BRUEL & KJAER 2031 User manual

Instruction Manual
2031
Narrow Band Spectrum Analyzer Type 2031
033-0267 .
A narrow band spectrum analyzer featuring·
almost entirely digital construGtion. The
spectral analysis takes place in
400 con-
stant bandwidth lines across a frequency
range
which is selectable from 0- 10Hz
to 0- 20kHz in a 1-2-5 sequence, and ex-
tensive transient recording facilities make
the
2031 equally at home with continuous
and impulsive signals. The analysis takes
place in real-time
for all frequency ranges
up to
0- 2kHz. The results produced can
be exponentially or linearly averaged prior
to display on an
11" calibrated display
screen,
which may also be used to show
the time function, the instantaneous spec-
trum, and a hold max. spectrum.
+ BrUel & Kjcer

BROEL & KJ~R instruments cover the whole field of sound and vibration measurements.
The main groups are:
ACOUSTICAL MI;ASUREMENTS
Condenser Microphones
Piezoelectric Microphones
Microphone Preamplifiers
Hydrophones
Sound Level Meters
Precision Sound Level Meters
Impulse Sound Level Meters
Noise Dose Meters
Noise Level Analyzers
Standing Wave Apparatus
Calibration Equipment
Reverberation Processors
Sound Sources
ACOUSTICAL RESPONSE TESTING
Sine Generators
Random Noise Generators
Sine-Random Generators
Artificial Voices
Artificial Ears
Artificial Mastoids
Hearing
Aid Test Boxes
Audiometer Calibrators
Telephone
Measuring Equipment
Audio Reproduction Test Equipment
Tapping
Machines
Turntables
VIBRATION MEASUREMENTS
Accelerometers
Force Transducers
Impedance Heads
Accelerometer Preamplifiers
Vibration Meters
Accelerometer Calibrators
Magnetic Transducers
Capacitive Transducers
Complex
Modulus Apparatus
Bump Recorders
VIBRATION TESTING
Exciter Controls- Sine
Exciter Controls- Sine- Random
Exciter Equalizers, Random or Shock
Exciters
Power
Amplifiers
Programmer Units
Stroboscopes
STRAIN MEASUREMENTS
Strain Gauge Apparatus
Multipoint Selectors
MEASUREMENT AND ANALYSIS
Voltmeters
Phase Meters
Deviation Bridges
Measuring Amplifiers
Band-Pass Filter Sets
Frequency Analyzers
Real Time Analyzers
Heterodyne Filters and Analyzers
Distortion Measuring Equipment
Psophometers
Statistical Distribution Analyzers
Tracking Filters
RECORDING
Level Recorders
Frequency Response Tracers
Tape Recorders
Alphanumeric Printers
Digital Event Recorders
DIGITAL EQUIPMENT
Computers
Tape Punchers
Tape Readers
Bruel & Kjcer
DK-2850 NJERUM, DENMARK · Telephone: + 45 2 80 05 00 · Telex: 37316 bruka dk
1 Instruction Manual
2031
Narrow Band Spectrum Analyzer Type 2031
033-0267 .
'I r
---~--..-
.. •: ..
" . ·
•·
·· wr·
A narrow band spectrum analyzer featuring
almost entirely digital construction. The
spectral analysis takes place in
400 con-
stant bandwidth lines across a frequency
range
which is selectable from 0- 10Hz
to 0- 20kHz in a 1-2-5 sequence, and ex-
tensive transient recording facilities make
the
2031 equally at home with continuous
and impulsive signals. The analysis takes
place in real-time for all frequency ranges
up to
0- 2kHz. The results produced can
be exponentially or linearly averaged prior
to display on an 11" calibrated display
screen,
which may also be used to show
the time function, the instantaneous spec-
trum , and a hold max. spectrum.
+ BrUel &Kjrer
"'r q··
L ~- -- -- -- , - - -~ --~-- -- -! ~ '~- - - - - - - -- ·- - - - - - - .../

NARROW BAND SPECTRUM ANALYZER
TYPE 2031
September 1 978

This apparatus has been designed and tested according to class II of IEC publication
348, Safety Requirements for Electronic Measuring Apparatus, and has been supplied in
safe condition. The present instruction manual contains information and warnings which
should be followed by the user to ensure safe operation and to retain the apparatus in
safe condition.

CONTENTS
1. INTRODUCTION AND SPECIFICATIONS (PRODUCT DATA) ................................................. 1
2.
CONTROLS .......................................................................................................................... 13
2.1. FRONT PANEL OF NARROW OF NARROW BAND SPECTRUM ANALYZER
TYPE 2031 ............................................................................................. 13
Controls used for Analog Input and General Operation .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 13
Controls used to set the Triggering Conditions .. . ........ .. .... . .. .. .. .. .. .. .. .. .. ... .... . 15
Controls used in Recording and Transformation of Data ............................... 16
Controls used in Averaging of Data ... .... .. .... .. . .. .. .... .... .. . .. .. .. .. .. .. .. ... .. .. .... . .. 17
Controls used in the Display of Data .. .... .. .... .. . .. .. .... .. . .. .. .. .. .. .. .... .. .. .. .. .. .. .... 18
Controls used in•. the Input/Output of Data .. .... .. . .. .. .. .. .. .... .. .. . .... .... .... .. .... .. . 20
2.2. THE 2031 DISPLAY SCREEN .................................................................... 21
The 2031 Display Screen in Display of a Time Function . .... .. .... .. .. ........ .. .... . 21
The 2031 Display Screen in Display of a Spectrum .. . .. .. .. .. .. ... ........ .... .. .. .... . 22
2.3. REAR PANEL OF THE NARROW BAND SPECTRUM ANALYZER
TYPE 2031 ............................................................................................. 23
3. OPERATION ......................................................................................................................... 27
3.1. PRELIMINARY ADJUSTMENTS ................................................................. 27
3.2. CALIBRATION ......................................................................................... 28
Calibration using a Sinusoidal Reference ................................................... 29
Calibration using an External Noise Source .. .. .... .. .. .. .. .. .. .. .. .. .. . ...... .. .. . .. .. .. .. 30
3.3. GENERAL OPERATIONAL PROCEDURE ...................................................... 31
3.4 . CONTROL OF TRIGGERING ...................................................................... 31
Selection of Triggering Source .. .... .. . .... . .. .. .. . .. .... .. . .. .. .. .. .. .. .. .... .. .. .. .. .... .. .. .. 31
Use of the "Records after Trig." Control .... .. ............ .. .... .. .... .. . .. .. .. .. .. .. .. .... .. 32
3.5. CONTROL OF RECORDING ................. .. .................................................... 33
3.6. CONTROL OF AVERAGING ....................................................................... 34
Linear Averaging ..................................................................................... 34
Exponential Averaging . .. .. .. .. .. .. .. .... . .. . .... .. .. .... .. .... .. .. .. .. .. .. ... .. .. .... . .. .. .. .. .. .. 36
Hold. max. .............................................................................................. 36
3.7. USE OF THE DISPLAY CONTROLS .. .. .... .... .. .. .. .. .. .... . .. . .. .. .. . .. .. .... . .. .. .. .. .. .. .. 36
Display of a Time Function ...... .... .. .. .. .... .. .. ...... .... . .. .. .. .. .. .. .. .. .. .. ... .. ........... 37
Display of a Spectrum .. .. .. .. .. .. .. .... .. .... .. .. .. .. .... .... .... .. .. .... .. .. .. .. .. .. .. .. .. .. .. .. .. 39
Use of a Reference Memory .. .... .. ...... . .... .. .. . ........ .. . .. . .. .. .... . .. .. .. ... ........ . .. .. 39
Display of Difference Spectra .................................................................... 40
3.8. ANALYSIS OF TRANSIENTS ..................................................................... 40
Analysis of Single Transients .. .. . .... .. .. . .... .. .. . ........ .... .. .... .. .... . .. .. . ...... .... .. . .. 42
Automatic Transient Averaging .. ..... .... .. .. .. .. .... .. .... .. .. . .. .. .... .. . .. .. .. .. .. .. .. .. .. .. 43
Manual Averaging of Transients with Rejection of Bad Data .. .. . .. .. .. .. .. .. .. .. .... 43

3.9. ANALOG OUTPUT FROM THE 2031 .......................................................... 43
Analog Output to a Level Recorder Type 2307 ............................................ 44
Analog Output to an X-Y Recorder Type 2308 .......................................... 46
Exchange of the Level Recorder and X-Y Recorder interfaces ...................... 48
3.1 0. USE OF AN EXTERNAL SAMPLING FREQUENCY ........................................ 49
Use of the Tracking Frequency Multiplier Type 1901 with the 2031 ..... .. .. .... 50
3.11. RACK MOUNTING OF THE 2031 .................. .. ........................ .. .... .. .. .. .. .... 50
4. DIGITAL DATA TRANSFER AND REMOTE CONTROL OF THE 2031
VIA THE IEC INTERFACE ..................................................................................................... 52
4.1 . IEC FUNCTIONS IMPLEMENTED ................. . ............................................... 52
4.2. SELECTION OF ADDRESSES ..................................................................... 52
4.3. CODES FOR ADDRESSING 2031 CONTROLS .............................................. 53
F.S. FREQUENCY ..................................................................................... 54
TRIGGER "Records a. Trig." ............................................. .. ....... ................ 54
AVERAGING "No. of Spectra" . .................................................................. 55
TRIGGER ''Level'' ............. .............................................. .. ........................ 55
INPUT ''Att 0-100 dB'' ............................................................................ 55
TIME FUNCTION "Move" ........................................................................... 56
SPECTRUM GAIN ..................................................................................... 56
LINE SELECTOR ....................................................................................... 57
Other Controls, Upper Front Panel of 2031 ............................. . ................... 57
Other Controls, Lower Front Panel of 2031 ....................... .......................... 58
Differences Between Remote and Manual Control of the 2031 .......... .... ........ 59
4.4. REQUIRED FORMATS FOR DIGITAL INPUT
AND OUTPUT OVER THE 2031 IEC INTERFACE ........................................... 60
Mode# 0, Digital Input/Output of an ASCII Encoded Spectrum ...................... 60
Mode# 1, Remote Setting of Pushkeys and Output of Pushkey Settings .......... 61
Mode# 2, Digital Input/Output of Time Function .......................................... 62
Mode#3. Block Transfers MSBY-LSBY to and from the 2031 ....................... 62
Mode#4. ASCII Sensing of Pushkeys ......................................................... 63
Mode# 5, Digital Output of 2's Complement Encoded Spectrum .................... 64
Mode# 6, Block Transfers LSBY to and from the 2031 ................................. 64
Mode# 7, Block Transfers MSBY to and from the 2031 ................................ 64
Mode#8, Carry On Processing ...................................... . ........................... 64
Other Input/Output Modes, Manual Digital Input/Output ............................. 65
Other Input/Output Modes, Input of Alphanumeric Text to the Display Screen 65
4.5. USE OF INTERRUPTS IN 2031 DATA TRANSFERS ....................................... 66
Read-out of the Time Function with Interrupt Control ................................... 66
Read-out of the Complex Spectrum with Interrupt Control ............................. 67
Read-out of the Averaged or Instantaneous
Spectrum
with Interrupt Control ................................................................. 67
4.6. FORMATS USED IN BLOCK TRANSFERS .................................................... 68
Format of the 16-bit Time Function ......................................... ......... .......... 68
Format of Complex Spectrum ..................................................................... 68
Format of Instantaneous Power Spectrum ................................................... 68
Format of Averaged MS Spectrum .............................................................. 69
Format of Spectrum Display Buffer ............................................................. 69
Format of Reference Memory ..................................................................... 70
Format of the 8-bit Time Function .............................................................. 70
4.7. SPEED OF DIGITAL INPUT AND OUTPUT .................................................... 71

FEATURES:
• High resolution, constant
bandwidth frequency
analysis in 400 frequency
channels
• Eleven selectable
frequency ranges from 0-
10 Hz through 0-20 kHz
• Full transient recording
facilit ies with internal,
external, or free running
trigger and control of after
trigger recording
• Display of time signal,
instantaneous spectrum,
averaged spectrum, and
spectrum difference on
11" display screen
• Greater than 70 dB
dynamic range plus 9 dB
crest factor capabiI ity
• Built-in antialiasing filters
• Exponential and linear
averaging over 1 to 2048
spectra
• Hold max. facility allows
maximum level in each
channel to be held
• Selectable
flat
(rectangular) or Hanning
weighting
• Display of spectrum over
80 dB, 40 dB, or 20 dB
display range
• Memory allows storage of
spectrum for comparison
with later data
• First estimate transfer
function measurements
18-200
type 2031
Narrow Band Spectrum Analyzer
possible using memory
subtract feature
• Alphanumeric read-outs
directly from display
screen using line selector
• Read-outs using line
selector can be made
relative to values in other
channels or lines
• Output of displayed data
via analog
output
• Digital input and output via
IEC interface
• Settings of major controls
shown on display screen
via alphanumeric displays
• Pushkey
controls
externally programmable
via IEC interface
,..,_ • ...,f;ti'K'tOW<OANI\>a ..
l'"aol'
r :;-::: 1 1 ;~,
--
L- y~oo,.,_,._ _.:J
' '
.: .: .. : .: .
kll'l ... _ ........,.,
• Connection to IEC or IEEE
Std. 488 interface bus via
IEC interface
USES:
•
Analysis of acoustical and
vibrational signals
• Analysis of transients and
shocks
• Preventive maintenance
and machine monitoring
• Speech analysis and
phonetics
• Order analysis
• Physiological and
Neurological research

Introduction
The Narrow Band Spectrum Anal-
yzer Type 2031 is an instrument de-
signed
for the narrow band fre-
quency analysis of continuous and
transient data coming from acousti-
cal, vibrational, and other signal
sources. It calculates a 400 channel
constant bandwidth RMS power
spectrum, in dB, of the input signal.
The frequency range of the instru-
ment is pushkey selectable in a 1-2-
5 sequence from 0-10 Hz to 0-
20 kHz. Alternatively, by using an
external sampling source, the fre-
quency range can be made to float,
allowing the 2031 to track pheno-
mena of variable frequency, as in,
e.g.
order analysis . The time taken
to generate a single spectrum is
less than 200 ms, giving real-time
operation up to greater than 2kHz.
The 2031 operates by Fourier
transforming records of 1 024 sam-
ples of the input signal into the fre-
quency domain. The recording takes
place in a transient recorder having
two parallel memories . The trigger-
ing facilities available on the recor-
der enable the 2031 to be used in
the analysis of a wide range of sig-
nals.
The transient recorder's free run-
ning mode is used to analyze contin-
uous signals. Here, new records are
continuously taken and trans-
formed . Where the signals to be an-
alyzed are intermittent or transient,
the triggered mode is used. Here,
transformation only takes place on
receipt
of a trigger, the source of
which may be internal, with a trig-
ger level variable in 200 steps ac-
ross
the input voltage range, or ex-
ternal from a TTL pulse.
In
the triggered mode, the after
trigger recording setting selects pre-
cisely which record is transformed
with respect to the trigger. Its wide
range of adjustment, namely 0,0 to
9,9 record lengths in steps of 0,1
record lengths, enables transient
data to be analyzed even when they
are quite remote from the precise
moment of triggering. Further, in
the analysis of cyclic processes, it
allows part of the signal to be gated
and analyzed
independently from
the rest, e.g. the opening or closing
of a valve in a machine cycle.
2
The 2031 can be set such that
each new record taken is trans-
formed. Alternatively, in the analy-
sis of single transients, recording
and transformation may be stopped
as soon as
the first record has been
transformed .
Transformed data can be exponen-
tially or linearly averaged over 1 to
2048 spectra. Exponential averag-
ing is
useful in looking at running
processes, where changes are im-
portant, while linear averaging be-
comes preferable
where it is import-
ant that all data being input has an
equal influence on the final average
generated, e.g. in spatial averaging
or averaging of cyclic processes.
Note
that linear averaging is con-
trolled such that a true average is al-
ways displayed, this converging on
the final result. A hold max. mode
is also available,
allowing the maxi-
mum value occurring in each chan-
nel to be held.
Exponential and
linear averaging
and
the hold max . mode are equally
applicable to the analysis of continu-
ous and transient data. This means
that, e.g. in the previous example,
the signal generated by the opening
or closing of the valve could be aver-
aged over
many cycles.
Display
of the analyzed spectrum
takes place on an 11" calibrated dis-
play screen, which may also be
used
to show the time function.
Both the instantaneous and the av-
eraged
spectrum can be shown ,
and
either can be entered into a ref-
erence
spectrum memory and re-
called later for comparison with
new , incoming data. A further dis-
play mode gives the difference be-
tween new, incoming data and the
stored data. This allows changes in
spectra
to be identified, and where
sufficient stationarity exists, allows
the magnitude of a transfer function
to be measured.
Read-outs
from the display can be
plotted using a Level Recorder Type
2307 or, (optional), an X-Y Recor-
der Type 2308 via the analog out-
put, or digitally transferred to an
IEC compatible peripheral via the
IEC interface. This interface also
permits the digital read-in and read-
out of data, and the reading and the
setting of the front panel controls,
as used in the remote programming
of the 2031.
The frequency and RMS ampli-
tude of any channel in the spec-
trum, or the relative position of any
sample in the time function can be
read
from alphanumeric displays on
the display screen, by using the line
selector. (Note that when external
sampling is used, spectral read-outs
are
in terms of amplitude and chan-
nel number.) Further alphanumeric
displays give the settings of the ma-
jor controls such as the full scale le-
vel,
full scale frequency, etc.
Amplitudes are read from the
2031 in terms of dB. The reference
level is usually set to 1 JJV RMS,
but may be varied throughout a
range of ± 50 dB from that level. Alt-
ernatively, the RMS level in any
channel of a displayed spectrum
can be set to 0 dB, and all the fol-
lowing amplitude measurements
made relative to that level. Like-
wise, the position of a sample in
the time function can be read in ab-
solute terms, or relative to the posi-
tion of any other sample, as a sam-
ple number and as a time in sec-
onds.
The
amplitude display range of
the 2031 can be set to 80 dB,
40dB, or 20dB . The full scale level
can be varied over 80 dB in 10 dB
steps. Hence,
this allows a 40 dB or
a 20 dB amplitude window from the
displayed spectrum to be expanded
to fill the entire display screen. Simi-
larly, when a time function is dis-
played, the time axis can be ex-
panded by a
factor of 3.
Input to the 2031 is via a stand-
ard B & K measuring amplifier in-
put, allowing connection of most
B & K microphone preamplifiers and
other signal sources. An internal
amplitude reference is provided to
ease calibration . Frequency calibra-
tion is made unnecessary, since all
measurements are referenced to an
internal crystal controlled oscillator .
Prior to recording and transforma-
tion, the input signal is passed
through an antialiasing filter, auto-
matically selected with the fre-
quency range. This filter, which
cuts off at the selected full scale fre-
quency, is to prevent the introduc-
tion of aliasing distortion, and may
be bypassed if required. For exter-

nal sampling, either the internal or
external antialiasing filters may be
used.
Before transformation, the time
records may be weighted using a
Hanning weighting. This is usually
used in the analysis of continuous
signals to improve selectivity. For
the analysis of transients, however,
where the Hanning weighting might
modify the data under analysis, a
flat weighting, (i.e. rectangular or
no weighting), may be selected .
Frequency Analysis
The mathematical basis for all fre-
quency analysis procedures is the
so-called Fourier Transform. In its
most genera I form, it is expressed
by the Fourier integral pair as fol-
lows:
1 1
00
F(w) = - f(t) exp (-jwt) dt
21T -oo
(1)
f(t) = ~-~ F(w) exp (jwt) dw
(2)
Equation 1 transforms the time
function f(t) into its complex fre-
quency components F(w) for each
(angular) frequency w radians /s,
(w = 21rf,
where f is the frequency
in Hz). Equation 2 shows how the
original time signal f(t) can be com-
pletely synthesized from its original
frequency components.
It is possible to write discrete
equivalents of equations 1 and 2 as
follows:
1 N-
1 ( 21rnk)
F(k) = - L f(n) exp -j --
N n=O N
(3)
f(n) =
N-1
I
k=O ( 21rnk)
F(k) exp j -N-
(4)
These two equations are known
as the Discrete Fourier Transform,
or DFT, and each of the continuous
functions F(w) and f(t), extending
from + oo to -oo, is replaced by a fi-
nite number, N, of discrete sam-
ples, these being F(k) and f(n) respec-
tively. If Af is the frequency incre-
' ment between the samples, and At
is the time increment, then kAf is
the frequency corresponding to F(k),
750177
Fig.1. Discrete Fourier Transform of Complex Time Function, (real part only). Note the periodic-
ity of the spectrum
Real
Imaginary
761025
Fig.2. Discrete Fourier Transform of Real-valued Time Function. Note the symmetry of the real
part and the antisymmetry of the imaginary part
and nAt is the time corresponding
to f(n). The infinite integrals are
hence replaced by finite sums, but
many of the properties of the DFT
are parallel to those of the integral
transform. Equation 3 is known as
the forward transform, and equa-
tion 4 as the inverse transform.
Although it is of a discrete na-
ture, and hence lends itself to digi-
tal computation, direct evaluation of
the DFT is a lengthy and inefficient
procedure requiring some N2 com-
plex multiplications to perform an N
point transform. Fast Fourier Trans-
form, or FFT, is an algorithm which
allows a more efficient means of
evaluation, reducing the number of
multiplications required to approxi-
mately N log2 N.
In the generalized forward trans-
form, N complex time domain points
are transformed into N complex fre-
quency domain points, where N is
usually chosen to be a power of 2.
The N complex frequency domain
points are evenly distributed from
DC to the sampling frequency. Fur-
ther, the spectrum becomes peri-
odic, with a period equal to the sam-
pling frequency, i.e. the N complex
points are repeated up and down
the frequency axis with a period
equal to the sampling frequency.
Taking this periodicity into ~ account,
and referring to Fig.1, it is evident
that those points falling between
the Nyquist frequency, (i.e . half the
sampling frequency), and the sam-
pling frequency represent the nega-
tive frequency components of the
spectrum, while those points from
DC up to the Nyquist frequency re-
present the positive frequency com-
ponents.
A simplification leading to an in-
crease in efficiency occurs whe ·n it
is assumed that the time domain
points can only be real valued, (as
is
usually the case). The spectrum
will then be conjugate even, mean-
ing that the positive frequency com-
ponents can be generated from the
negative ones, and vice versa, i.e.
only the positive or the negative
components are needed to define
the whole spectrum. Using this, it
enables N real-valued time doniain
points to be transformed as N/2
complex points, a manipulation be-
ing made on the results to obtain
the spectrum of the original time
function from DC to the Nyquist fre-
quency.
Each of the points generated by
the above process wi II have the
form F(i) = ai + jbi. For each positive
component F(i), there must be a ne-
gative frequency component F(-i) =
3

ai- jbi. The power at this fre-
quency is the sum of the powers of
the positive and negative frequency
components, i.e. 2(ai2 + bi2). This
value may then be averaged over
multiple spectra to form the MS
value, and the square root taken to
form the RMS value.
Description
A simplified block diagram of the
2031 is given in Fig .3 . It can be
split into four sections, the Analog
Input, the Processor , the Display
and Analog Output, and the IEC In-
terface . Its architecture is based on
two 16 bit buses, the Y-bus and the
D-bus, over which all data flow be-
tween the various sections takes
place .
Analog Input Section
A block diagram of the Analog In-
put Section of the 2031 is shown
in Fig.4. The Input Amplifier has
two inputs. One, the Preamplifier In-
put, is the standard B & K 7-pin
type, allowing the connection of
most B & K Microphone Preamplifi-
ers, to which it also supplies power
and microphone polarization volt-
ages. The input of signals from
other sources is made via the Direct
Input . Both inputs have separate
sensitivity adjustments, such that
differing transducer sensitivities can
be accommodated .
Calibration of the Input Amplifier
and the rest of the 2031 can be
checked using an internal Refer-
ence Oscillator, whose frequency is
controlled such that the calibration
signal appears in channel 256 of
the resulting analysis. Its amplitude
is 100 dB referred to 1 J1V RMS.
(Alternatively, an external calibra-
tion signal can be used.) The 1 11V
RMS reference level can be altered
by± 50 dB in steps of 10 dB.
The Input Attenuation setting gov-
erns the level setting for the rest of
the analyzer . It can be used to re-
duce the input signal level in steps
of 10 dB over a 1 00 dB range, us-
ing the front panel controls. When
set to 0 dB, full scale deflection
corresponds to 1 mV, and since the
4
Front
Panel
Controls
Analog
Output
To Level Recorder
Digital Input and Output
Remote Programming
Information over I EC Bus
770258
Fig.3. Simplified Block Diagram of the 2031
Y-Bus
l[
Mode Set Register Master
Oscillator
1 l
·~
Reference
Oscillator Sample Rate Divider
Sa
~
Ext .
mpling
,~, L l
~
~ ~
!
D
I
nput o---
eamp.o--- Pr
I
nput
Input
f---
Amplifier
Input
Anti-
Attenu-
I-- aliasing
ation Filter
Sample
lnt
I--
and
I-- ADC
....__
errupt
Hold
Jl
D-Bus
770264
Fig.4. Analog Input Section of the 2031
lowest measurable voltage is
1
J1V, this gives a total operating
range of 1 60 dB. Any input over-
load is indicated by an overload indi-
cator, and an upper 10 dB indicator
shows when the input signal is
within 10 dB of full scale.
From the Input Attenuation, the
signal enters the Antialiasing Filter .
This is a 7-pole elliptical low pass
filter having a roll-off of
113 dB/octave, automatically se-
lected with the frequency range so
that it cuts off at the fu II seale fre-
quency. Its purpose is to avoid alia-
sing errors when the input signal is
later sampled, and it may be by-
passed, if required . The Antialiasing
Filters may also be used when sam-
pling is externally controlled. In this
case, the cut-off frequency of the
connected filter will correspond to
the frequency range setting .
The low pass filtered signal now
enters the Sample and Hold net-
work and Analog to Digital Conver-
ter, (ADC), where it is sampled at a
frequency 2, 56 times the full scale
frequency, the sampling frequency
being derived from a crystal con-
trolled Master Oscillator. Alterna-
tively, an external sampling fre-
quency can be used, whereupon
the full scale frequency becomes
the sampling frequency divided by
2,56 . Each sample taken is entered
into the ADC, where it is converted
into a 1 2-bit two's complement
word.
Processor Section
The block diagram of the Proces-
sor section of the 2031 is given in
Fig.5. The functions of the Proces-
sor are to control the input of data,
to Fourier transform the data, to av-
erage the spectra produced , to con-
trol the display of data, and to
sense the front panel controls .

All calculations are carried out in
the Arithmetic Logic Unit, (ALU), un-
der control of the program in the
Program Memory . The ALU uses
floating point arithmetic, and is pro-
vided
with 16 accumulator registers
and extension registers . The pro-
gram flow is controlled by the Pro-
gram Counter and a 16 level Stack
is used for nesting subroutines. The
Constant Register is used to enable
constants, held as part of the pro-
gram, to be transferred into the
ALU in a single instruction cycle.
The instructions contained in the
program are decoded in the Instruc-
tion Decoder. Since the front panel
control settings will influence which
subroutines are used, it also de-
codes these . The decoded instruc-
tions and control settings are con-
verted into control signals in the In-
struction Register. These signals
control all functions in the Proces-
sor.
The Main Memory, where all data
is
stored, has a capacity of 4K 16
bit words. Data is entered via the
Memory Data register into which-
ever address has been placed in the
Memory Address register . The Bit
Reversal allows memory addresses
to be read to the processing sec-
tion, (i.e. the ALU, etc.), in bit re-
versed
order, giving an increase in
efficiency when reordering the fre-
quency components after Fourier
Main
Memory
Fig .5. Processor Section of the 2031
Transformation. A further increase
in efficiency is obtained by use of
the sine table, which may also be
used
to generate the Hanning
weighting of the time function.
The Fourier Transformation pro-
cess uses floating point arithmetic
on a block basis, i.e. the same expo-
nent is used for the entire block.
The Maximum Magnitude Detector
is used to optimize the exponent.
The power calculation and averag-
ing
routines use true floating point .
Conversion of the analyzed spec-
trum from linear to logarithmic form
is aided by the Log Table and the
Shift Counter . The Shift Counter is
also used in floating point calcula-
tions .
Output and Display Section
The block
diagram of the Output
and Display section of the 2031 is
shown in Fig.6. The displayed data
is held in
the Display Memory. De-
pending on the settings of the front
panel controls of the 2031, either
the Input Function, (i.e. the time
function , the instantaneous spec-
trum, or the averaged spectrum), the
Stored Function, (i .e. the contents
of the reference spectrum memory,
which might be an instantaneous or
an averaged spectrum), or the differ-
ence between them will be passed
to the Display Memory by the Pro-
Level
Recorder
Output
Display
X-Y
Recorder
Output
(Option)
Y-Bus
D-B us
770265
770259
Fig.6 . Output and Display Section of the
2031
cessor under program control. Alt-
ernatively, the Input Function and
the Stored Function can be alter-
nately displayed. All updating of the
Display Memory takes place during
the X-flyback of the display screen,
under interrupt control .
The display screen is a 400 line
display, each frame lasting approxi-
mately 20 ms . Fig .7 shows how it
may be used to display a time func-
tion . Here, the first 8 lines are used
to display the trigger level set. The
next two lines are left blank, and
5

Indication of major control settings
Read-out of the Line Selector
Indication of the selected line or
channel
Indicators of input overload or upper
1 OdB
For entry of a signal from co-axial
sources
For input of signals from most 8 & K
Microphone Preamplifiers
Allows adjustment of the reference
level in steps of 10 dB
Gives continuous adjustment of the
input gain
6
Select whether the signal analyzed
comes
from the Preamplifier or Di-
rect Input . Also select the internal
calibration signal
Control trigger mode
Set input attenuation in steps of
1 OdB
Ret.,_,ce Adju1t. d8
0
- 10 • •10
•
I I I
--

....._ ___
Select full scale frequency
Select after trigger recording setting
from 0,0 to 9,9 record lengths
Select external sampling and allow
the internal antialiasing filters to be
bypassed
---•• Control trigger level
N•rrow 8•nd Spectrum An•lvze'
Type2031
Select time weighting function
Select number of spectra to be aver-
aged
Select digital input or output
Control the analog output
Control which data is displayed on
the display screen, and which data
is
output
Control the entry of data into the
memory
Control the entry of data to the dis-
play
For setting a reference on the dis-
play screen
Select the displayed amplitude
range of the displayed spectrum
Select a line or channel
Expand and move a displayed time
function, and move a displayed spec-
trum up and down the screen
For
control of the appearance of the
display
Resets the 2031 to a preset combi-
nation of control settings
For control of recording
For control of averaging
•
Select the averaging mode
7

Fig.7 . Display of a Time Function on the 2031, (expanded mode)
then 341 of the remaining 390
lines are used to display every third
sample of the time function, (since
the display has 400 lines, and the
time function contains 1 024 sam-
ples, it is not possible to display all
of the samples simultaneously) . Alt-
ernatively, in expanded mode, any
390 consecutive samples of the
time function can be displayed. The
390 samples displayed can be
moved
within the time function.
Display of a 400 channel spec-
trum can take place over an 80 dB,
a 40 dB, or a 20 dB display range.
The top
line of the display can be
made
to represent from + 10 dB to
-70 dB with respect to the full
scale input level, in steps of 1 OdB.
This enables either the whole spec-
trum to be shown, or a 20 dB or
40 dB window from it to be ex-
panded
to fill the entire display
screen. Display of the difference be-
tween the Input Function and the
Stored Function is in terms of posi-
tive and negative dB values dis-
placed about a 0 dB line. Again, an
80 dB, 40 dB, or 20 dB amplitude
range may be selected, with a dis-
play gain of± 40 dB in 10 dB steps.
Simultaneously with the display
of a time function, spectrum, or
spectrum difference, the Character
Generator gene~ates alphanumeric
displays on the display screen to
show the settings of the full scale
8
input level, the full scale frequency,
the number of records of after trig-
ger recording, and the number of
spectra to be averaged, (and when
a linear average is in progress, the
number of spectra already aver-
aged), along
with an indication of
whether an overload has occurred
during the generation of the dis-
played data. They also give the read-
out of the line selected by the Line
Selector, which in the case of a dis-
played time function will be the
time and sample number of the se-
lected sample in the time function,
measured either in absolute terms
or relative to the position of any
other sample. In the case of a dis-
played spectrum or spectrum differ-
ence, the read-out will be the chan-
nel number, the frequency in hertz,
and
the RMS level in dB of the se-
lected channel . The dB read-out
may be referenced either to the ref-
erence level set in
the Input Ampli-
fier, or to the level in any channel
of any displayed spectrum or spec-
trum difference . (Note that when ex-
ternal sampling is selected, the fre-
quency read-out is suppressed.)
Any displayed information may be
output to a Level Recorder Type
2307 or an X-Y Recorder Type
2308, (X-Y Recorder output via op-
tional card ZN 0204 which replaces
Level Recorder output).
Output is al-
ways of the displayed information.
IEC Interface
Digital input and output, and the
transfer of remote programming in-
formation takes place over the IEC
Interface of the 2031, which is de-
signed according to the IEC propo-
sal, "Standard Interface for Pro-
grammable Measuring Apparatus,
Byte-Serial Bit-Parallel" TC 66. The
interface can be operated with or
without a controller.
When operated without a con-
troller, the interface allows the read-
in and read-out of any spectrum,
(be it averaged, instantaneous, or
difference), and the read-in and
read-out of the front panel settings.
Read-ins and read-outs are in AS-
CII.
When the interface is operated
with a controller, then it becomes
possible to read in or to read out
from any data block in the Main Me-
mory. This extends the read-in and
read-out facilities outlined above to
include, e.g. read-out of the time
function, read-out of the instantane-
ous spectrum with a linear ampli-
tude scale, access to the alphanum-
eric displays on the display screen,
etc.
Further, read-in and read-out
may be in ASCII or in binary form.
The 2031 is completely program-
mable over its IEC interface. All of
the front panel pushkeys can be
remotely sensed and controlled.
Firmware Routines
The program used to control oper-
ation of the 2031 can be divided
into three parts, namely, the Main
Program, the Analog Input Interrupt
Routine, and the Output to Display
Interrupt Routine . The Main Pro-
gram operates continuously with
new samples being placed in the
Main Memory, and data being trans-
ferred from the Main Memory to the
Display Memory under the interrupt
control of the two other routines .
The
time taken for the Main Pro-
gram, which controls the weighting,
transformation, power spectrum cal-
culation, and the averaging rou-
tines, is less than 200 ms, includ-
ing the time lost due to interrupts.
Since operation of the 2031 is buf-
fered, i.e. recording and transforma-
tion take place in parallel, this gives
full real-time operation in all fre-
quency ranges up to 0-2 kHz.

The averaging routines use three
different algorithms. The first of
these, used in linear averaging, is
as
follows:
-------, 1 ~ n ~ N
n
where Yn is the current average,
Y n-1 is the previous average, Xn
is the current instantaneous value,
and N is the number of spectra to
be averaged. This algorithm will al-
ways cause a true arithmetic aver-
age of X1 to Xn to be displayed, irre-
spective of the value of n. Yn con-
verges on YN as n approaches N,
and when n = N, the average stops
and the result is held. The averager
may then be reset, and a new aver-
age started. Alternatively, the new
average may be started and added
to Y Ni. Since the averager accumu-
lates on an energy basis, this al-
lows a number of averages to be
added, the RMS amplitudes of the fi-
nal spectrum formed being cor-
rected for the number of additions.
The exponential averaging algor-
ithm is identical to the linear averag-
ing algorithm, except that n is re-
placed by N/2, i.e.,
y =
n
(N/2- 1) Yn_
1 +
Xn
N/2
This produces an exponentially
weighted average, i.e., the most re-
cent data has the greatest influence
on the result, and older data is grad-
ually forgotten. The average contin-
ues for n~ N/ 2. Note that both ex-
ponential and linear averages can
have pauses introduced while they
are in progress, allowing, e.g., a
change of signal source.
The value of N is selected via the
2031 front panel controls, and for
the same value of N, exponential
and linear averaging will produce re-
sults having similar statistical accu-
racies but not necessarily the same
numerical values. In exponential av-
eraging, the first spectrum output is
equal to the first instantaneous
spectrum input, whereafter, averag-
ing proceeds according to the algor-
ithm given above. In the case of de-
terministic signals, this gives a fas-
ter convergence.
Exponential and linear averaging
for the same data points are com-
15
QJ
14
"0
.e
13
a.
E
12
<{
~ -o-- -o--0-<>---<>--- -o- .()...- -0
11
10
9
8
6
5
4
3
2
\\
X
\ \\\\ x
\
'><
'\.
><,
---+- Instantaneous values
-X- Exponential Average
-~-Linear Average
'>< ......
X..-x:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17· 18 19 20
770267 / 1
Fig.8 . Comparison of Exponential and Linear Averaging
Frequency
0,1 0,2 0,4 0,6 0,8 1
1
x-
T 2,0 4,0 6,0 8,0
0
~
r-- ......
.........
~
.........
\ h
\
,.....,
\
If
\\
f\n
(\
1--T-i
I \ I ..-
II
I
Flat
cb
\ I \
l I \
Hanning /t\
II
If\
II I/ I
II
,,
il
I I \
I I I I I
I,.._
I,\
CD
"0 10
20
30
40
50
60
770260
Fig.9. Comparison of Flat and Hanning Weighting Functions
pared, for N=8, in Fig. 8. At n=8,
the linear average stops, while the
exponential average continues .
The third averaging algorithm
gives the hold max. mode of the
2031. Here, the maximum level oc-
curring in each channel is held, and
a spectrum is developed consisting
of these maxima.
The Analog Input Interrupt Rou-
tine controls the input of data, set-
ting the Trigger conditions and the
After Trigger Recording according to
the front panel controls. The After
Trigger Recording is used in the
2031's triggered mode to allow the
record transformed with respect to
the trigger to be varied from that im-
mediately before the trigger to that
ending 9, 9 record lengths after the
trigger, in steps of 0, 1 record
lengths .
The records taken by the 2031 in
both its free running and triggered
modes can be transformed with a
flat or a Hanning weighting. Where
the resolution, {3, is defined as be-
ing one fourhundredth of the analy-
sis range, the 3 dB bandwidth using
flat weighting is 0,88 {3, while Han-
ning gives 1 ,44{3. The possibility of
sidelobes, however, normally
means that flat weighting gives the
worse selectivity of the two, mean-
ing that Hanning weighting is usu-
ally used with continuous signals.
In the analysis of transients,
though, where the Hanning weight-
ing might modify the captured data,
use of a flat weighting is usually
preferred. Flat and Hanning weight-
9

worse selectivity of the two, mean-
ing that Hanning weighting is usu-
ally used with continuous signals.
In the analysis of transients,
though, where the Hanning weight-
ing might modify the captured data,
use
of a flat weighting is usually
preferred. Flat and Hanning weight-
ings are compared in the frequency
domain in Fig.9.
Although said nominally to start
at 0 Hz, the centre frequency of the
first channel of the analysis is in
fact f3 Hz. The frequency range of
the analysis is hence f3 Hz to
400 /3Hz.
Examples of Use
Preventive Maintenance
The vibration spectrum of a me-
chanical part of a machine has long
been recognized as a means of pre-
dicting the impending failure of that
part . The 2031 provides a fast and
efficient method of obtaining the vi-
bration spectrum. The Accelerome-
ter Type 4366 is attached to the
part of the machine under investiga-
tion, and the acceleration signal
which it produces is entered into
the Charge Amplifier Type 2635 .
The
2635 gives a conditioned out-
put in terms of acceleration, veloc-
ity, or displacement, which is anal-
yzed using the 2031. Whether the
machine may continue operation, or
should be shut down for repair may
then be decided by comparing the
Transmitting Room
Tape Recorder
7003
Accelerometer
4366
Charge Amplifier
2635
~
~
• •
I I
Machine under investigation Narrow
Band Spectrum Analyzer
2031
170508
Fig.1 0. Use of 2031 in Preventive Maintenance
/' D~o l
~ ~
/pt ~ I
..
~ .< ?, ~ · :~i=~=:
--
Tektronix 4051 Narrow
Band Spectrum Analyzer Hewlett-Packard
HP 9825A
2031
770262
Fig.11 . Use of 2031 with a Desk-top Calculator
results obtained with known allow-
able limits.
For off-line preventive mainte-
nance, a short example of the vibra-
tion signal could be recorded, e.g.
on a Tape Recorder Type
7003, and
brought back to the laboratory for
analysis . Where the 2031 is con-
nected to an IEC or IEEE compatible
desk-top calculator, the results ob-
tained could be compared with a lib-
rary of reference spectra held on a
cassette.
Where the results indicate
Receiving Room
that further measurements are re-
quired, the 2031, with or without
the calculator, could be taken to the
measurement site.
Connection to Desk-top
Calculators
Since the 2031 is equipped with
an IEC interface, it can be con-
nected into an IEC interface bus sys-
tem. Further, since the only signifi-
cant difference between this inter-
face and one designed according to
the IEEE standard 488-1 975 is the
Microphone Power Supply
2807
Microphone
4165 Microphone
4165 Microphone
Preamp!ifiers 2619
10
Power Amplifier
2706 Noise Generator
1405
Fig.12. Use of 2031 for measurement of Transmissibility
Narrow Band Spectrum Analyzer
2031
770261

Accelerometer
4368
~
~
Tracking Frequency
Multiplier 1901
Narrow Band Spectrum Analyzer
2031
77 0263
Fig. 13. Use of 2031 in Order Analysis
Of part i cular interest here is the
connection of a 2031 to an IEC or
IEEE compatible desk-top calculator.
The ca leuI at or may then be used to
operate on data supplied by the
2031, as well as completely rem-
otely control the 2031 in automatic
test sequences. Data may also be re-
called
from the calculator to the
2031 for display and comparison
with new, incoming data. Further,
Fig.14. Rear Panel Sockets of the 2031
Input Characteristics:
Input: Either "Direct Input " or standard
B & K 7 pin
"Preamplifier Input"
Input Impedance:
Direct. 1 MO / I 1 00 pF
Maximum Input Voltage: 2031 is a
Safety Class II instrument (IEC 348) . For
safe operation in accordance
with IEC
348, the voltage of the signal or signal
ground relative to earth must not exceed
the 2031 can become part of a
larger test system, where it is
called
up on demand over its IEC in-
terface to carry out tests and meas-
urements.
Bruel & Kjrer suggests the Tek-
tronix 4051 and the Hewlett-Pack-
ard HP 9825A for use with the
2031 and other 8 & K instruments
equipped with an IEC interface . Rec-
ommended memory size for the
4051 is 32 K. Recommended mini-
mum memory size for the 9825A is
15 K (Memory option 001 ). The
9825A should also be equipped
with Hewlett-Packard options
98034A, 9821 OA and 9821 3A .
Estimates of Transfer Functions
and Transmissibilities
The spectrum difference mode of
the 2031 can be used to measure
the magnitude of a transmissibility
or transfer function, provided that
there is sufficient stationarity in the
process under investigation. Fig.12
shows how this might be used to
measure the transmissibility of a
wall or panel . The transmitting
room is excited with white noise.
The spectrum in the transmitting
room is obtained using the micro-
Specifications 2031
42 V RMS (sine). To ensure safe opera-
tion within IEC 348 at higher voltages,
the user must limit all input currents to
0,7 mA peak.
Sensitivity: 11 full-scale sensitivity
ranges, (overload
limit of ADC), 10 dB
steps
from 66 dB pV to 1 66 dB pV RMS
(sine).
Input Attenuator: 0 to 1 00 dB in 10 dB
phone and the 2031, and placed in
the reference memory. The 2031 is
then switched to the microphone in
the receiving room. When it is
switched to its difference mode, the
2031 will show the difference in
dB
between the spectrum in the
transmitting room and the spectrum
in the receiving room, i.e. after cor-
rection for the room constants, the
transmissibility of the wall.
Order Analysis
t=ig .1 3 shows how the 2031 can
be used
for order analysis. Here,
the Tracking Frequency Multiplier
Type 1901 is used as an external
sampling source for the 2031. The
1
901 inputs a signal related to the
fundamental frequency of the mo-
tor, such as the pulses provided by
the Magnetic Transducer MM
0002 . It then outputs some multi-
ple of this frequency, selected on
the 1901, which becomes an exter-
nal sampling frequency for the
2031 . The 2031 analyzes the sig-
nal coming from the accelerometer
end preamplifier in a range from
0 Hz to 1/ 2,56 of the frequency
supplied by the 1901. In a normal
analysis, with internal sampling con-
trol, if the speed of the motor var-
ied, the positions of components of
the signal under analysis related to
the motor speed would change on
the 2031 display screen. However,
use of the 1901 ensures that the
highest frequency analyzed is re-
lated
to the motor speed. Hence, if
the motor speed changes, the analy-
sis range changes in sympathy, en-
suring that the speed related compo-
nents stay in the same channels on
the screen . This allows the relation-
ships between the amplitudes of
the various motor speed related
components , or "orders", and the
motor speed itself to be studied.
steps, accurate to± 0,1 dB
Gain Control: 0 to 10 dB
Sensitivity Adjustment (Direct) :
+ 4, 7dB to - 10dB
Sensitivity Adjustment (Preamp):
+ 4, 7dB to - 10dB
Amplitude Reference: 1 00 dB referred
to 1
pV , frequency 64% of selected full
scale frequency
11

Antialiasing Filters: 11 filters automati -
cally selected
with frequency range. Max
±0 ,2 dB ripple in the passband,
113 dB/octave roll off . Provide at least
70 dB attenuation of any components in
the input signal having frequencies
higher than 1, 56 times the full scale fre-
quency. The filters can be bypassed, if re-
quired
Sampling: 2,56 x full scale frequency ,
automatically selected with the fre -
quency range for internal sampling. Ex -
ternal sampling sets the full scale fre-
quency to 1/ 2,56 of the external sam-
pling rate
Analog to Digital Conversion: 12-bit
two's complement. Quantizing error maxi-
mum± 1/ 2 LSB
Overload Indicator: Indicates overload in
either the input amplifier or the ADC
Upper 10 dB Indicator: Indicates when
either the input amplifier or the ADC is
operating in
the upper 10 dB of its dy-
namic range
Analysis Characteristics :
Full Scale Frequency: 10Hz to 20kHz
selectable in a 1- 2-5 sequence
Weighting: Flat (rectangular), or Hann ing
Number of Synthesized Filters: 400,
generated from 1 024 input samples
Filter Spacing, /3 : reciprocal of the input
record time duration , or full scale fre -
quency / 400
3 dB Bandwidth : 0, 88/3 for flat weight-
ing, and 1, 44/3 for Hanning weighting
Noise Bandwidth : /3 for flat weighting,
and 1 ,5/3 for Hanning weighting
Amplitude Linearity : (for 32 spectra av-
erage) ± 0,1 dB or ± 0,01% of full scale
at overload, whichever is greater, (no ad -
j ustment necessary, two-tone test)
Frequency Accuracy and Stability :
0,01 % w i thout warm-up , (no adjustment
necessary)
System Frequency Response: ± 0,2 dB
with antialiasing filters bypassed. or
± 0,3 dB
otherwise, 2Hz to 20kHz
Noise Level: (for 32 spectra average)
less
than 0 dB/ 1 1N or greater than
76 dB below overload, whichever is grea-
ter
Real Time Frequency Range: > 2 kHz
Transient Analysis:
Trigger Modes: Internal, External, or
Free- run
Trigger Level: Adjustable in 200 steps
across
the input voltage range. Trigger le-
vel indicated
on display screen
Trigger Slope: positive for positive trig -
ger level and negative for negative trig-
ger level
After Trigger Recording: Adjustable in
steps
of 0,1 record length from 0,0 to
9, 9 record lengths after trigger
Memory Period: equal to in seconds,
400 / full scale frequency in Hz
Record Control: manual transient cap-
ture, or automatic capture on the next
trigger
Transient Averaging: automatic averag-
ing of a succession of transients , or aver-
aging with manual control after verifica-
tion of data
Averaging :
Linear:
linear average of a preset num -
ber of spectra producing a true power av-
erage. A true average is always dis -
played, and
the number of spectra aver-
aged is indicated on the display
Exponential: the number of spectra indi-
cated on the display gives the effective
averaging
time
Hold Max.: the maximum level occurring
in each channel is held
No. of Spectra: 1 - 2048 in 12 binary
related ranges
Controls: Start , Proceed, and Stop
Memory:
"Store" stores the instantaneous or aver-
aged spectrum and the corresponding al-
phanumerics shown on the display
"Protect" prevents further updating of
the memory
Display:
The display screen shows the input func-
tion, (i .e. time function , instantaneous
spectrum , or averaged spectrum), the
function stored in the memory , the differ-
ence between the input function and the
stored
function , or the input and stored
functions alternately with a fast or slow
alternating frequency
Size: 11 "
Display Area: 150 x 210 mm (6 in x
8,25 in)
Scale Lines: 41 hor izontal lines are elec-
tronically generated directly on the
screen
for parallax free readings
Line
Frequency : 20.48 kHz
Frame Frequency: 47,7 Hz
Spectrum Range : 20 dB, 40 dB or
80dB
Spectrum Gain : in 10 dB steps over an
80dB range
Alphanumeric Read-outs: full scale le-
vel,
fu II seale frequency . records after
trigger , number of spectra averaged, se-
lected frequency line or sample number
selected frequency or time , selected level
Line Selector: line moved either continu-
ously or in single steps to the right or
left , indicated by an intensified column
Amplitude at Line Selector: shown on
display referenced to
11N RMS, or 1pV
RMS ±50 dB in 10 dB steps, or refer-
enced to the RMS level in any other
channel of the spectrum
Time at line selector: shown on display
referenced to
first sample or relative to
any sample in the record
Time Function Display: Every third sam-
ple,
or in expanded mode 390 conti-
guous samples selectable by 'Time Func-
tion Move"
Level Recorder Output :
Output Impedance: 1 kO
Maximum Output Voltage: + 15 V (nomi-
nal 12V)
Digital Input and Output :
IEC TC 66 Standard Interface* for pro-
grammable measuring apparatus
Functions: SH 1, AH 1, T 5, L 3, SR 1,
RL 0 . PP 0. DC 0 and DT 0
Data: Functions selected by Display Se-
lector,
or with controller, any main me-
mory block
Code: ISO 7-bit code, (i.e. ASCII but
without the parity bit) or binary (selected
from controller)
Remote Control: Front Panel settings
can be
input and output via standard in-
terface
Miscellaneous:
Power Supply: 100, 115 , 127 , 200 ,
220 , 240 V AC ± 10%, 50 to 60Hz, ap-
prox. 120W
Environment: temperature range, (for op-
eration within specifications) + 5°C to
+
40 °C
Storage Temperature -25 °C to + 75°C
Cabinet: supplied as model A (light -
weight metal cabinet) or C (as A but with
flanges for standard 19 " racks)
Dimensions and Weight: (A-cabinet with-
out feet)
Height:
310.4 mm (12 ,2 in)
Width : 430 mm (16,9 in)
Depth:
500mm (19 , 7in)
Weight : 22 kg (48 ,5 lb)
Accessories Included : Mains cable AN
0020
1 B & K plug JP 0101
1 BNC plug JP 0035
Cam disc for 2307 OD 0253
Remote control cable for 2307 AQ 0035
Cam switch cable for 2307 and Remote
1 cable
for 2308 AQ 0034
Signal cable for 2307 AO 0064
Signal cable for 2308 AO 0087
X-Y Recorder Output ZN 0204 (optional):
(replaces Level Recorder Output)
X-Deflection:
Read-out time: switch selectable,
45, 118, 263 s
Output Voltage: staircase 0 to + 10 V in
400 steps, (linearity better than 0,1 %) .
Full scale deflection in " cal" mode
Output Impedance: 1 00 n
Y-Deflection:
Output Voltage: 0 to + 10 V, in 256
steps (accuracy better than ± 10 mV).
Full scale
deflection in "cal" mode
Output Impedance: 1 00 n
Accessories Available:
Cable AO 0129 for connection to IEC in-
terface bus
Cable AO 0157 for connection to IEEE
std 488 interface bus
Preprinted Level Recorder paper
QP
1103
Preprinted paper for X-Y Recorder QP
1002
*(IEEE Standard 488 - 1975 and ANSI MC1.1 1975)

2. CONTROLS
2.1. FRONT PANEL OF NARROW BAND SPECTRUM ANALYZER TYPE 2031
The front panel controls of the 2031 can be conveniently grouped as follows:
1. Controls used for analog input of data and general operation.
2. Controls used for recording and transformation of data.
3.
Controls used in averaging of data.
4.
Controls used in the display of data.
Many of the controls have an LED indicator to show their status. The indicator comes on
when the function is active.
2.1.1. Controls used for Analog Input and General Operation
Preamp. Input --:!--I f •-
Signal Ground --:::1--=--- t• J
Gain Control
w--~&IMetrllftiAMI\'.Mr
1.,.,_1\l,U
liii ' ll--
--::--System Reset
1--::-- Power On
780023
Fig.2 . 1. Controls used for Analog Input and General Operation
13

PREAMP. INPUT 7-pin input for signal entry from most Bri.iel & Kjcer Microphone pream-
plifiers. Input impedance 1 MO in parallel with 1 00 pF. Accepts stand-
ard Bri.iel & Kjcer plug part number JP 0701. The connection diagram
is given in Fig.2.2.
Heater Voltage
+ 12,6 V (100 rnA)
Heater Voltage
+ 6,3 V (200 rnA)
Signal Input
Ground
Polarization
Voltage
+ 200 V (10 J.LA)
780028
Fig.2.2. Connection diagram, 2031 PREAMP. INPUT, external view
DIRECT INPUT Co-axial input to the 2031. Input impedance 1 MO in parallel with
1 00 pF. Accepts standard Bri.iel & Kjcer plug part number JP 01 01 or
BNC adapter part number JP 0144. Maximum input voltage, 250V
RMS (sine), see Section 3.1.
SENSITIVITY Two screwdriver operated potentiometers placed below their respective
INPUTS, allowing the input gain to be adjusted to compensate for dif-
fering transducer sensitivities. They operate independently over a
range
of+ 4,7dB to -10dB.
REFERENCE ADJUST dB Allows adjustment of the 2031 reference level in steps of 10 dB.
When it is set to "0", then assuming the input gain is correctly ad-
justed, the reference level is 1JiV RMS. Range of adjustment,± 50 dB.
GAIN CONTROL Gives continuous adjustment of the input amplifier gain. When set to
"Cal.", the gain is set to maximum. The maximum variation from this
position is approximately -1 0 dB. Operates simultaneously on both
the PREAMP. and DIRECT INPUTS.
UNCAL Indicator showing when the GAIN CONTROL is not set to "Cal.".
OVERLOAD Indicates when either the input amplifier or the ADC is overloaded.
UPPER 10 dB Indicates
that either the ADC or the input amplifier is operating in the
upper 10 dB of its dynamic range.
INPUT "Preamp."- connects the PREAMP. INPUT to the input amplifier.
14
"Direct" -connects the DIRECT INPUT to the input amplifier.
"Reference 1 00 dB 11V" - injects a reference signal into the selected
INPUT
for internal calibration adjustment. With a correctly adjusted in-
put gain and the REFERENCE ADJUST dB set to "0", the amplitude of
the signal is 1 00 dB RMS referred to 11J.V. The frequency of the signal
is derived from the sampling frequency such that it always falls at the
centre of line number 256 of the resulting spectrum, irrespective of
the F.S. FREQUENCY SETTING.