EG&G 128A Operating And Service Manual

MODEL
128A
LOCK-IN
AMPLIFIER
SEE SAFETY NOTICE
PRECEDNG SECTION I
;EFORE OPERATING INSTRUMENT
AND
SERVICE
MANUAL
M128A; 9/83·500·MIC
MODEL
128A
LOCK-IN
AMPLIFIER
OPERATING
AND
SERVICE
MANUAL
n
~~
G
BROOKDEAL
ELECTRONICS
~~
.:;;;;;~I:..
PRINCETON
APPL/ED
RESEARCH
Copyright
© 1983 EG&G PRINCETON APPLIED RESEARCH
Printed
In U.S.A.
SHOULD YOUR EQUIPMENT REQUIRE SERVICE
A. Contact the factory
(6091452-2111)oryour
local factory representative to discuss the problem. In many cases it will be possible to expedite servicing by localizing the problem to a particular plug-in
circuit
board.
B. If it is necessary to send any equipment back to the fac-
tory, we need the following information.
(1) Model number and serial number. (2) Your name (instrument user).
(3) Your address.
(4) Address to
which
instrument should be returned.
(5) Your telephone number and extension.
(6) Symptoms (in detail,
including
control
settings).
(7) Your purchase order number for repair charges (does
not apply to repairs in warranty).
(8) Shipping
instructions
(if you wish to authorize ship-
ment by any method
other
than normal surface
transportation).
C. U.S. CUSTOMERS-·-Ship the equipment being returned
to: EG&G PRINCETON APPLIED RESEARCH
7 Roszel Road (Off Alexander Road, East of Route 1) Princeton, New Jersey
D. CUSTOMERS OUTSIDE OF
U.S.A.-To
avoid delay in customs clearance of equipment being returned, please contact the factory or the nearest factory
distributor
for
complete shipping information.
E. Address correspondence to:
EG&G PRINCETON APPLIED RESEARCH P. O. Box 2565
Princeton, NJ 08540 Phone:
6091452-2111
TELEX: 84 3409
WARRANTY
EG&G PRINCETON APPLIED RESEARCH warrants each in­strument of
its
manufacture to be free from defects in material
and workmanship. Obligations under
this
Warranty shall be
limited
to replacing, repairing or giving credit for the purchase price, at our option, of any instrument returned, freight prepaid, to our factory
within
ONE year of delivery to the
original purchaser, provided
prior
authorization for such return
has been given by our authorized representative. This Warranty shall not apply to any instrument which our in-
spection shall
disclose
to our satisfaction, has become detec­tive or unworkable due to abuse, mishandling, misuse, acci­dent, alteration, negligence, improper installation or
other
causes beyond our control.
Instruments
manufactured by
others, and included in or supplied
with
our equipment, are not
covered by
this
Warranty but carry the original manufacturer's warranty which is extended to our customers and may be more restrictive. Certain subassemblies, accessories or
corn-
ponents may be specifically excluded from
this
Warranty, in
which case such exclusions are listed in the
Instruction
Manual supplied
with
each instrument.
We reserve the right to make changes in design at any time without
incurring any obligation to install same on
units
previously purchased. THERE ARE NO WARRANTIES WHICH EXTEND BEYOND THE
DESCRIPTION HEREIN. THIS WARRANTY IS IN LIEU OF, AND EXCLUDES ANY AND ALL OTHER WARRANTIES OR REPRE· SENTATIONS, EXPRESSED, IMPLIED OR STATUTORY, IN·
CLUDING MERCHANTABILITY AND FITNESS, AS WELL AS ANY AND ALL OTHER OBLIGATIONS OR LIABILITIES OF EG&G PRINCETON APPLIED RESEARCH, INCLUDING, BUT NOT LIMITED TO, SPECIAL OR CONSEQUENTIAL DAMAGES. NO PERSON, FIRM OR CORPORATION IS AUTHORIZED TO ASSUME FOR EG&G PRINCETON APPLIED RESEARCH ANY ADDITIONAL OBLIGATION OR LIABILITY NOT EXPRESSLY PROVIDED FOR HEREIN EXCEPT IN WRITING DULY EXE­CUTED BY AN OFFICER OF EG&G PRINCETON APPLIED RESEARCH.
Section
II
III
IV
V
TABLE
OF CONTENTS
CONDENSED
OPERATING
INSTRUCTIONS
CHARACTER
ISTICS
2.1
Introduction
2.2 Specifications
INITIAL
CHECKS
3.1
Introduction
3.2 Equipment Needed
3.3 Procedure
OPERATING
INSTRUCTIONS
4.1
Introduction
4.2 Preliminary Considerations
4.2A
Power Requirements
4.2B Fusing
.....
4.2C Warm-Up Period
4.20
Operating Frequency
4.2E
Grounding...
4.2F Noise
.....
4.3 Operating the Model 128A
4.3A
Introduction..
4.3B Reference Channel
4.3C Signal Channel
4.30
Output
Channel Controls
4.4 Mixer
Function
and Harmonic Sensitivity
4.5 Interface Connector .
4.6 Battery Operation .
4.7 Operation
with
the Internal Reference Oscillator
4.7A
Introduction
4.7B Operation . . . . . . . . . . .
4.7C Installation .
4.8 Operation
with
the Internal Tuned
Amplifier
4.8A
Introduction
4.8B
Operation.........
4.8C Installation .
4.9 More Reference Channel Operating Hints
4.9A
Reference Channel Slewing Rate
4.9B
Phase
Errors
with
Small Reference Signals
ALIGNMENT
.
5.1
Introduction
5.2 Required Equipment
5.3 Preliminary Steps
5.4 Procedure
5.4A
+15 V
Adjust
(R310),
-15
V Check, and +5 V Check
5.4B Reference Board Adjustments
5.4C Signal Board
Adjustment
5.40
Mixer Adjustments
5.4E Other Adjustments
Page
1-1
11-1 11-1 11-1
111-1 111-1 111-1 111-1
IV-l IV-l IV-l IV-l IV-l IV-l
IV-l IV-l IV-3 IV-5 IV-5 IV-5
IV-6 IV-8
IV-9 IV-10 IV-10 IV-l0 IV-10 IV-10 IV-ll IV-12 IV-12 IV-13 IV-15 IV-15 IV-15 IV-15
vi
V-l V-l V-l V-l V-l
V-l V-3 V-3 V-4
VI
VII
Number
1·1 IV·l IV-2 'V-3 IV-4 IV-5 IV·6 IV·7 IV-8 IV-9 IV-10
rvu
IV·12 IV-13 IV·14 IV·15
vi
Number
IV-l .V-2
TROUBLESHOOTING
6.1 Introduction
6.2 Equipment Required
6.3 Initial Steps . . . .
6.4 Power Supply Checks
6.5
Reference Checks
6.6 Signal Channel Amplifiers
6.6A Preamplifier
6.6B Intermediate AC Amplifiers
6.6C Final AC Amplifier
6.7 Mixer
6.8 DC Amplifiers .
SCHEMATICS, Table of
FIGURES
Model 128A Lock-In Amplifier . Ground-Loop Suppression by Ten-Ohm Input Ground Differential Measurement of "Single-Ended" Signal Differential Measurement of
"Off-Ground"
Signal
Example of Everything
"Done
Wrong"
Errors Depicted in Figure IV-4 Corrected
Typical Model
l28A
Noise Figure Contours
Pulse Train as Reference Drive
Net Phase
Shift
Between Signal and Reference Channels as a
Function
of Frequency Amplitude and Phase Characteristics of Hi-Pass Filter Amplitude and Phase Characteristics of Low-Pass Filter
Examples of
Output
Filter Interactions . . . . .
Mixer
Output
for In-Phase and Quadrature Signals
Internal Oscillator Board Installed . . . . . . .
Tuned Amplifier Installed .
Phase/Amplitude Characteristics of Tuned Amplifier Model 128A Adjustments and
Testpoints
.....
TABLES
Interface Connector Pin Assignments Frequency Range as a Function of Capacitors
ii
VI-' VI-' VI·' VI-' VI·' VI-' VI·2 VI-2 VI-2 VI-2 VI-2 VI-2
VII·'
Page
.
1-2
IV-l IV-'
IV-2 IV-2 IV·3 IV·3 IV·5 IV-6 IV-7
IV-7 IV-8
IV-9 IV·12 IV·13 IV·14
. V·2
Page
IV·l0 IV·l1
SAFETY CONSIDERATIONS
A. INTRODUCTION
The apparatus to
which
this
instruction
manual
applies has been supplied in a safe
condition.
This manual
contains
some
information
and warn-
ings that have to be
followed
by the user to ensure safe operation and to retain the apparatus in a safe condition. The described apparatus has been designed
for
indoor
use.
B. INSPECTION
Newly received apparatus should be
inspected
for
shipping damage. If any is noted,
immediately
notify EG&G PARC and file a claim
with
the car-
rier. The shipping
container
should be saved
for
possible
inspection
by the carrier.
WARNING! THE PROTECTIVE GROUNDING COULD BE RENDERED INEFFECTIVE IN DAMAGED AP­PARATUS.
DAMAGED
APPARATUS SHOULD NOT BE ·OPERATED UNTIL ITS SAFETY HAS BEEN VERIFIED BY QUALI­FIED SERVICE PERSONNEL. DAMAGED AP­PARATUS WAITING FOR SAFETY VERI­FICATION SHOULD BE TAGGED TO INDI­CATED TO A POTENTIAL USER THAT IT MAY BE UNSAFE AND THAT IT SHOULD NOT BE OPERATED.
C. SAFETY MECHANISM
As defined in IEC Publication 348 (Safety Require-
ments
for
Electronic
Measuring
Apparatus), the
Model 128A is Class I apparatus, that is, apparatus that depends on
connection
to a protective con-
ductor
to earth ground for
equipment
and operator
safety. Before any
other
connection
is made to the apparatus, the protective earth terminal shall be connected to a protective conductor. The protec­tive connection is made via the earth ground prong of the M128A's power cord plug. This plug shall only be inserted
into
a socket
outlet
provid­ed with the required earth ground contact. The protective action
must
not be negated by the use
of an extension cord
without
a protective conduc-
tor, or by use of an
"adapter"
that
doesn't
main-
tain earth ground
continuity,
or by any
other
means. The power cord plug provided is of the type illus-
trated in Figure 1. If
this
plug is
not
compatible
with the available power
SOCkets,
the plug or
power cord should be replaced
with
an approved
type of
compatible
design.
WARNING! IF IT IS NECESSARY TO REPLACE THE POWER CORD OR THE POWER CORD PLUG, THE REPLACEM ENT CORD OR PLUG MUST HAVE THE SAME POLARITY AS THE ORIGINAL. OTHERWISE A SAFETY HAZARD FROM
ELECTRICAL
SHOCK,
WHICH COULD RESULT IN PERSONAL INJURY OR DEATH, MIGHT RESULT.
L =
LINEORACTIVE
CONDUCTOR (ALSO CALLEU
"LlV£"OR"HOT")
N =
NEUTRALORIDENTIFIED
CONDUCTOH
E
=EARTH OR SAFETY GROUND
Figure 1. POWER CORD PLUG WITH POl.ARITY INDICATIONS
D. POWER VOLTAGE SELECTION AND
LINE FUSES
Before
plugging
in the
power
cord, make sure that
the
equipment
is set to the voltage of the ac
power
supply.
CAUTION! THE APPARATUS DESCRIBED IN THIS MANUAL
MAY BE DAMAGED IF IT IS SET FOR OPERATION FROM 110 V AC AND TURNED ON WITH 220 V AC APPLIED TO THE POWER INPUT CONNECTOR.
A detailed
discussion
of how to
check
and.tt
nec-
essary, change the power-voltage
setting
follows.
The line voltage is selected by means of a rear­panel
switch.
FOR SAFETY, UNPLUG THE POWER CORD WHEN CHECKING THE LINE VOLTAGE SETTING OR WHEN CHECKING THE FUSES. FUSES SHOULD ONLY BE CHANGED BY QUALIFIED SERVICE PERSONNEL WHO ARE AWARE OF THE HAZARDS INVOLVED. Depend­ing on the
switch
position,
either
"115"or"230"
(both are printed on the
swltchj-wl!lbevisible
to the viewer. For operation from a line voltage from 100 V ac to 130 V ac, 50-60 Hz,
"115"
should
show.
SECTION I
CONDENSED OPERATING INSTRUCTIONS
The
following
condensed
operating
instructions
are pro-
vided as an assistance in placing
the
Model
128A
Lock-In
Amplifier
into
operationasquickly
as possible.
Generally
speaking,
these
condensed
instructions
will
allow
good
re-
sults
to be
obtainedinmost
instances.
However,
because
of
the
brevity of
these
instructions,
many
considerations
hav-
ing
importanceinparticular
applications
have
been
fore-
STEP ONE
PRELIMINARY
STEPS;
Check
that
the
rear-panel
115/230Vswitch
is in
the
proper
position.
Remove
the
top
cover
and
verify
that
the
two
Reference
Lock-On
Speed
switches
are
properly
set.
For
operation
below5Hz,
they
shouldbesettoSLOW.
For
operations
above5Hz,
they
should
be
settoFAST.
(Units
are
shipped
with
these
switches
settoFAST.)
Replace
the
cover
and
pluginthe
line
cord.
Turn
the
power
on.
STEP TWO
REFERENCE
CHANNEL:
Check
that
the
Reference
Mode
switchissettothe
proper
position.
With
the
switch
setto"f",
the
detectorisdrivenatthe
frequencyofthe
applied
reference
signal.
Setto"2f",
the
detectorisdrivenattwice
the
frequency
of
the
applied
reference
to
facilitate
second
harmonic
measurements.
Next
connect
the
reference
signal
(100
m V pk -pkorgreater)
to
the
Reference
Input
connector
and
wait
for
the
REF.
UNLOCK
lighttogo
out
before
proceeding.
Usually
this
willbebut
a
few
seconds.
However,atlow
frequencies,
a
longer
timeisrequired.
STEP
THREE
SIGNAL
CHANNEL:Set
the
Input
Selector
to
"A"
(single-ended),
"A-B"
(differential)
or
"-B"
(single-ended),
whichever
is
appropraite.
Set
the
Sensitivity
switch
to
"250
mV".
If
signal
channel
filtering
is
desired,
set
the
HI-PASS
and
LO-PASS
Filter
switches
as
required.
Connect
the
signal
sourcetothe
Model
128A
Input.
1-1
gone.
For
this
reason,
it is advisabletoread
Section
IV,
the
complete
Operating
Instructions,tobe
assured
of achieving
optimum
performance.
NOTE:Aswritten,
these
condensed
instructionsdonot
applytounits
having
either
the
Internal
OscillatororTuned
Amplifier
modifications.Ifthe
unit
in
question
has
eitherorbothofthese
modifications,
the
operatorisreferredtoSubsections
4.7
and
4.8.
STEP
FOUR
OUTPUT
CONTROLS;
Set
the
Zero
Offset
toggle
switchtothe
center
(OFF)
position.
Then
set
theDCPrefilterto100msand
the
Time
Constant
switchto0.3
SEC.
STEP
FIVE
FINAL
ADJUSTMENTS:
Rotate
the
Sensitivity
switch
counterclockwise
until
the
panel
meter
beginstodeflect.
Adjust
the
Phase
Quadrant
switch
and
Phase
dial
for
maximum
meter
indication.
When
the
Sensitivity
and
Phase
adjustments
are
optimally
set,
the
meter
should
indicate
the
maximum
possible
on-scale
reading.Ifthe
Overload
light
comes
on,
increase
the
prefilter
and
time
constant
settings.
Also
try
"bracketing"
the
signal
frequency
with
the
HI
PASS
and
LO
PASS
filters.Ifthese
techniques
don't
help,
rotate
the
Sensitivity
switchasfar
clockwise
as is
required
to
eliminate
the
overload.
Set
the
time
constant
as
requiredtoreduce
the
output
noisetoan
acceptable
level.
STEP SIX
READING:
The
input
signal
levelisread
from
the
panel
meter.Ifthe
input
signal
is a
sinewave,
the
meter
directly
indicates
the
rms
amplitudeofthe
input
signal.
However,
if
the
signal
frequencyisnear
enoughtoa
selectedHIorLOPASS
filter
frequency
to
be
influenced
by
the
filter,
the
filter
attenuation
effects
must
be
taken
into
account.
If
the
input
signalisnot
a
sinewave,
then
the
harmonic
responseofthe
instrument
mustbetaken
into
account
as
well.
1-2
a:
w
u.
-'
Q.
~
~
~
c
0
t
-'
~
co
N
...
-'
W
0
i
...
e
::l
'"
u::
SECTION
II
CHARACTERISTICS
2.1 INTRODUCTION
The
Model
128A
Lock-In
Amplifier
enables
the
accurate
measurementofsignals
contaminatedbybroad-band
noise,
power
line
pickup,
frequency
drift,orother
sources
of
interference. It
does
thisbymeans
of an
extremely
narrow
band
detector
which
has
the
centerofits
passband
locked
to
the
frequencyofthe
signaltobe
measured.
Because
of
the
frequency
lock
and
narrow
bandwidth,
large
improvements
in signal-to-noise
ratio
can
be achieved,
allowing
the
signalofinteresttobe
accurately
measured,
even in
situations
where
it is
completely
masked
by noise.
Signals applied
to
the
input
(single-ended or
differential)
1
routed
through
a series of amplifiers
which
allow
..rll-scale sensitivity ranges
downtoone
microvolt.
Switch
selectable low-pass
and
high-pass filters allow
considerable
noise
reduction
aheadofthe
phase-sensitive
detector.
This
pre-detector
noise
reduction
canbefurther
enhanced
by
making
use of
the
optional
plug-in (internal) selective
amplifier. At
the
phase
sensitive
detector,
the
signal is
compared
with
the
reference
signal derived
from
the
experiment.
Only
those
signal
components
which
are
synchronous
with
the
reference
yieldanetdcdetector
output.
Noise
and
other
non-synchronous
signalsdonot
contributeanetdcoutput,
but
onlyacfluctuations
which
canbereducedtoany
arbitrary
value
accordingtothe
amount
of filtering
selected
with
the
Time
Constant
switch.
This switch allows
time
constants
as large as
100
seconds
to
be
selected,
with
provision for achieving larger
externally
determined
time
constants
if necessary.
Post-detector
dc
amplifiers drive
the
panel
meter
and
signal
output
connectors.
Other
features
include
provision
for
calibrated
zero
suppression of up to10x full scale, a
two-position
dc
prefilter,
and
the
capability
of driving
the
reference
input
the
detector
at
double
the
frequency
of
the
signal
applied to
the
Reference
Input
connector
to
facilitate
second
harmonic
measurements.
An
optional
plug-in
oscillator
(internal)
is available for use in
applications
where
the
experiment
does
not
produceareference
signal itself,
but
is capableofbeing
driven
by a signal
furnishedbythe
Lock-In Amplifier.
With its wide range of capabilities
and
easeofoperation,
the
Model
128A
Lock-In
Amplifier
should
find
extensive
applicationinsituations
where
the
accurate
measurement
of signals is
complicated
by
the
presenceofnoise and
interference.
2.2 SPECIFICATIONS
SIGNAL
CHANNEL
(1) INPUT
TYPE
Single-endedordifferentialasselectedbyfront-panel switch.
11-'
(2)
INPUT
IMPEDANCE
100
Mr2
shuntedbyno
more
than20pF.
(3)
SENSITIVITY
12 full-scale ranges in
1-2.5-10
sequence
from1J.1V
to
250
mV.
(4) FREClUENCY
RANGE
0.5Hzto
100
kHz.
(5) COMMON MODE
REJECTION
At least
100dBat
1 kHz.
(6) MAXIMUM COMMON MODE
VOLTAGE
3 V
pk-pkto20
kHz;
then
--6
dB/octave
above
20
kHz.
(7)
DETECTOR
BIAS
Internal
network
allows dc bias
current
of
either
polaritytobe
providedatthe
"A"
Inputtofacilitate
operation
with
diode
detectors
which
require
biasing.
(See page VII-3
and
Parts
Location
Diagramonpage
VII-2.)
(8) NOISE
At 1 kHz
the
signal
channel
noise will
not
exceed
10
nV/H
zY2.
(9) LOW PASS
FILTER
Switch
selectable6dB/octave
low-pass filter
which
can be
set
to 3 dB
down
frequenciesof100
Hz,
10
kHz, or MAX
(greater
than
100
kHz).
(10)
HIGH PASS
FILTER
Switch
selectable6dB/octave
high-pass filter
which
can be
setto3 dB
down
frequenciesof50
Hz, 5 Hz,
or MIN
(below
0.5
Hz).
(11)
OVERLOAD
DETECT
Front-panel
indicator
lights if
applied
signal plus noise
is large
enoughtocause
overloadatany
of several
critical
overload
monitor
points.
(12)
GAIN
STABILITY
(13) GAIN
LINEARITY
0.05%.
(14)
OVERALL
GAIN ACCURACY
±2%.
REFERENCE
CHANNEL
The
Model
128A
reference channel
automatically
locks
onto
and
tracksanapplied
reference signal over the
entire
operating
frequency
range of
the
instrument.
As a result,
the
instrumentisimmunetofrequency
and phase
shifts
as
long as
the
reference and signal to be recovered change
together.
(1)
TRACKING
RANGE
5 Hz to
100
kHz (FAST) or
0.5
Hz to
100
kHz
(SLOW) as
determinedbythe
settingoftwo
internal
switches.
Faster
lock-on
time
and slewing
rate
obtained
with
switches
settoFAST
make
this range
preferable
except
when
operating
below 5 Hz.
(2) MODES
Eitheroftwo
modes, f and 2f, can be selected by
means of a front-panel switch. In
the
"f"
position,
the
phase-sensitive
detector
is driven at
the
same
frequencyasthe
applied reference signal. In
the
"2f"
position,
the
phase-sensitive
detector
is driven at twice
the
frequency
of
the applied reference signal
to
facilitate
second
harmonic
measurements.
(3) INPUT IMPEDANCE
10
Mrl
shuntedbyno
more
than20pF.
(4) MINIMUM
REFERENCE
SIGNAL
REQUIREMENT
100
mV pk-pk,
any
waveshape crossing its
mean
only
twice each cycle. Minimum
time
required on
either
sideofthe
meanis100
ns.
Amplitude
excursions
must
be at least50mV on each sideofthe
mean.
Maximum
input
signal is 5 V (pk-to-rnean}. Best phase
accuracy
is
obtained
with
a 1 V rms sinewave.
(5) LOCK-ON TIME
A
function
of internal
switch
setting
as follows.
Selected
Range Lock-On
Time
SLOW (0.5 Hz to
100
kHz)
20
sec. per
octave
FAST (5 Hz to
100
kHz) 2 sec. per octave
(6) PHASE
Calibrated Phase
controls
allow the phase of
the
reference drive to
the
Phase-Sensitive
Detector
to be
set at any angle relative to
the
input
signal.
The
controls
consist of a Phase Dial with a range of
100°
and a Phase
Quadrant
switch
which
provides
incremental phase shifts of
90°.
The
phase
shift
11-2
accuracy of
the
dial is
better
than
0.2
0
over the
entire
frequency
range.
The
resolutionofthe
dial is
better
than
0.10.The
incremental
phase shiffs provided by
the
Quadrant
switch are
accurateto0.2°.
The
overall
phase accuracy of
the
instrument,
including
shifts
in
both
the reference and signal channels, is typically
better
than
50.
(7)
DETECTOR
BIAS
Internal
network
allows de bias
currentofeither
polarity
to be provided at
the
REF.INconnector
to
facilitate
operationinsituations
where
the
reference
signal is
taken
from
diodes requiring biasing. (See page
VII-6 and Parts Location Diagram
on
page VII-5.)
PHASE
SENSITIVE
DETECTOR,DCAMPLIFIER
(1)
OUTPUT
DRIFT
(2)
OVERLOAD
CAPABILITY
1000
times full scale up to a
maximumatthe
input
of
650
mV rms. Overload capability is
definedasthe
ratio, at
the
input
of the Model
128A,
of
the
maximum
pk-pk
non-coherent
signal
which
can be
applied
without
overloading the Model
128Atothe
pk-pk
coherent
signal required to yield full scale
Model
128A
output.
Note
that,
expressed as
the
ratio
of
the
pk-pk
non-coherent
signaltothe
rms value of
the
coherent
signal
required
for full-scale
output,
this
number
can be as
greatas2800.
Maximum
acceptable
signal is a
650
mV rms sinewave.
(3)
NON-COHERENT
REJECTION
50
ppm
maximum.
Non-coherent
rejection is
defined
as
that
offset
which
results
from
applying a
non-coherent
signal having a
pk-pk
amplitude
1000
times
the
pk-pk
amplitude
of
the
coherent
signal
requiredtoobtain
full-scale
output.
Example: With a
non-coherent
signal applied having a
pk-pk
amplitude
1000
times
the
pk-pk
coherent
signal
required
to
obtain
full-scale
output,
there
will
occur
an
offsetatthe
output
caused by
the
non-coherent
input
signal.
The
amplitude
of this
offset
will be no
greater
than:
50 X
10-
6
X
1000
= 50 X
10-3of f.s.
output
=50 mV (f.s. =1 V)
(4) TIME
CONSTANT
Front-panel switch allows selection of 6
dB/octave
filter
time
constants
of 1 ms, 10 ms, and .1 s
through
100
s in 1-3-10 sequence. Also MIN (time
constant
:::=0.7
ms) and EXT, which allows
time
constants
longer
than
100
s to be achieved by means of
external
capacitors. A
separate
front-panel
toggle switch allows
another6dB/octave
filter to be
inserted,
if desired.
This filter has a
time
constantofeither
100
ms or 1 s,
whichever is selected.
(5)
ZERO
OFFSET
A
calibrated,
ten-turn,
Zero
Offset
dial,
with
up to
ten
times full-scale
capability
is provided.
(6) FULL-SCALE
OUTPUT
±1 V.
(7)
OUTPUTS
(a) Panel
meter,
± full scale.
(b)
Front-panel
BNC
connector.
One volt
out
corresponds
to full-scale panel
meter
deflection.
Output
resistance is
600
ohms.
(c) Rear-panel
Recorder
Out
binding posts,
spaced
to
accept
standard
double-banana
connector.
Output
resistance is
600
ohms.
GENERAL
(1) AMBIENT
OPERATING
TEMPERATURE
RANGE
(2)
AUXILIARY
POWER
OUTPUT
±15.5
V regulated dc
at
20 mA is
provided
at
rear-panel
connector.
(3) POWER
REQUIREMENTS
100-130
V ac
or
200-260
V ac,
50-60
Hz. May also be
powered
from
±24 V de
source
(such as
batteries).
Power
consumption:15watts.
(4) SIZE
17-3/4"
W x
3-1/2"
H x
14"D(45cmW x 9 cm H x
36 cm D).
11-3
(5)
WEIGHT
14 Ibs
(6.4
kg).
(6)
MODIFICATIONS
(a) Model
128A/97
Monitor
Modification
Three
rear-panel BNC
connectors
are installed
which
permit
monitoring
of:
(1)
OutputofSignal
Channel
before
demodulation
(please
note
that
the
Signal Channel
monitor
is provided as
part
of
the
Tuned
Amplifier
modification
as well), (2)
Squarewave
output
of Reference
Channel,
and
(3) Full-wave
demodulated
Mixer
output
before
Time
Constant
filter.
(b) Model
128A/98
Tuned
Amplifier Modification
Internal plug-in
board
is available
which
provides
a
tuned
bandpassornotch
characteristic
at a
Q-of-5.
The
frequency
can be
adjusted
over a 3: 1
range by
meansofa rear-panel
adjustment,
and
can be
settoany
frequency
from
1 Hz to
100
kHz by changing
capacitors
mounted
on
component
clips
located
on
the plug-in
circuit
board.
(c) Model
128A/99Internal
Oscillator
Modification
An internal low
distortion
oscillator
is available
which
provides a sinewave
output
adjustable
from
0-to-l0
V pk-pk
at
600
ohms.
The
frequencyisadjustable
over
about
a 3: 1 range by
means of a rear-panel
adjustment,
with
the
actual
frequency
range
spannedbythe
adjustment
being
determined
by a pair of
internal
capacitors
mounted
on
the
oscillator
circuit
board.
Operation
from
about
1 Hz
to
100
kHz is
possible. This
option,inconjunction
with
the
2f
mode
of
operation,
is
particularly
useful for
harmonic
detection
where
the
modulation
frequency
must
be at
one
half
the
detected
frequency.
SECTION
III
INITIAL
CHECKS
3.1 INTRODUCTION
The
following
procedureisprovided
to facilitate initial
performance
checkingofthe
Model 128A. In general, this
procedure
should
be
performed
after
inspecting
the
instrument
for
shipping
damage
(any
notedtobe
reported
to
the
carrier
and
to
Princeton
Applied
Research
Corporation),
but
before
using
the
instrument
for
experimental
measurements.
Should
any
difficulty
be
encountered
in carrying
out
these checks,
contact
the
factoryoroneofits representatives. It
mightbenoted
that
it is
not
the
purpose
of these
checkstodemonstrate
that
the
instrument
meets
all specifications,
but
rather
simply
to
show
that
it is
functioning
normally.
If normal
indications
e
obtained
for
the
functions
checked,
one
may
reasonably
assume
that
those
functions
which
are
not
checked
are
working
properly
as well.
3.2 EQUIPMENT NEEDED
(1) Sinewave Oscillator
to
providea100
mV rms sinewave
at 1 kHz.
NOTE:Ifthe
instrumenttobe
checked
is
equipped
with
the
Tuned
Amplifier
modification,
then
the
oscillator
will have
to
provide a
100
mV rms
sinewave at
the
tuned
frequency.Ifthe
instrument
in
questionisequipped
with
the
Internal Oscillator
modification,noexternal
oscillator will be
required.
Th is applies
whetherornot
the
unitisequipped
with
the
Tuned
Amplifier
modification.
(2) General
purpose
oscilloscope. This
itemisrequired
only in
the
caseofunits
having
the
Tuned
Amplifier
modification.
'3) Suitable cables
for
interconnecting
the
above instru-
ments.
3.3 PROCEDURE
(1) Check
the
positionofthe
rear-panel
115/230
switch.
Be sure
the
number
showinginthe
window
corre-
spondstothe
line voltagetobe used.
(2) With the Power switch
settoOFF,
plug in
the
line
cord.
(3)
Set
the
Model
128A
controls
as follows.
Input
Selector: A
Sensitivity:
100
mV
Filters
HI-PASS: MIN. LO-PASS: MAX.
Phase
Quadrant
switch:
270°
Phase dial:
90° Mode: f Zero
Offset
switch:
OFF
(center
position)
111·1
dial:
1.00
(one
turn
from
fully
counterclockwise
position)
Time
Constant:
.3 SEC.
DC Prefilter:
OUT
Reference
Tracking-Rate
switches
(two
internal
switches);
FAST
(unless
unitisequipped
with
Tuned
Amplifier'
or Internal Oscillator set for
frequency
below
5 Hz, in
which
case
switches
shouldbeset
to SLOW). NOTE:
Instruments
are
normally
shipped
with
these
switches set
to
FAST.
Power.
ON
(4) Applies
onlytounits
not
equipped
with
the
Internal
Oscillator
modification.Ifthe
Model
128A
has this
modification,goto
step
5.
Connect
the
outputofthe
external
oscillator
(set for
100
mV rms sinewave
out
at 1 kHz)toboth
the
"A"
and
Reference
Inputs. If
the
unitinquestion
has
the
Tuned
Amplifier
modification,
set
the
oscillatortothe
tuned
frequency
specified
when
the
unit
was
ordered.
(5) Applies
onlytounits
having
the
Internal Oscillator
modification.
Connectacable
from
the
rear-panel
REF.
OSC.
OUT
connectortothe
front-panel
"A"
Input.
There
is no
needtoconnect
th is signal to
the
Reference
Input
connector
(labeled MONITOR in
units
equipped
with
the
Oscillator
modification).
The
connectiontothe
Reference
Channel of
the
Model
128Aismade
internallyatthe
factory.
The
rear-panel
REF.
AM-
PLITUDE
adjustment
should
be
set
so
that
the
amplitude
of
the
signal at
the
REF.
OSC.
OUT
connectoris100
mV rms. An
accurateacvoltmeter
may be useful
for
setting
this
level (units leave the
factory
set
foranominal
100
mV rrns
out).
(6) Applies
onlytounits
having
the
Tuned
Amplifier
modification.
In
the
case
of
units
not
having
this
modification,
go d irectl ytostep
7.
(a)
Connect
the
oscilloscope to
the
rear-panel SIG.
MON.
connector.
(b)
Vary
the
frequencyofthe
signal applied to
the
"A"
Input
(use
the
rear-panel
REF.
OSC.
FREQ.
ADJ.
controlinunits
equipped
with
an Internal
Oscillator)
for
peak
signal
amplitude
as observed
with
the
oscilloscope.
(7)
Set
the
Phase
Quadrant
switchto0°.
Then
adjust
the
Phase dial
for
"0"
panel
meter
indication.
(8)
Set
the
Phase
Quadrant
switch
backto270°.
The
panel
meter
should
indicate
full scale to
the
right ± a
few
percent
of full scale.
The
accuracyofth is readinq
will
depend
on the
amplitude
accuracy of the signal
applied
to
the
"A"
Input.
(9)
Adjust
the
amplitudeofthe
signal applied to
the
"A"
Input
as required to
obtain
exactly
full-scale panel
meter
indication.
(10)
Set
the
Phase
Quadrant
switch to
180°.
The
panel
meter
should
indicate
"0"
±5%offull scale.
(11)
Set
the
Phase
Quadrant
switchto90°.
The
panel
meter
should
indicate
negative full scale ±5% of full
scale.
(12)
Set
the
Phase
Quadrant
switch to
270°
to restore
the
111-2
positive full-scale panel
meter
indication.
(13)
Set
the
Zero
Offset
toggle switch to "+".
The
panel
meter
indication
should
go to
"0"
±5% of full scale.
(14) Begin
rotating
the
Offset
dial
counterclockwise.
The
panel
meter
indication
should
increase linearly, track-
ing
the
dial setting. When
the
dial is fully
counter-
clockwise,
the
panel
meter
should
indicate full scale
±5%. Reset
the
Zero
Offset toggle switch to
the
center
(OFF)
position.
This
completes
the
initial checks. If
the
indicated results
were
obtained,
one
can be reasonably sure
that
the Model
128Aisfunctioning
normally.
SECTION
IV
OPERATING
INSTRUCTIONS
4.1 INTRODUCTION
Even
though
operation
of
the
Model
128Aisstraight-
forward,
there
are a
number
of
factorsofwhich
the
operator
should
be aware
to
be assured of achieving
optimum
performance
in all
situations.
This
sectionofthe
manual
treats
these
considerationsinsome
detail.
Topics
covered include
grounding,
noise
performance,
harmonic
sensitivity,
operation
in
conjunction
with
the
plug-in
accessories,
and
others.
For
an overall
"quick
look"athow
the
instrumentisoperated,
the
operatorisreferred
to
SectionI,the
condensed
operating
instructions.
4.2
PRELIMINARY
CONSIDERATIONS
4.2A
POWER
REQUIREMENTS
The
Model
128A
requires
100-130
V acor200-260
V ac,
50-60
Hz.
The
power
consumptionis15
watts.
The
unit
may also be
poweredbybatteriesbyapplying
±24 Vtothe
appropriate
pinsofthe
rear-panel octal-
connector
(see
BATTERY
OPERATION,
page
IV-l0).
A rear-panel slide
switch
determines
whether
the
ac
power
circuits
are
connected
for
operation
from
100-130
V or
from
200·260V.For
operation
from
100-130V,"115"
should
showinthe
window.
For
operation
from
200-260V,"230"
should
show.
4.28
FUSING
The
Model
128Aisprotected
by a single fuse
mounted
on
its rear panel. A
slow-blow
1/4
A fuse is used for
operation
from
115
V. A slow-blow
1/10
A fuse is used for
operation
from
230
V.Itoccasionally
happens
thataslow-blow
fuse
fails in
shipment
as a
resultofshock
and
vibration.
Hence,
if
the
fuse is
found
to be bad
when
the
instrument
is
operated
for
the
first
time
it is advisable to
try
and
change
the
fuse. If
normal
operation
follows,
chances
are
there
are
no
other
problems.
However, if
the
replacement
fuse fails,
thereissomething
wrong
which
will have to be
corrected
before
proceeding.
4.2C
WARM-UP
PERIOD
For
most
applications,
five
minutes.
Where it is
desired
to
achieve
the
best
possible gain
and
output
stability,
allow an
hour.
4.20
OPERATING
FREQUENCY
Although
one
can,
in principle,
make
equally
accurate
measurements
at any
frequency
within
the
operating
range
of
the
instrument,
operationissimplest
and
least
subject
to
error
over a range having as its
lower
limit,
perhaps
a few
hundred
Hz,
and
as its
upper,
perhaps
10k
Hz.Atvery low
frequencies,
phase
offsets
occur
which
could
matterifone
is
interestedinthe
absolute
phase of
the
input
signal.
Another
problemoflow
frequency
operationisthatofl/f
noise, including
both
that
which
developsinthe
Model
128A
and
that
which
originatesinthe
experiment
itself
to
degrade
the
signal-to-noise
ratio
ahead
of
the
lock-in
amplifier. Increased
response
and
settling
time
could
be a
IV-1
significant
problemifone
were
operatinginconjunction
with
the
optional
plug-in
tuned
amplifier.
At high fre-
quencies,
radiation
and
associated
pick-up
tend
to
be
bothersome.
Another
high
frequency
problemisthat
of
signal
attenuation
as a resultofthe
input
cable
capacitance.
Th is is especially a
problem
when
working
from
a high
source
resistance.
Other
frequencies
to avoid are60Hz
and
its
lower
order
harmonics.
By avoiding
these
frequencies,
the
operator
assures
that
he will be
measuring
the
signal
of
interest
only,
uninfluenced
by
power
frequency
pick-up,
either
internalorexternal.
4.2E
GROUNDING
In
any
system
processing low-level signals,
proper
grounding
to
minimize
the
effectsofground
loop
currents,
usually
at
the
power
frequency,
is an
important
consideration.Inthe
case of
the
Model
128A,
special design
techniques
have
been
employedtogive a high
degreeofground-loop
signal
rejectioninsingle-ended
applications.
Even so, it will
often
prove advisable to
operate
differentiallv.
even
when
ex-
amining
a single-ended signal
source,
10 achieve
the
greatest
possible
rejection.
Figures IV-l
and
IV-2
illustrate
this
point.
Note
from
Figure IV-l
that
the
signal
source
is
located
inside a
grounded
enclosure
(shield), to
which
signal
source
commonisattachedatone
point.
The
braid of
the
/
'U
,
INDUCED emf
(DISTRIBUTED AROUND
LOOP)
RESISTANCE OF AC GROUND
PATH--J
Figure
IV·l.
GROUND·LOOP
SUPPRESSION
BY
TEN-OHM
INPUT
GROUND
SHIELD
l_
10"
L
INDUCE~.mf
'"
(CJSTRIBUTED AROUND LOOP) J
RESISTANCE OF AC GROUND PATH
Figure
IV-2.
DIFFERENTIAL
MEASUREMENT
OF
"SINGLE-ENDED"
SIGNAL
Figure
IV-4.
EXAMPLEOFEVERYTHING
"DONE
WRONG"
The
reductionofpower
frequency
interferenceisnot
the
only
benefit
to be
derived
from
proper
grounding
and
differential
operation.
A
much
more
serious
source
of
interferenceiscoherent
interferenceatthe
signal
frequency
which
results
when
drive signal
currentisallowedtoflow
through
the
braidofthe
signal
cable.
Figures IV-4
and
IV-5
are
providedtoillustrate
this
problem
and
the
steps
which
can
be
takentoprevent
it.
To
begin
with,
Figure
IVA
shows
the
experiment
with
just
about
everything
possible
"done
wrong".
The
lock-in
amplifierisoperated
single-
ended.
The
ground
connectionsatthe
experiment
are
made
to
the
enclosure,
allowing
currentstoflow
through
it,
and,
in
particular,
the
drive signal
currents
have
the
opportunity
to
flow th rough
the
braidofthe
signal cable.
The
drive
signal, in
additiontoproviding
the
reference
input
signal
to
the
Model
l28A,
can be
presumedtobe driving
other
componentsofthe
system
as well.
Dependingonthe
nature
of
the
experiment,
these
currents
could
range
from
very
small to
quite
large,
perhaps
even
amperesifthe
experiment
involves driving a
low
impedance
coil.
Note
that
the
various
loads
for
the
drive
are
represented
by a single
resistor
returnedtoground
somewhere
on
the
enclosure.
Most-of
this
drive
signal
current
canbepresumedtoflow
through
the
shield
backtothe
drive
signal
source.
However,
a small
but
significant
part
of it will
flow
through
the
parallel
path
consistingofthe
braidofthe
signal cable,
the ten
ohm
resistor,
and
the
braidofthe
reference
signal
cable.
The
voltage
dropofthis
current
across
the
resistanceofthe
signal
cable
brai-J,
even
though
attenuated
by
the
ratio
of
the
ten
ohm
resistortothe
braid
resistance, can
constitute
a
serious
sourceofinterference
at low signal levels,
particular-
ly in
that
this
interferenceiscoherent,
in
phase,
and
directly
addstothe
signal of
interest.
It is
not
hard
to
envision
situations
where
this
interference
signal
could
well
be larger at
the
Inputofthe
lock-in
amplifier
than
the
signal
of
interest
itself.
signal
cableisgrounded
directly
to signal
source
common
as
well,
thereby
assuring
that
no signal
currentsorground-
loop
currents
will
flow
through
the
shield,'adesirable
condition
for
the
best
possible
shielding.
The
Model
l28A
is
operated
single-ended,
using
the
"A"
input.
Note
that
the
"low"
side of
the
amplifier
inputisnot
groundedtothe
chassis
directly
butbywayofa
ten
ohm
resistor.
Further
note
that
the
braidofthe
signal
cableisreturnedtothis
resistor,
and
nottothe chassis. A
ground
loop
generator
is
indicatedasbeing
connected
between
the
chassis of
the
Model
l28A
and
siqnal
source
common.
This
path
would
ordinarily
consistoftheacground
"third
wire",
paralleled
by
the
braids
of
other
cables
connecting
the
system
components.
The
ground
loop
generator
will
cause
currents
at
the
power
frequencytoflow
through
the
braid of
the
signal cable,
through
the
ten
ohm
resistor,
and
back
through
theacwound
pathtocomplete
the
loop.
Because
of
the
ten
ohm
ground
employedinthe
Model
l28A,
these
currents
are
attenuated
over
what
they
would
be if
the
Model
128A
input
were
returned
directlytothe
chassis.
More
importantly,
most
of
the
ground
loop
signal is
dropped
across
the
ten
ohm
resistor
and
little
across
the
braid of
the
signal
cable,
the
ratio
being
the
ten
ohmsofthe
resistor to
the
10 to
20
milliohms
(typical)ofthe
braid
resistance, As far as
the
input
of
the
Model
l28A
is
concerned,
the
ground
loop
signal is
reduced
by this
ratio,
and
the
ground
loop
interferenceisthus
perhapsafactor
of
five
hundredorone
thousand
less
than
wouldbethe
case
without
the
ten
ohmqround.
However, in
some
appl
ications, th is
would
notbeenough.
Figure IV-2
shows
how
this
same
signal
couldbemeasured
operating
the
Model
128A
differentially.Inthis
instance,
the
Model
l28A
Input
Selectorissetto"A-B"
and
two
input
cables
are used,
one
connectedtothe
signal
source
and
the
other
to siqnal
source
common.Atthe
source
end,
the
braid
of
both
cables is
returned
to
signal
source
common.Atthe
lock-in
amplifier
end,
the
ten
ohm
ground
serves to
attenuate
the
wound
loop
currents
and
maintain
a
small
ground
loop
signa!
drop
across
the
braids
the
same
as
in Figure
IV-l.
However,inthe
first
instance,
the
amplifier
"looked"
at
the
potential
difference
between
the
center
conductorofthe
cable
and
the
braid.Inthe
second,
it sees
the
potential
difference
between
the
"A"
Input
and
"8"
Input.
The
ground
loop
siqnal
current
flowinginthe
signal
cable
braid is of no
consequence.
The
very high
common
mode
rejectionofthe
amplifier
assures
that
common
mode
power
frequency
pickup
will
not
be a
problem
either.
However,
when
operating
differentially,
it is
important
to
takealittle
trouble
to assure
that
common
mode
inter-
ference arising in
ground
loopsisjust
that,
that
is,
without
a
significant
differential
component.
This
should
not
prove
a
problemaslonqasboth
signal
cables
follow
the
same
path,
Figure IV-3
shows
the Model
128A
operated
differentially
to
measurean"off
~JI
ound
" siqn al.
The
most
important
consideration
in an
application
of this
type
is to be sure
that
the
common
mode
siqnal
componentisnot
so large as
to
exceed
the
common
mode
input
limitofthe Model
128A.
(See Specs.)
IV-2
DRIVE
~
~:£"
0'"''''''
,.0'"
-1
AS
FUNCTIONOFEXPERIMENTAL
PARAMETEI
L..-
...I
Figure
IV-3.
DIFFERENTIAL
MEASUREMENT
OF
"0
FF-GROUND"
SIGNAL
DRIVE
LOAD
k =
Boltzmann's
constant=1.38x10-
2 3
joules/kelvin
T =
absolute
temperature
in kelvins
R
s
= Resistance in
ohmsofthe
resistive
component
of
the
impedance
across
which
the
voltage
iii
mea-
sured
B =
Bandwidth
over
which
the
measurementismade
Mathematically
expressed,
noise figure
canbestated
as:
NF(dB)
=
2010
9 10 Noise
V.olta~_~~g~!':l_~t
of
Amplifier
That
PortionofNumerator
Attributable
IV-2
to
Source
Thermal
Noise
Figure IV-6.
TYPICAL
MODEL
128A
NOISE
FIGURE
CONTOURS
~~
where
Etis
the
total
noise
referredtothe
input
in volts rrns
and
all
other
terms
are as
defined
previously.
Note
thar
IV-3
10~
Et=V4kTBRX10NF/20
o
11>
N
~
Ul
2
:r
o
~
""
o
z
c(
Ii;
~
0:
""
U
~
~
102::-:~__~~---9:::'-'-_--.-l.-
__
.L,
__
--.J
0.5
10
CENTER fR
Noise figure is
not
constant
but
varies as a
functionofthe
source
resistance,
frequency,
and
temperature.
When
the
lociofall
points
having
the
same
noise figure are
plotted
as
a
functionoffrequency
and
source
resistance
(temperature
fixed),
the
result
is a noise figure
contour.
A full
set
of
contours
completely
specifies
the
noise
characteristics
cf
the
antpti#er
ruler its
working
range. Figure IV-6
contains
a
full
setofcontours
for a
typical
Model
128A.
The
utility
of
these
contours
are, first of all,
that
they
clearly
indicate
the
best
noise
performance
region in
terms
of
operating
frequency
and
source
resistance,
and
secondly,
that
they
allow
onetodirectly
compute
the
total
noise
accompany-
ing
the
signal
(amplifier
noise
and
source
thermal
noise
considered,
other
noise
sources
neglected).
.The
relating
formula
is:
IV-l
4.2F NOISE
Any
electronic
signal processing
system
adds
noise to
that
already
accompanying
the
signal to be
measured,
and
a
Lock-In
Amplifier
is no
exception.
Even
though
the
method
of signal processing
used
in a Lock-In
Amplifier
allows very large
improvements
in signal-to-noise
ratio
to be
achieved,
the
amount
of noise
contributedbythe
Lock-In
Amplifier itself
affects
its
performance
and
limits
the
achievable
improvement.
Figure IV-5
shows
the
steps
which
can be
taken
to
circumvent
this
problem.
Allofthe
drive signal
current
is
returned
directlytothe
drive signal
source
except
for
the
very small
component
(reference
input
resistance of Model
128A
is 10
MS1)
whichisappliedtothe
Model
128A
by
way of
the
Reference
Input.
Second,nocurrent,
whether
drive
current,
reference
current,
or signal
current,isallowed
to
flow
through
the
experiment
shield;
the
shield
contacts
groundatone
point
only.
The
only
coherent
signal
which
can
flow
through
the
parallel
pathofthe
signal cable braid
is a small
portionofthat
allowed by
the
ten
megohm
Reference
Input
resistance.
Furthermore,
the
use of
differ-
ential
operati
on assures
that
even th is small
amountcan
have no
effect.
By using
the
arrangements
indicated,
one
could
operate
with
very large drive
currents
without
concern
that
they
might
contaminate
the
signalofinterest.
If
electrostatic
couplingofthe
drive signaltothe
detector
is
a
problem,
mountingaconducting
material
around
the
signal
source
detector
should
prove
helpful.
The
electro-
static
shield
shouldbeconnectedtothe
systematbut
one
point,
signal
source
common.
Figure IV-5.
ERRORS
DEPICTEDINFIGURE
IV-4
CORRECTED
One
convenient
wayofspecifying
the
noise
performance
of
an
amplifieristo
speakofits
noise
figure,
which
indicates
the
amount
of noise
the
amplifier
adds
to
the
source
thermal
noise.
Source
thermal
noise is used as
the
basis for
comparison
because
it is
completely
predictable,
always
present,
andisthe
least
amount
of noise
which
can
possibly
accompany
any
signal. Its value, in volts rms, is given by
the
following
formula.
where:
En = rms noise voltage
within
the
bandwidthofthe
measurement
IV·3
with
a noise figure of 3 dB,
the
amount
of
noise
contributedbythe
amplifier
is 1.4
times
the
source
thermal
noise.At1.4
times
the
thermal
noise,
the
amplifier
noise
just
beginstobe
noticeable.
At
lower
noise
figures,
the
amplifier
for
all
practical
purposes
mayberegarded
as
noiseless.
Generally
speaking,ifone
can
operate
anywhere
inside
the
3 dB
contour,
amplifier
noise
considerations
may
be
neglected.
As critical as
amplifier
noise is in
certain
applications,
it is
nevertheless
possibletooveremphasize
its general
impor-
tance.
For
example,ifthe
signal
amplitudeissignificantly
higher
than
the
amplifier
noise,
the
subject
becomes
purely
academic.
Similarly,ifpreamplificationisprovided
ahead
of
the
Model
128A,
with
the
result
that
the
amplified
source
noise at
the
inputtothe
Model
128A
is far
greater
than
the
amplifier
noise,
thereislittle
point
in striving to
operate
inside
the
3 dB
contour
of
the
Model
128A.
However,
whereapreamplifierisused,
it is
important
that
these
same
considerations
be
carefully
evaluated
for
the
preamplifier.
In
other
words,
when
using a
preamplifier,
try
to
operate
inside
the
3 dB
contourofthe
preamplifier.
A
quick
check
of
Fiqure iV-G
followed
by a
computation
of
the
total
noise
(Equation
IV-3)
should
give
one
a realistic
idea of the
importanceofamplifier
noise
considerations
to
the
measurementathand.
Where
amplifier
noise is a
consideration,
one
should
try
to
operate
inside
the
3 dB
contourbyappropriately
adjusting
the
operating
frequencyorsource
resistance.
The
choice
of
operating
frequencyisusually
determined
by
the
type
of
sensor
used
andbythe
capabilitiesofthe
chopper,
where
oneisused.
Often,
the ex
perimenter
has
some
control
over
his
choiceoffrequency
and
so can
adjust
thingssothat
he
is
operatingatthe
low
noise
endofthe
frequency
range
available to
him.
The
situation
with regard to
source
resistance
may
be less
flexible
because
the
source
resistanceisusually
determined
solely by
the
typeofsensor
used.
Once
the
commitment
to
a
particular
sensor
has
been
made,
it can be
difficult
to
adapt
the
systemtoanother
one.
Hence, in
applications
where
noise is a
uotcntial
problem,
the
choiceofsensor
shouldbemade
carefully
and
with
full regard for
the
noise
characteristics
of the amnlifier.
Two
situations
deserve special
attention,
the
first being
operation
fromasource
resistance very
much
lower
than
optimum,
and the
second
being
operation
fromasource
resistance very
much
higher
than
optimum.Ineither
case,
there
may be a
temptation
to
"improve"
the
source
resistance
situation
by
the
use of resistors. In
the
case
of
the low
source
resistance,
one
maybetemptedtoconnect
a
resistor in series
with
the
source.Inthe
caseofthe
high
source
resistance,
one
may be
temptedtoconnectaresistor
in parallel
with
the
source.
Unfortunately,
neither
approach
to
the
problem
does
any
good
and,infact,
both
will
result
in
further
degradationofthe
signal-to-noise
ratio
ahead
of
the
lock-in
amplifier.
The
series
resistor
adds
its
thermal
noisetothat
already
accompanying
the
signal.
Although
the
amplifier
showsa"better"
noise figure
than
before,
it is
only
because
the
ampl ifier
noiseisnow
less relative to
the
thermal
noise
of
the
combined
resistances (source
plus
IV-4
series
resistor).
The
signal is no larger (in
fact,itmay
well
be
attenuated),
the
noise is
greater,
and
the
improved
noise
performance
is illusory. Recall
that
noise
figure
only
relates
amplifier
noisetothermal
noise, and
does
not
denote
the
absolute
valueofamplifier
noise.
Connecting
a parallel
resistor
to
lower
a high
source
resistance has a similar
effect.
Even
though
the
thermal
noise
doesgodown,
the
signal
amplitude
goes
down
even
more.
For
example,
if a
source
of
resistanceRwere
paralleledbyanother
resistorofthe
same
value,
the
signal
ampl
itude
wouldgodown
by a
factoroftwo.
However,
the
thermal
noise
would
onlybereducedto.707ofits initial
value
(thermal
noise
varies
directly
with
the
square
root
of
the
resistance),
withanet
degradation
in signal-to-noise
ratio.
In
operating
from
low
source
resistances,
however,
one
can
usually
improve
the
situation
dramatically
by using a
transformer
to
raise
the
source
resistance
seen
by
the
amplifier.
The
improvement
one
obtains
withatransformer
is real
because
the
amplitudeofboth
the
signal and
the
noise is
increasedbythe
turns
ratio.
The
source
resistance is
increased by
the
squareofthe
turns
ratio.
For
example,
if
one
hadaten
ohm
source,
one
could
use a 1:
100
step-up
transformer,inwhich
case
the
amplifier
would
see a
source
resistance of
100
kn.
At
100
kS1
the
amplifier
adds
little
additional
noise. Even
though
the
thermal
noiseofthe
transformer
addstoth at of
the
source,
a very
considerable
improvement
is usually
achieved.
P.A.R.C.
manufactures
a
line of
suitable
signal
transformers,
each
designed
for
optimum
operation
over
a given
frequency
range. Perfor-
mance
information
can be
obtained
from
the
factoryorone
of its
representatives.
In using an
external
transformerinconjunction
with
'the
Model
128A,
single-ended
operationofthe
lock-in
amplifier
is advised.
The
extremely
high
inherent
common
mode
rejectionofthe
transformer
makes
differential
operation
of
the
lock-in
amplifier
unnecessary.
When
working
from
a high
source
resistance,
one
could,
in
principle,
use a
transformerinthe
same
mannertoimprove
noise
performance.
Unfortunately,
practical
transformer
design
considerations
usually
prevent
one
from
doing
so. As
a
result,
the
options
availabletoan
experimenter
working
with
a high
source
resistance device,
such
as a
photo-
multiplier,
are
limited.
Practically
speaking,
the
best
one
candoistomake
the
load
resistor
as large as possible.
The
larger
the
source
resistor,
the
less
the
shunting
effect
it will
have,
and
the
better
the
signal-to-noise
ratioatthe
input
of
the
amplifier
will be.
That
this
is so
becomes
clear
when
one
recalls
that
the
signal
amplitude
varies
directly
with
the
load
resistance, while
the
thermal
noise varies
with
the
square
rootofthe
resistance.
Note
that
the
entire
preceding
discussionofnoise
is based
on
comparing
the
noise
generatedbythe
amplifier
with
the
source
thermal
noise. In
many
situations,
other
types
of
noiseofinterference
may
accompany
the
signal as well and
could
even
dominate
it.
Where
this is
the
case,
the
amplifier
can
only
perform
"better"
than
the
noise figure
contours
indicate
because
the
noise
figures are basedona cornpari-
son of
amplifier
noise
with
the
minimum possible noise
which can
accompany
any
signal,
namely,
the
source
thermal
noise.
4.3 OPE;RATING THE MODEL
128A
4.3A
INliRODUCTION
Operation/of
the
Model
128Aisstraightforward.Inmost
instances,
the
operator
simply
connects
the
reference
signal,
waits
for
the
REF
UNLOCK
light to go
out,
and
then
connects
the
signal to be
measured.
The
Sensitivity
and
Phase
controls
are
then
adjusted
for
maximum
output
without
overload.
Should
overload
occur,
thedcfiltering
(Prefilter and
Time
Constant)
is
increased
and/or
the
sensitivity is
reducedasrequiredtoeliminate
the
overload.
The
readinq can
thenbetaken
.
•n
many
Situations,
achieving a successful
measurement
will
depend
notsomuchoncritically
adjusting
the
Model
128A
controlsason
taking
the
proper
steps
indicated
by
the
preliminary
considerations
discussed in
Subsection
4.2.
Factors
such as
proper
grounding
and
operating
inside
the
3
dB
contour
are
most
important
in
making
low level
measurementsofnoisy
signals.
4.38
REFERENCE
CHANNEL
(see
Subsection
4.9
for
additional
information)
Referende
Signal
Requirements
An outsltanding
featureofthe
Model
128A
is its
unique
reference
channel
circuitry
which
allows ittolock
onto
and
trackawide
range of possible
reference
input
waveforms.
Once
locked
on,
the
reference
remains
locked
on,
even if
the
reference
input
signal
changesinfrequency.
There
are
no
Reference
Channel
controlsofany
kind
which
must
be
adjusted
for
proper
reference
channel
operation.
Once
the
light goes
out,
all
that
remains
is to
adjust
the
Phase
controlssothat
the
Reference
signal
appliedtothe
mixer
is
-t
the
proper
phase
relativetothe
signal to be
measu
red. As
statedinthe'
specifications,
the
only
requirements
on the
reference signal are
that
it swing at least plus
and
minus
50
mV
with
respect
to its
mean,
thatitcross
its
mean
twice
each cycle,
and
thatitremainoneach
sideofthe
mean for
at least
100
ns. Sinewaves,
square
waves, triangle waves,
and
many
others
are all
suitable.
However,
for
best
phase
accuracy,
a 1 V rms sinewave is
recommended.
One
waveshape
whichatfirst
glance
may
seem
to suffice,
but
which
does
not,isthe
very
narrow
low
duty
factor
pulse.
For
example,
suppose
one
intended
to use as a
reference
signal
the
pulse
train
depicted
in Figure IV-7,
that
is, pulses
having an
amplitude
of 1 V, a
durationof1
JiS,
andaperiod
of 1 ms.
The
mean
of this signal
wouldbeabout1mV,
and
the
excursions
relativetothe
mean
wouldbe+999
mV
and
-1
mV.
Thus,
the
individual pulses, even
though
they
exceed
the
minimum
excursion
requirementonone
side by
some
950
mV,
lack
meeting
the
excursion
requirement
on
the
other
side
of
the
mean
by a full
49
mV
(50
mV
required),
with
the
result
that
the
Reference
Channel
will
not
function
properly.
Thus,
when
using pulses as
the
reference
input,
take
care
that
the
durationofthe pulses,
relative
to
the
pulse
period,isgreat
enough
to give a
mean
of at least50mV.Inthe
example
just
cited,
if the pulse
duration
were
increased to
100
JiS
as
shown
in Figure
IV-5
AMPLITUDE : 1 V
J
~
ERIOO
: 1
ml
DURATION:
1
III
MEAN:
I mV
'-
-'-_____
EXCURSIONS:
.!"f~vmv
A. PULSE TRAIN WHICH WILL NOT TRIGGER BECAUSE
EXCURSION CRITERIA NOT
MET
J
~
~
AMPLITUDE
: 1V
PEROO:
1ms
~tti~~~~
~~
liS
EXCURSIONS
:+900
mV
- 100 mV
-~
B PULSE TRAIN WHICH MEETS
All.
CRITERIA, INCLUDING
EXCURSION REQUIREMENTS
Figure
IV-7.
PULSE
TRAINASREFERENCE
DRIVE
IV-7B,
the
mean
would
increase to 100 m V,
and
the
excursions
(900
mV
one
side, 100 mV
the
other)
would
be
more
than
adequate
for
proper
triggerinq.
Even
though
the
Model
128A
can
accept
and
trackawide
range
of
possible
reference
siqnals, it is
nevertheless
important
that
the
reference
signal used be relatively
noise
free. Any
superimposed
noise
can
cause small
zero
crossings
to
occurinthe
region of
the
main
waveform
zero
crossings,
with
the
result
that
the
Reference
Channel
momentarily
"sees"amuch
higher
reference
frequency
than
what
is
really
there.
When
this
happens,
the
reference
"lock"
can
be lost.
Frequently,
moderately
noisy siqnals
canbecleaned
up
sufficiently
for
satisfactory
operationbyinterposing
a
single-section low-pass
filter
between
the
reference
siqnal
source
and
the
ReferenceInput
connectorofthe
Model
128A.
On
later
instruments,
spaceisprovided
for
mounting
the
filter
components
on
the
Reference
printed
circuit
board.
Locationofthe
mounting
holes
is
indicated
on page
VII·5.
To
install
the
filter,
transfer
the wire
which
norrnally
goes
to
quick-disconnect
J419
over
to
quickd
isconnect
J429.
Then
install
the
filter
components,aresistor
for
RX4
and
a
capacitor
for CX 1
(RX4
and CX 1 are de
siqnations
given
on
the
schematic
and
parts-location
diagram
for
these
corn-
ponents).
In
most
instances,
optimum
performance
is
obtained
by
setting
the
filter
corner
frequency
(f
::::
1/4rrRC)tothe
intended
reference
frequency.
Switches Three
switches
are
associated
with
operation
of the
Reference
Channel.
Oneofthem,
the
f/2f
SWitch, is
located
at
the
front
panel.
The
other
two are
internal.
The
front
panel
switch
determines
whether
the
mixer
will be
driven
at
the
frequency
of the
applied
reference
signal or at twice
the
frequencyofthe
applied
reference
signal.
The
"2f"
position
is
used
for
second
harmonic
studies.
For
normal
operation,
the
switchissetto"f".Toexamine
harmonics
higher
than
the
second,
the
operator
would
havetosupplyareference
signal at
the
frequencyofthe
harmonic
to be
measured.
One
should
check
the
position
of this
switch
before
connecting
the
Reference
signal,
when
the
switch
posi
tion
is
changed
during
operation,
the re"ference
channel
will
unlock,
and
time
will be
lost
III
waitill~]
for it to lock on
Loading...
+ 42 hidden pages