vs. time display of a waveform on a gram of the basic circuits that these
coming familiar with a piece of test cathode ray tube (CRT). Inside the controls operate.
equipment is concerned about three
0 Knowing where and how to con-
nect the test instrument,
!I lF 0 Knowing how to adjust the
oscilloscope. Why the oscilloscope?
Because it is probably one of the
most versatile troubleshooting in-
struments you have on the bench.
You can use it to measure voltage
levels (from dc to microwave), phase
differences, signal presence (or ab- POWBI Tngg%Aniro1s
Figure 1. Block diagram of a basic oscllloscope
distortion, and complex
analysis (wave shape, Part 1 of this article describes most of
etc.) to name a few. We the basic controls of an oscilloscope IN THIS ISSUE
n’t show you how to use and how to set them up to make a WAVEFORM MEASUREMENT
all these endeavors. We basic measurement. Part 2 will con-
clude the series with SERVICE NOTES
cal period, pulse rise
propagation delay, and time-delay
I 5420A OWNERS
- , , J
The Vertical Input
As shown in Figure 1, the input sig- trol incmatpea time interval meas-
nal is connected to the vertical input exampie 1740A-sCape has four p i -
amplifier. The vertical amplifier %ions: AC, GND,DC and Solb. The
either attenuates or amplifies the
signal for convenient viewing.
The next block the incoming signal
encounters is the delay line. The
delay line allows the sweep
generator circuitry time to start a
sweep before the signal reaches the
CRT vertical deflection plates. This
coordination of vertical and horizon-
t l timing by the delay line enables
a Now l&e look at the h n t panel of
viewing of the leading edge of the an HP oacillowqk in Figwe 2 and
signal. This will be explained in see how the controls that operate low impedance sollIm?s.
greater detail later on. The vertical these circuits are identified and
output amplifier provides additional AC Position - The AC position
grouped on this particular model. couples the input signal through a dc
amplification that is required by the
CRT vertical deflection plates. bl capacibr, allowing only the
VERTICAL INPUT CONTROLS ac component t be viewed. AC mu-
The Time M s
Although precise horizontal deflec-
tion rates are not required in many
general purpose applications, the
more sophisticated scope applica-
tions require precise control of the
quency digital-type signals. The
internal dc blocking capacitor will
distort the waveform as shown in
GND Position - The ground
(GND) position is useful when you
want to set a @;roundor zero volts
reference level on the CRT screen
VVVVVV HPARCHIVF CnM
WAVEFORM MEASUREMENT 1:: :m i -1 1 I-
and low level signals will
attenuationlmore gain. The
VOLTSDW 6ont;mll allme you &o
change the v e r t h l mdtivity in cal- t u r n g d offin &hitit ym would
This means that you can leave the ibrated fixed steps,from 20 volts per see the retrace lines with every
input signal connected to your scope. division to 5 millivolts per divisionon sweep.
You won’t short it out when you the 1740A. The vernier portion of the
switch to the ground position. input attenuator provides continu-
ous sensitivity control between the
NOTE calibrated volk-perdivision ranges.
Whenever you move the vernier out
High frequency signals can of its detent position, the UNCAL
create special problems for light will be on, letting you know that
switches in scopes as well as the steps marked on the VOLTslDIv
other instruments. Therefwe,
when mearsuring HF signals, it
allows you t view both dc and ac
compoaents of the input signal. For 40 MHz. The vertical magnifier is
example, if you have set the 0 volts umhl when you’re tryingto measure
reference level at the center of the low-level signals such as power sup-
screen (using the GND position) and ply ripple.
then switch to DC, the waveform
nent, if any, and the signal will
offset either up or down depending The sweep generator, sometimes
on whether the dc component is posi- called the time-base generator, pro-
tive or negative. DC coupling is also duces the sawtooth wavBform which Sweep Speed Control
used when you are measuring s the rate the beam is drawn
digital-type signals or quare waves. horiaontdly the face of the The sweep generator’s sawtooth
CRT. The generator‘s most impor- waveform is controlled by a front
n is tant function is to ensure linear panel control called TIMEDIV or
a dc input (no blocking capacitor) beam movement, meaning the beam SEC/DIV. This calibrated control
with the Xc of the input amplifier moves at the same rate from start to lets the operator select many M e r -
very large compared to 50 ohms. The finish. Without this precise rate, ac- ent sweep speeds in order to view
5 input is used to measure high
W curate time measurements are not waveforms that vary from a few
speed pulses and square waves from possible. Another factor of accuracy Hertz up to the bandwidth limit of
50 ohm sources with minimum dis- depends on the delay line. Its func- the scope. The control is usually
tortion and VSWR reflections. Most tion is to delay the vertical input divided into steps i n a 1-2-5
oscilloscopes with a built-in SOCi signal just enough so that the trace sequence covering the ranges of sec-
input have internal compensation being displayed is the signal that onds, milliseconds, microseconds,
that make it a better match than an started the sweep (see Figure 4). and nanoseconds. These ranges cor-
external load. relate to how faet the beam is drawn
Another function of the sweep across the CRT. The faster the beam
The Input Attenuator Control generator is CRT unblanking. An is drawn across the CRT, the faster
Most modern scopes use a combina- unblanking pulse is a positive the time reference (i.e., the shorter
square wave that turns the trace on the scale). For example, if the
a tion of variable attenuation and ad-
justable vertical amplifier gain to
control input signal levels. High
level signals will require more
in relation to the rising portion of
the sawtooth. What this means is
that the trace is turned on during its
TIME/DIV control is set for 0.5
seconds-per-division, the time refer-
ence over the full 10 major divisions
attenuatiodless gain so that the left-to-right movement across the (vertical graticule on the CRT face)
I trace is not deflected off the screen, screen and then turned off during is 5 seconds. If it’s set at 5
11 WAVEFORM MEASUREMENT
scale time reference is 60 milli-
steps. Whenever you move the ver-
nier out of its detent CAL position,
factor the magnifier is labeled. For
example, if yo& scope has a 10 divi-
sion time arris (10 squares on the
seconds. Figure 4 shows how the the UNCAL light will be on letting
sawtooth waveform produced by the you know that the steps marked horizontal axia) and the magnifier
sweep circuit develops a sinewave on the TIME/DIV dial are not has a factor of ‘Xlo’, you would have
pattern on the CRT. calibrated. an effective 1OO-division wide signal
and a lo-division window. This also
P r of the TIMEDIV control is a
at Another control that interacts with
means the signal has 1 times the
s e p vernier control that provides
we the sweep speed control is the hori-
horizontal resolution as before.
continuous adjustment of the sweep zontal magnzier. This control ex-
speed between the fixed TIMEfDIV pands the sweep time by whatever
Rise Time and Bandwidth
OsciIIoscoDe Amplifier Considerations
Oscilloscope users generally frequency components will reach requirements for scope band-
consider a scope’s bandwidth the deflection plates at the same width using the 5 times faster
and rise time as its primary pa- time. This results in minimum and 0.35 factors together can be
rameters. And rise time is usually distortDon of complex waveforms. estimated using the following
considered the more important Note that this Gaussian re- rule of thumb:
parameter when working with sponse is always falling in gain, 1.70
faster waveforms. This is mainly therefore, accurate voltage (minimal) Fastest Rise Time
because the primary axis of the measurements can only be
scope’s display is the horizontal made at dc. The frequency re- But remember, very accurate
or time axis, and it offers the sponse will be down 1.5dB at absolute rise time meas-
greatest rmolution - less than 20% of the 3dB bandwidth, so urements are not always i m p -
2% for timing f n e a w r m t s . 3% accurate amplitude meas- tent. When simply comparing
Why is t horkontal or time
b urements of sinewaves can’t be the rise times of two signals,
axis txmkfwd the major axis? made for frequencies greater scopes with a tfse time equd t o
ComMar that the vertiial axis than 20 MHz on a 100 MHz the rise time of the signals
has an 8cm window, whereas oscilloscope. However, the applied am usually consided
the horizontal axis has a lOcm amplitude of a pulse is dc so ac- adequate.
window. 10cm provides more curate pulse amplitude meas- In conclusion, it can be said
resolution than 8cm. Also, the urements can be made up to the that the modern oscilloscope
range of the vertical axis (the HP full bandwidth of the scope. with its Gaussian response is
1749 for example) is 2,000 to 1 designed for pulse parameter
Constraints make bandwidth analysis but not sinewave
or from 1 mv/cm to 20 v/m. The and rise time numerically related
time axis has a range of analysis. The characteristicsof a
in well designedgeneral purpose sinewave can be better meas-
4O,OOO,ooOto1from 2 W c m to oscilloscopes. Bandwidth in
50 ndcm. This is 20,oOO times ured with instruments other than
megahettz multipliedby rise time the oscilloscope. For true RMS,
greater than the verticai axis of- in nanoseconds is approximately
fers. a voltmeter can give better
0.35. Therefore, if your oscillo- amplitude measurements, a
Signal bandwidth is of course scope needs are defined in
defined as the frequency range counter better frequency meas-
terms of one factor, for example urements, and a spectrum
in which signals are handledwith rise time, Widing it into 0 3 will
less than a 3dB loss compared to analyzer better distortion meas-
produce bandwidth. urement$. However, for a com-
midband performance. However,
the wertical system of an oscilb In terms of rise time, scopes plex waveform such as a pulse,
scope is not flat like that of a ideally should have a vertical the oscillosocpe is clearly the
voltmeter - it is Gaussian. system capable of responding at best choise. The voltmeter can’t
What does Gaussian re- least three to five times as fast as respond fast enough to make
sponse mean? It means that the the fastest applied step signal. In this measurement. The trigger
vertical system of the scope al- such a case,the rise time of the f
uncertainties o a counter mask
ters the input signal and delays it signal indicatedon the scope will its accuracy for pulse meas-
in such a way that it produces a be in error by less than 2%.For urements, and nothing but a
linear phase response. The example, if you are going to ac- scope can measure parameters
linear phase response has a curately measure ‘X’ micro- such as overshoot, droop, and
constant group delay so all the second pulses, the minimal ringing.
o€s;hie series will show you
how to make the Same measurement
using other controls t o obtain
Frequency is the d p m d of the
time period for one cycle. For exam-
ple, the time period of the signal
shown in lFigtve 7 is obtained by
counting fbjg.nam&erof h&&d
ma). Then take the reciprocal.
t 6X 0.2m8 10-0
. I I
WAVEFORM MEASUREMEN 1
of horizontal tangent points, and di-
vide this number by the number ob
Figure 9 represents a frequency re-
sponse curvi and is obtained by con-
vertical tangent points. If you u e
s necting a sweep generator to both
this method, always make certain the input of the circuit under Cest
that the tram contains visible cross- X
and the ' 'axis. The output of the
overs, that they are not masked Y
circuit is connecW to the ' 'axis.
by trace coincidence; that is, the The oscilloscope becomes a simple
horizontal tangent points don't fall network analyzer that is a swept re-
together. ceiver that pmvi?es a visual display
The ode of operation is a two-
dimensional representation of two ac
voltages. The vertical or ' 'input
signal deflects the beam up and
down while the horizontal or ' ' X
input signal replaces the scope's
sweep generator and deflects the
beam horizontally. A third dimen-
sion can be added by modulating the
beam's intensity through the ' ' 2
One of the more co 868 of
X-Y mode is to generate Lissajou
patterns to check phase. For exam-
ple, the transistor checker discussed
in the Sept.-Oct. and Nov.-Dec. 1974
issues of Bench Briefs provides a
Lissejous pattern that indicates the
voltage-to-current characteristic of a
diode junction. Another more sophis-
ticated use is in the area of circuit
freauency response where you turn
Figure 8 shows wme of the various
Wajous patterns you can expect
using the X-Ymode. Note that Fig-
ing a 1 2 frequency ratio. T obtain
the ratio of vertical and horizontal
deflec%ion frequencies from any
Lissajous pattern, count the number
derence. It also
%weep gemmator in order k, produce
a reference bamline if there is no Normal Mode -The Normal
between the horizontal and
signal connected to
ut of the wow, you
gger signal -hence
trace. This loss of trace with loss
AUTO Connects sweep to multivibratar so
sweep free-runs in absence of adequata,
NORMAL Connects trigger circuit to one o three
sources: input signal, external W l or
TRIGGER CONTROLS SINGLE Sweep will start only upon the occur-
rence of the trigger signal that meets the
condiths of level and slope after the
sweep signal discussed earlier so INT Uses sample of input signal to start
that each trace is written right on
top of the previous one. You see one
single trace, but it is actually being EXT Uses sample of some external signal to
refreshed on each sweep. start sweep. External signal is usually
related to input signal.
Several controls allow you to select
the source, positive or negative
Source EXT t tO Same as EXT except attenuates external
mode, and level of the synchronizing signal by a factor of 10.
trigger signal as shown in the
simplified diagram Figure 1.The fol- LINE Uses sample of power source. Useful for
lowing table is an abbreviated de- viewing events related to power line
scription of the basic controls and frequency.
VARIABLE Permits selection of triggering at any
-1.5~to +1.5~(EXT) r
point (level) on the posithre- o negative-
level -15v to +15v (EXT+10) going edge of the displayed waveform.
This switch is probably the greatest
source of ‘bilot error” in oscilloscope
pas (+) Sets up the t r i m n g circuit so that the
operation. In simple terms, Normal Slope NEG (-) displayed signal is triggered on the
m d requires a trigger signal to
oe positive- or negative-going edge.
generate a sweep - Auto does not. 1
of input can be a valuable trou- (depending on how you have set the One method of viewing the time re-
bleshooting aid. Say, for example, Slope and Level controls).This lationship between the input signal
you are probing a circuit looking allows you to view a time event and external trigger signal is with a
only for presence or absence of a sig- related to the input signal. If you dual-channel scope. U e one input to
nal. If you adjust the trigger level are using a dual-channel scope, you look at the signal and one input to
control for an optimum level, and must know which input channel will look at the trigger. You must know
then probe a point in the circuit that trigger the sweep circuit and use which input channel will trigger the
has no signal present, there will be that channel for your input. Part 2 of sweep circuit and use that channel
nothing to trigger the display and this article will explore some of the for the trigger input. Then set the
the screen will be blank. Figure 1 various triggering options available source switch for INT.
shows a simplified representation of for dual-channel viewing. If you are going to use an external
how the trigger controls are If you are using the internal trigger trigger signal, it is advised that you
interlocked. mode for troubleshooting, you may first look at that signal on the input
have to re-adjust the trigger level of your scope. You must determine if
Trigger Level and Slope control to maintain a trace as you it has a dc component or noise
probe different points in the circuit greater than the trigger level you
Trigger Level and Slope controls under test. The reason this occurs is are trying to set up (or possibly ex-
allow you to select any point on the because the trigger circuit has been ceeding the limit of the input). For
positive or negative edge of the dis- initially adjusted (by you1 to trigger example, the trigger level range of
played waveform to trigger the the sweep at some positi e or nega- your scope may be 2 1 5 volts (215
sweep circuit (see Figure 10). Usu- tive voltage level. There re, as you through EXT + 10). If you try to use
ally, when the scope ia in the Inter- move the probe from pomnt-to-point an external trigger signal with a dc
nal Trigger mode, the level control monitoring different si$nal levels, component greater than 1.5 volts,
will select any point on the vertical the voltage level to the trigger cir- you won’t be able to trigger the
waveform displayed. With external cuit is also constantly changing. To
trigger signals, the control has a &
voltage limit (refer to the operating
eliminate this inconve ience, use
the External Trigger m de and con-
nect the external trigger to a low rep
sweep unless you block that dc.
Some scopes have ac coupling
(selectable) built in - other do not.
At any rate, you must use dc cou-
rate timing signal from the circuit pling for trigger signals below about
Internal Trigger under test. In a digital circuit, use a 20 Hz.
sub-multiple of the clock pulse rate.
When the switch is set for Internal Your external trigger signal may
Triggering, it means that a portion External Trigger also have power line pick-up or
of the input signal is tapped off, as When the switch is set for External possibly RF noise. In either case, you
shown in Figure 1, and sent to the Triggering, you must provide a sig- need to filter out the unwanted por-
trigger circuit. The CRT will display nal to a connector on the scope tion in order to obtain a stable dis-
aportion of the input signal related marked EXT TRIGGER. If the sig- play. Some scopes have built-in fil-
to the first occurrence of a positive or nal voltage exceeds the input volt- ters while others do not. The point is,
negative slope of the input signal age limit (refer to your manual), if you use external triggering, make
then use the EXT + 10 trigger input. certain the signal is clean.
A good rule-of-thumb is use a 1O:l
probe on EXT and no probe on EXT Line Trigger
+ 10. This will help prevent satura- In the Line mode, the display is
tion of the trigger cornparitor. The triggered by a sample of the power
External Trigger signal is usually line which is usually 50 or 60 Hertz.
derived from a low rep-rate timing Line triggering is often used when
signal related to the input signal. you want to determine if there’s any
The CRT will display the input sig- relationship between the displayed
nal on each occurrence of the trigger signal and the line frequency (often
signal. This allows you to view an called power line hum).
event time-related to the trigger
source. The Trigger Level and Slope Trigger Holdoff
controls work the same for an exter-
nal triggered signal as an internal Some oscilloscopes may have this
triggered signal. specialized variable control that is
WWW. H PARCHIVE.COM
111 :: i
used in coqjunction with the Trigger Set the input coupling switch ac- Axiom #2: If the trigger signal has
Level control. Trigger Holdoff cording to the foIlowing criteria: a large dc component, it
increases the time between sweeps
and helps stabilize the display when
- ifthe source is a pulse or must be blocked by a
capacitor (e%., 0 1 p )
internally triggering off a complex If your sc#p hae d d c
digital signal or RF signal. - DC if the source is a low- s e l d o n built into the
frequency digital signal trigger input controls,
PUTTING IT ALL TOGETHER (square wave). use the ac position to
Now that you have an idea of what -AC ifthe source has a large DC block t b unwanted dc
all the basic controls are for, let‘s put companentthat needs blocking compo.lkent.
I them all together in step-by-step or for general purpoge probing.
L Select the trigger murce. You can
I order to actually set up your scope. Connect the input eignal and ad-
4 just the input atbwtor control trigger the sweep from an exter-
1 Turn-On and Preset,(beforethe
. to obtain a reasonable display. nal, internal or line frequency
0 Adjust the sweep speed control
0 Turn on the power and allow ap- until you get a display you can
proximately 30 secohds for warm recognize. 0 If the frequency of the trigger sip
nal is less than approximately 40
2.Fine Tuning H r z change the mode switch to
0 Preset the trigger mode switch to Normal.
Auto and turn the intensity con- The following steps are contingent
trol up. on the type of oscilloscope that you If you have selected external trig-
If there’s still no display, use the have. A lot of brands won’t have all ger, select either ac or dc trigger
beam finder together with the of the controls that we have been s
coupling. U e ac if the trigger sig-
horizontal and vertical position discussing, and wme brands may nalcontains-a l e e dc component.
controls to bring the trace to cen- have more. “he point is, the theory Uee de if the trigger eignal fre-
ter meen. is the same regardlees d what the queney is less than tlJOi%rt%.
control is caned or even if you have
0 Adjust the intensity control to a Om.
comfortable viewing level. Adjust 0 Always u ~ the EXT + 1 input if
the focus control for sharpest NOTE you are not using a divider piwb
trace. The following axioms apply to connect the external trigger
0 Adjust the input attenuator con- when you are using an external signal to the scope.
trol to its highest setting. This trigger signal.
will prevent the trace from being Axiom #1: The trigger signal must at
P r 2 will go into more detail on
deflected off screen if the signal be clean and free of triggering and probe equaliza-
has a large dc component or is a noise. If your soope has tion in preparation to making
very large ac signal. built-in filters,use them. measurement&
345549 Digital Voltmeter
c Does your 3455A “hang-up” in its
turn-on mode? Or &oes it eometimea
display out-of-spec low frequency
c readings whea in normal AC, or in-
termittently read incorrect data
when autoranging? If so, there are
some resistodcapacitor changes
listed in service notes 3455A-7A and
3455A-15 that can correct and im-
prove the voltmeter’s reliability.
Order the 3455A service notes with
the order form at the rear o Bench
I * l
B OBClLLOSCOPE seride1622AW3Wandbelcrw.
ala PrefemdreplacementforhQh imprarebUeRMSbWhqUOnCy
A m . a reliaMuty.
-16. Serials 1622AO53W and below.
SAIB OSCILLOSCOPE to replace relay AlOKl .
1 senals. Preferred replacementtor high
12Q7AIBO ~ l L L O s c o p E 3465A
1207AlB5A. All serials.Preferred mphcmmt for high 80024.
voltage oscillator A 34656-2. AU serials. R
34656 b e t t m Battery P/N tor HP uI)(1: OOO91-
1208A OPTION H11 80013. Customer use: Accesllay 82(M3A
H11-1208A-12. serlals 3581C (C MODEL ONLY) WAVE ANALYZER
repkroement for high voltage oscillator A604. 3581AIC-5. Sdals 1411A01107 and below, and 88
Hll-1208A-13. Serials 1330A arxl below. Preferred rids 1411A01110, 1411A01 113. 141lAOl114,
rwbment for hlgh voltap OsOHlptor A m . 1411A01116. Audlo amplifier modification to
121SAIB OSCILLOSCOPE p a ~pertormaw.
VOItage oscillator A m . 3582A SPECTRUM ANALYZER
358ZA-2. Serials 03582-00151 to 035BUMBlD. Mdi-
voltage oscillator A6Q4. 3745NB SELECTIVE LEVEL
1812U and b.low. R
-. serials 184344
replacementto inxcwsB the relicrbi#tyofthe X and Y 1812U and bebw.
instruments you own, remove the amplifiers. f& mlopUons H15 and H16.
374WB-298. SerlaJs 1815U. 1831U and 11t;98u,
order form and mail it to the HP 1304A DISPLAY Modificatlon to prwnt ptmft&a remote SlMtUp
distribution center nearest you. 1715A and bebw. problem.
374wB-30. serials 192Ou and below. Preoanrrd re-
placement of Memory Assembly A109.
1333A X-Y DISPLAY
374WB-31. 3745A S r a s below 193011-00347
. - .-
and 37458 Serials belaw 192AUaa201. Prefer-
- . .
red replecemeot Capecitor N% CI
1333A-6. All prefixes. Preferred parts r to 3745A/&32A. 3745A - s r a s below 1
r l i of z-uds amplifier
w y and 37458 - Serials belaw 1924uQ0201. P@W-
red replacement of Integrated circuit A211IC@,
ptetened replacemenl to im 3747AI0 SELECTIVE LEVEL
3747AlB-W. Serials 1804U and below. #&Man d
1350A GRAPHtCS TRANSLATOR lOMHz crvstal tmar assemblv to improve
135OA-5. All serials.Program modication to eliminate
a bright spot which appears on screen when writing 3
into W o n 2047 or the data in location 2047 is not
16OOA LOQlC STATE ANALYZER 3
16WA-2. Serials 1714AOW llnd bekvr. ModiRcation
to prevenl inconed deta storage into “E” taMe. 3
tion to prevent unstable Input autorangin&
1 W A SERIAL DATA ANALYZER 3747NB-10. 3747A - Serials belovv 1924U-00139
164OA-7. All serials. Prefermd parts replacement for -
and 37478 seriela below 1924U-00115. Prefer-
8251 model USARTs. red replacement of cepacita A311c3.
3747-11. 3747A - Serials WW 1924U-00139
1722NB OSCILLOSCOPE and 37478 - Serials below 1924U-00110. P ~ c -
1722A-4A. All serials. Extender board kit to facilitate red replacement of integrated circuit A21 1IC%.
troubleshooting. 3747AlB-12. All serials. Installation of Option 022 to
1722A-6A. All serials. Preferred replacement for at- facilltnteCMessage weighted filter and phaseJitter.
W8UF SIGNAL GENERATORS tenuator detd wheel.
F-20. -E: Serials 1826A and below and 608F 3770A AMPUTUDUDELAY
serials. Prefermi part replacement for series 1744A OSCILLOSCOPE MSTORTION ANALYZER
w transistor 01. 17-1. SeriaQ, 1915A and below. Modiication to r e 377OA-37A. All serials. RetroRtting instructkwm for
612A SIGNAL GENERATOR duce ‘‘writethrough’ and defocurJng. option 001, +l&JBm switoh.
612A-11. SeMs 1826A and below. Preferred pert re- 3771NB DATA UNE ANALYZER
pkcammt for Wes regulator transistor 01. 2 W A QUARTZ THERMOMETER
612A-12. W’W 1914A and W. ModificaHon tor 2804A-2. All serials. Modification to prevent random 3 7 7 1 N M Serlals below 1926U. Modificatiantopre
reset. vent gain hits fmm causlng dropouts.
impovcld~moduhtkn~~uacy. 3771AlB-6. Serials bebw 1937lJ-00160 (3771A) end
680 STRIP CHART RECORDER 3OWA BOARD TEST SYSTEM 1937U-00123 (37716). replacement of E l
306OA-1. All serials. Proper configuration to prevent Mains Module.
-74. All serials before 1200. Modification to im 377WB-7. All serials. Retrofitting instructionS for
p o safety grounding.
missing charactem on 26408 CRT.
OptlOn 061, Rack Mount.
6E0-8.AU serials. Preferred gear plate in tratmmhkm 377lNB-8. All serials. Preferred replacement of
mmnblbe. 3437A SYSTEM VOLTMETER switchesA38s1andA4os1.
3437A-3. All serials. S r i e procedure to fix sticky
evc 3771NB-9. All serials. Table of link positions for
12oouB OSClLLOCOPE hwa PQnel pushbuttons. 377lA, 37716 and Options.
w-mentforhigh 3437A-4. SefW 1630-70 aml below. Recom-
m d d fa Aau124. 3779NB PRIMARY MULTIPLEX ANALYZER
377W-10. serial 1919U-00155 and Modwiorr-
SCOPE 3466A DIGITAL VOLTMETER t
tkn to prevent keekdom o f m 2 a wlt&-on.
rep(ecementforh@ 3456A-7A. Serials 1622A05650 and below. ModiRca- 3779A-11. Serials 191BU-00170 and below. Modlflca-
uon to improve tum-on reliebility. t h to pIII)v(Mt feilwe of Int~atad h d t k w 9 .
WWW. H PARCHIVE.COM
Attention 5420A Owners
This issue of Bench Briefs contains AN 240-0 TIME AND
several service notes that may be of FREQUENCY DOMAIN
interest to you. of primary impor- MEASUREMENTS
tance is service note 542OA-21 that
Defines, discusses, and illustrata a
groups all necessary modifications
wide range of important meas-
that are important to the reliability
urements made by the 5420A Digi-
of the 5420A Digital Signal
Analyzer. tal Signal Analyzer using sampled
data techniques. Includes Time
Record Averaging, Auto and Cross
Owners of the 542011 can order the Correlation, Auto and Cross Spec-
service notes with the order form lo- trum, Transfer Function, and more.
AN 240-1FEEDBACK CONTROL
Illustrates the application of the
Also available are three application 5420A Digital Sigpal Analyzer to
notes designed to help you obtain both time and frequency domain
maximum use from your 5420A. measurements of feedback systems.
Safety-Related Hewlett-Packard Corporate Stand-
ards have been changed to require
4960A/BP i Identifier
Service Notes from H p relating to
an instnunent to be able to pass 25
amps through the safety earth
ground. Instruments with serial
numbers below those listed below
personal &ety and possible equip
ment damage are of vital importance
to our customers. To make you more
were manufactured before this
safety standard came into being.
While there does not appear to be a
aware of these important notes, they
are printed on paper with a red bor- hazard with the old circuit design,
der, and the service note number has we are notifying our customers that
a modification is available and can hock hazard may exist
a "-S"s e . In order to make you
immediately aware of any potential be made at customer's expense, if
safety problems, we are highlighting desired.
safety-related service notes here
with a brief description of each prob-
lem. Also,in order to draw your at- 204C - 0989A14605 and below 4960B serials 17371900750 and
tention to safety-related service 204C-H20, Option 02 - beI@V?
notes on the service note order form 0989A14855 and below
at the back of Bench Briefs,each ap- 2 4 - 1105A03415 and below
propriate number is highlighted by 209A - 1045A05846 and below
being printed in color. 608 - All serials before 1200
204C/D, 209A OSCILLATORS
AND 680 STRIP CHART pear at the front panel fel+me
RECORDER The modification consists of install- line connectors. Installation d an
ing a solder lug to a chassis point insulator, HP Part #04960-00212,
and connecting a 18 GA GRNNEL removes thi~ hazard.
wire from the power line connector
earth terminal to the solder lug. For For complete detailed instm&i 061181,
complete detailed instructions, order please order the approprirtte! wmim
the appropriate service notes using n o w using t e form at the
the form at the back of Bench Briefs. Ben& Briefs.
WVWV, H PARCH IVE,COM
0 Includes a technique for obtaining
the open loop gain from data taken
STIMULUS-RESPONSE TESTING USING THE IC TROUBLESHOOTERS
with the loop closed and the system
in normal operation. Several practi-
cal examples are illustrated.
AN 240-2 IMPROVING
ACCURACY OF STRUCTURAL
Vcc o Ground
I P W
Illustrates how to idencify and cor-
rect mass loading and accelerometer
loading errors. Shows the “ideal”
measurement system where the test
system and exciter appear on sepa-
rate sides of the load cell. Describes
“real” measurement set-ups and how
to correct for mass loading in both
shaker and impact test systems.
1. A node is an interconnectionbetween two or more IC’s.
2. Uee the Pulser to pIovide stimulus, o use normal circuit signals, whichever is most convenienl
3. 547A Current Tracer WlO!X?!jE,TLogic Probe
548A L q i i clip o$
5uw10526-r L ! c Pukef
practical terms the technician often The logic pulser changed this be-
J uneoldered and lif€ed an IC leg, or atwe it delivers both high current
cut a circuit trace, then used a pulse and low total energy by gelleratiag
generator to drive a gate’s input. very short pulses sufficient to
IC TROU6LESHOOTl)rso TOOLS momentarily override TTL logic
P The “why” of such destructive tech- LOW states. Pulse. width is never
In the digital troubleshooting world niques has to do with the internal su&ciently long to degrade a gate’s
the idea of stimulus-responsetesting structure of standard ‘ITL gates. A performance (the pulser will usu-
s is a relatively new one. Tools to ‘ITL gate in its LOW state is a satu- ally generate a ‘ITL HIGH for only
measure logic states have existed for rated transistor to ground. To move about 500 nanoseconds).
quite some time, but forcing a state a ‘ITL output HIGH requires a great
change, especially on a line being deal of current drive. The catch here To help show how straightforward
held LOW is something that was dif- is that continuous high current digital stimulus-response testing
ficult prior to the introduction of HP tends to destroy the ‘ITL gate’s out- can be, the following table outlines
Logic pulsers. T accomplish such a
o put transistor. So, it was usually seven node and gate troubleshooting
1 task meant disconnecting a device’s safer and easier to simply disconnect problems and how pulsers and other
! input from the circuit, and then a driver from a circuit and replace it IC troubleshooters would be used to
pulsing the input with a source. In with a low current stimulus tool. pinpoint the fault.
STlM U L US-R ES PO NSE TESTlNG
ening current density in the
7. When you think you've located the
moving the pulsing pint to the
paths should be &etected el
you pulse directly mrom a short.
As nondestrudively as possible, remove the suspectsd
component and verify that the Vcc to ground short no
Node and Gate Troubleshooting Examples
Example 1 b) Set the current tracer derence level by aitfi.
until the light just lighta
U e the current tracer to follow the current pafh to
A frequently o c d n g troubleshooting symptom is a
this erample,the problem is to deter-
er h dead, or ifa shorted i p t i clamp-
i nu,s Here are two D-Flip-Flops. One w o r b n o d
other doesn't change output state although
ditions are identical for both.
a) U e the pulser at the D input ami probe Q
888 if the outputs &.- (In this actuaL
vv\Ivw HPLLRCHIVF CnM
r STIMULUS-RESPONSE TESTING
0 b) The Reset line in the above case was found to be
stuck in a LOW state by using a logic probe. Pulsing
and probing simultaneously indicated the Reset
line couldn’t be driven HIGH,indicating the line
was shorted to ground.
c) Further use of pulser and current tracer showed
that the area near the Reset line drew current when
pulsed and that the D-FlipFlop would not perform I t
operations when the Reset line was pulsed.
d) Using the pulser and tracer,the operator foun
hairline solder bridge from the Reset line to
over two PC bo
I Example 3
a) Use the HP-10
node. The comparator is clipped onto the suspected
IC and us88 the in-circuit signals to drive an identi-
cal IC installed h i d e the comparator. The com-
parator identifies AlU8-13 as bad allowing you to
ClmnNlT troubleshoot further.
b) Probing and the node indicates it is stuck
The node between U1 and U2 was stuck LOW when LOW.
measured by a logic probe, although probing revealed
pulse activity at Ul’s input. c) Pulsing and current tracing at A
a) Probe U1 pin 2. If no voltage activity is present, current is flowing toward PC board
0 pulse pin 2 and see if the node’s state can be
d) A2U1 is sinking current andholding the node
LOW. As a result, U2 is not being clocked. The com-
b) In this case the state couldn’t be altered using parator located the failure (A1 U8 pin 151, but it
pulser so current was traced from U1-2 to U2-9, th required the tracer to indicate current flow to veTify
J-Kflip-flops input. A2U1 as the c a m
Example 4 Example 6
Outputs A, B, C, and D are LOW; other circuit inputs U1 tests bad using the probe and pulser. The problem
appear normal. is to find out the nature ofthe Eault before h t i w h g
a) Use the probe and pulser to make sure A, B, C, and what appears to be an internally aharted IC.
D aren’t grounded @robe and pulse each pin - if a) Pulse pin 12, and obsem with the probe cluzt pin 13
ungrounded the states will be changed by the direction ClB and 13
b) Probe other pins on the IC and check for normal/
abnormal indications. b) Pulse pin 12 and
c) Measure current at pins A, B, C, and D by pulsing
verse the two instruments. Current is.identica1 in
each pin, and tracing to see if current flow is indi- both directions.
cated from the pulser to the Shift Register.
0 d) In this example all signals are normal except A, B,
C, and D. They are stuck LOW, and are not indicat-
c) Pins 12 and 13 are
bridge on the back of the circuit board. Although
originally located by pulser and probe, the tracer
ing current flow, which suggeats an internal failure adds important information that keeps you from
in the IC, and not in the circuits connected to it. removing the IC.
vcNCti u n i u
Volume 20 Number 2
Service lnformatlon from
D obtain a quallficatlonform for a fre
Ion, send your rsquest to thi
Peader comments or technical article
contrlbutlon8 are welcomed. Please
send them to the above address,
attention Bench Briefs.
tor: Jlm Bechtold, HP Mt. View