Brüel & Kjær 2113 User guide

o::.:::-~ ~ ~ . 
-· ~
...... . , •(1..'
• Audio Frequency Spectrometer
Type 2113
A precision instrument for 1/3 and
1/1 Octave analysis
of sound,
vibration or voltage over the 25 Hz
to 20 kHz range which with optional
filters can be extended to cover
the 2 Hz to 160 kHz range. Linear ,
A, 8, C and D
networks are included .
With a 8 & K microphone the
Spectrometer conforms to
international standards for precision
and impulse sound level meters.

Audio Spectrometer
Type 2113
July 1970

CONTENTS
1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. TECHNICAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Input Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Filter Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Output Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Record and Meter Section .......................... 20
Power Supply ................................... 29
Overall Performance .............................. 29
3. CONTROLS ......................................... 34
Front Panel ............................. . ....... 34
Rear Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4. OPERATION ........................................ 40
Adjustments ... . ................................ 40
Sound ......................................... 42
Vibration ....................................... 46
Voltage ........................................ 48
Insert Voltage Calibration .......................... 50
Extension Filters ................................. 52
5. ACCESSORIES ....................................... 54
6. USE WITH OTHER INSTRUMENTS ...................... 57
Level Recording .................................. 57
Measurement Accuracy
............................ 67
Tape Recording
.................................. 69
Measurement of Absorption Coefficient ............... 71
Measurement of Power Spectral Density . . . . . . . . . . . . . . . 72

7. PRECISION SOUND LEVEL MEASUREMENTS ............ 73
Precision Sound Level Meters ....................... 73
Impulse Precision Sound Level Meters ................. 74
Complementary Equipment ......................... 75
Operational Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Frequency Weighting
Network 0 ..................... 77
8.
DEFINITIONS ....................................... 79
Ideal Filter ...................................... 79
Effective Bandwidth
.............................. 79
Filter Centre Frequency ........................... 81
Filter Band-Edge Frequency ........................ 81
Bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Noise .......................................... 83
9. SPECIFICATIONS .................................... 84

1. INTRODUCTION
The 2113 Audio Spectrometer is a precision solid state instrument for
the analysis of sound and vibration. When employed with a suitable B & K
condenser microphone
it becomes a precision sound level meter and an
impulse sound level meter to the respective IEC 179 and DIN 45 633 recom-
mendations.
The
2113 is basically a wide range measuring amplifier plus a set of
contiguous 1/1 and 1/3 octave bandpass filters, covering the 25 Hz to
20 kHz audio range. With additional filters this range can be extended from
2Hz to 160kHz. The filters conform to the following recommendations.
IEC 225- 1966
ANSI S1.11- 1966
DIN 45 652
(
1/3 octave Class Ill and
1/1 octave
Class II).
"Lin", "A", "B" and "C" networks are also provided as well as the
recently introduced "D" network for aircraft noise analysis. Both filters and
weighting networks
can be selected manually, or if automatic synchroniz-
ation
with other equipment is preferred selection is via remote control.
Operation is considerably simplified by the inclusion of automatic input
and output overload indication together with direct indication of the meter
range on the meter scale. The
scales are interchangeable and calibrated for
sound, vibration, voltage, power spectral density and sound absorption co-
efficient.
For normal use the equipment can be powered by an AC mains supply,
but provision is made for 12 V battery operation when field measurements
are required.
5

2. TECHNICAL DESCRIPTION
A block diagram of the 2113 is shown in Fig.2.1. It is divided into the
following parts:
1. Input Section.
Input Circuit and Attenuator I.
Reference Oscillator.
Amplifier Sections I and II.
Input Overload Indicator.
2. Filter Section.
1/3 Octave Filters.
1/1 Octave Filters.
Weighting Networks.
Linear
Ranges.
Remote Control.
3. Output Section.
Amplifier and Attenuator Sections I and II.
4. Record and Meter Section.
Recorder Amplifier.
Meter Amplifier I.
Output Overload Indicator.
RMS Rectifier.
Meter
Amplifier II.
5. Power Supply.
1. INPUT SECTION
Input Circuit and Attenuator I
There are two alternative input sockets. One is a normal B & K coaxial
socket. The other accepts the 7-pin plugs of the B & K preamplifiers and
supplies the ·stabilized voltages for the preamplifiers and a 200 V polarizat-
ion voltage for a condenser microphone.
6

Insert Voltage
.......,
Input Section
Amplifier Amplifier 1
I IT
12V~
100~40V L._j -DC
Power Supply
1/1 and 1/3 Octave
Filters + Weighting
Networks
Fig.2.1. Block Diagram.
HP and LP
Filter
Output Section
I Ampl ifier Amplifier
I IT
Meter
DC~
Output
Overload Indicator
-17o 1o b

Both inputs lead to the Input Attenuator which is divided into two
sections. It steps down the signal in accurate 10 dB steps and can give a
maximum
attenuation of 100 dB. In the least sensitive position it will
accept 300 V input, but if the Gain control is used as well, input voltages of
up to 1000 V may be accepted.
The input impedance of the instrument is 1 MDJ /50 pF.
Reference Oscillator
This is built in for calibration purposes and is selected by push-button. It
produces an accurate 50 mV sine reference voltage at 1000Hz. The signal
distortion is < 2% and the amplitude and frequency stability are also better
than 2%. A frequency of 1000 Hz was chosen since the weighting networks
have no attenuation at this frequency.
Amplifier Sections I
and II
When the 2113 is calibrated for a 50 mV reference deflection Amplifier
Section I has a gain of 50 dB. To obtain the best signal/noise ratio the
amplifier is divided into two sections with an attenuator section in between.
Both sections are similar,
but on the first there is a balanced input stage to
regulate the sensitivity of the two input sockets. The two sensitivity
adjusters and
the Gain Control, have an attenuation range of 14 and 10 dB
respectively and are situated in the same feedback loop. The amplification
of this stage as a whole can thus be varied between 5 and 29 dB. The
amplifier is protected against short circuit of the output (External Filter
Input
socket), and the output impedance is< 10fL
Input Overload Indicator
When using the 2113 in its linear position, overload of the amplifiers can
often be observed on the indicating meter by overdeflection of the needle.
However, when filters or weighting networks are used, certain signal com-
ponents will be attenuated and will not reach the output amplifier or show
up on the meter. It can be seen therefore that while the output amplifier
may not be overloaded the input amplifier may well be. Hence an overload
indicator is placed immediately after the input amplifier, as suggested by the
IEC extension to 179 for Impulse Sound Level meters.
8

The overload indicator is a form of comparator and measures the maxi-
mum peak voltage from zero level. The light will operate when the output
from the input amplifier exceeds 5.6 V peak and it will remain alight for
more than 0.5 sec. even if the peak is of 0.2 msec. duration.
Both the input and output overload indicators are connected in parallel
to the overload relay so that when either one (or both) operates the over-
load relay
is triggered.
2.
FILTER SECTION
Selection
of filters, plus weighting and linear networks is made via the 50
position Filter Switch on the front panel which can be manually or
.. emotely operated. The bold circles on the illuminated Filter Switch scale
indicate the two ranges available. Range I selects filters in the 25 Hz -
20 kHz band, plus D, A, B, C and "Lin" networks, whilst RANGE II utilizes
the same filters, but does not include the weighting or "Lin" positions.
However, positions are available
for the selection of extension filters if
connected. If not connected then no output will be available from the
Spectrometer in these positions. The extension filters cover frequency band
2 - 20 Hz and 25 - 200 kHz, and are available separately from B & K.
Stock numbers for these filters are given in the "Operations" chapter.
Selection of 1/3 or 1/1 octave filters is made via the push button MODE
switch. 1/3 octave centre frequencies are displayed as individual line blocks
on the filter scale, whilst those of the 1/1 octave are indicated by each
central block, of each group of three 1/3 Octave blocks.
1/3 Octave Filters
The Spectrometer is equipped with 30 1/3 oct. band-pass filters. The
filters are all 6-pole Butterworth networks and have centre frequencies and
pass bands in listed
in Table 1.
The filters in the range 25- 200 Hz, included with the Spectrometer and
the extension filters in the 2 - 20 Hz range, not provided, are all active
circuits. Each consists
of 3 stagger tuned Wien-bridge amplifiers, with an
emitter follower output to reduce the output impedance.
The other 1/3 octave filters are all passive 6 pole Butterworth bandpass
9

Centre Centre
Frequency Bandwidth Frequency Bandwidth
Hz Hz Hz
Hz
25 5.8
800
183
31.5 7.3 1000
230
40
9.2 1250
290
50 11.6 1600
370
63 14.5 2000
460
80
18.3 2500
580
100 23 3150
730
125 29 4000
920
160 37
5000
1160
200 46 6300 1450
250 58
8000
1830
315 73 10000
2300
400 92 12500 2900
500 116 16000
3700
630 145 20000
4600
Table 1. Filter centre frequencies and bandwidths.
networks. Schematic diagrams
for both active and passive networks are
shown in Fig.2.2 and 2.3.
The overall characteristics
conform to existing standards for 1/3 octave
filters. Typical examples
of filter characteristics are shown in Fig.2.4 and
2.5.
Peak -valley ripple in the pass band is less than 0.5 dB and rejection is
greater than 75 dB at frequencies five times and one- fifth of the centre
frequency. .
A typical
set of adjacent filter characteristics is shown in Fig.2.6. The
frequency characteristic cannot
be expected to intersect exactly at the
3.7 dB
point because of the steep attenuation slope at this point. A very
small displacement of the filter centre frequency will cause a large displace-
ment
of the point of intersection.
10

o--1
Input
c 201
cr--1
Input
L 201
Ground
Fig.2.2. Active 1/3 Octave filter network.
L 202 p
201 c 202 p 203 L 203
Fig.2.3. Passive 1/3 Octave filter network.
Emitt~r
tollow~r
c 203
Ground
A buffer amplifier is placed directly after the 1/3 octave filters and is
necessary because the filter characteristics are very dependent on terminat-
ing impedance.
1/1 Octave Filters
The 1I1 Octave filters are formed from three adjacent 1/3 Octave filters
and a summing amplifier combines them. In this way interaction between
the combined filters is reduced.
1/1 Octave centre frequencies are given in bold type in Table 1, whilst
Fig.2.7 shows
the top of a typical 1/1 Octave filter response. The peak-
valley ripple
in the pass band is better than -0.5, +1.0 dB, and the reject-
ion outside
the band is greater than 60 dB at eight times and one eighth of
the centre frequency.
11

12
I
.I
I
I
I
I
I
I
I
I
I
I
2 5
\\\
10
\\\
~Typical B&K
\ 1/3 Octave Fi Iter
\
\' ANSI
, class
'-~
20
Freguency Hz
-17o1 :l.'J
Fig.2.4. Typical 1/3 Octave filter response and tolerances.

c:
0
:;:;
0
::;)
c:
~
~
c
0
1
- 2
- 1
~ 2
10
11
12
13
14
15
16
17
18
19
2ll0.5 0.6 0.7
I
I
I
I
I
I
I
I
I
I
I
I
I
I
~
I
I
I
I
I
I
I
I
I
I
I
I
0.8 .9 1.0
1.2 1.4 1.6 1.8 2.0
Frequency relative to centre frequency tftm
Fig.2.5. Top of typical 1/3 Octave filter resJ;~~se.
O O ODOOOO OD OOODDDOO DDO DD D OOOOOOOOOOD D ODDDDODOOOODOOODODO
QP 11 02
10Hz
Frequency-
100Hz
Fig.2.6. Frequency response of several adjacent 1/3 Octave filters.
13

- 2
-1
0 dB
Alt.nuation max
rippl• {p•ak-to-vall•yJ
15 dB {- 0.5 to+ 1. 0dB)
10
11
12
13
14
15
0.5 0.6 0.7 0.8 0.9 10 11 12 13 14 15 f/ fm 2.0
Typical B&K
Octave Filter
Fig.2.7. Top of typical Octave filter response.
Effective Bandwidth
Table 1 gives the effective bandwidth at the 3.7 dB points. This band-
width is the 1/3 octave and also the effective bandwidth of the filter. See
"Definitions".
When measuring white noise and using an effective bandwidth of 1/3
octave, the RMS voltage from the filter can be in error by ± 0.4 dB, because
of the production tolerances on the filters.
14

Phase Response
Any practical filter produces considerable phase distortion in its pass
band.
The phase characteristics and tolerances for 1/3 and 1/1 Octave filters
for the 2113 are shown in Fig.2.8 and 2.9 respectively.
1/3 Octave Filter
dB
c
0
0
10
::>
c
"'
~
15
20
270°
180°
goo
lag
c oo
0
0
VI
goo
"'0
"'
VI
0
.c
180°
a._
1/ s
1;6
1
/3 octave
Frequency --
163
6+f
Fig.2.8. Phase response of 1/3 Octave filter.
15

I dB
~ 10
c
0
c
::J
c
"'
::( 20
30
630°
540°
360°
180°
log
. ~ oo
~ lead
:0 180°
"'
Ill
c
..c
c. 360°
540°
630°
I((
A
11
t7
fJ
Iff
~f
~
p
Octave Filter
'
~
\
'~
~
~
~
~
I)
"
I
I(
IJ
% "t :Y6 fs
1
16 - o+ 1fs
2/s
4-' s ".{; %octave
Frequen cy - /619Jf
Fig.2.9. Phase response of 1/1 Octave filter.
Filter Transient Response
The filter inputs of the Spectrometer are all connected in parallel to the
input amplifiers output via the RANGE I and II, and FILTEAS INT/EXT
push button selector switches, so that during analysis no switching
transients will be
produced when the filter set is being scanned. The
transient response of the filters is in fact quite complex, depending on the
input waveform as well as the filter type, but normally need not be taken
into account for measurements with the spectrometer.
However, the response of the filters to a suddenly applied sinusoidal
signal has been
determined, and is shown in Fig.2.1 0. The filter output
normally requires approximately 15 periods to stabilize to the steady state.
16

This becomes a problem when making low frequency measurements and
means
that 8 sees from the time of switching on would be required by the
2 Hz filter before its output could reach a steady state.
Input Waveform
11\~~~~\,....--.---
~
Output Waveform
at Fi Iter Centre
Freguency fo
Input Waveform
Output Waveform
at Filter Lower
Bandedge Freguency f1
410 0 71
Fig.2.10. Transient response of filters.
Weighting Networks
A graph showing the characteristics of the A, B, C and D networks is
shown in Fig.2.11. The A, B and C networks comply with the basic shape
and tolerances set by
the IEC Recommendation 179, whilst the more
recently introduced D network complies to the DIN and the proposed IEC
standards. Tolerances for these networks are given in Table 2.
llhC aiiOfl Typ1cJ
lr ~ quen~y cl'lo}oclf fiSt ! cs
I
In I I
/
['.._
L~n ewi8 Hz · 200kHz
1/
Llf'leart8Hz-200kHz
~
[/__.......-
__.-:.V'"
v
~ ::\
l\ ~kH z
""'
/ I
~ ~ I/
B·C~
\
/ 1/ c r
24Hz
Vo
/'
~
\
I /
J/
v
- ~
5 I w I
\'\
\
0 I /
,/'II
v \ ~'\ \
Fig.2. 11. Frequency response of A, 8, C, D and linear scales.
17

Frequenc~
Curve A Curve B Curve C
CurveD
IEC Tolerance
B&K Tolerance
Hz
dB
dB
dB
dB
limits dB
limits
dB
10 -70.4 -
38 .2 - 14.3
-27.6
5 -
00
3 - 3
12.5
-63.4
- 33 .2 - 11 .2
-25.6
5 -
00
3 - 3
16 -56.7 - 28.5 -8.5
-23.5
5
-00
3 - 3
20 -50.5 -24.2 -6.2
-21 .6 5 - 5 3 - 3
25 -44.7
-20.4
-4.4 -19.6 5
-5
2 - 2
31.5 -
39.4
-17 .1 -3.0 -17.6 3 - 3 1.5 - 1.5
40 - 34 .6 - 14.2 -2.0 - 15.6 3 -3
1.5 -1.5
50 - 30 .2 -11 .6 -1.3 -13.6 3
-3
1.5 -1.5
63 -26.2 -9.3 -0.8 -
11 .6 3 - 3 1.5 -1.5
80
-22.5 -7.4 -0.5 -9.6 2 - 2 1.5 -1.5
100 -19.1 -5.6 -0.3 -7.8 1 -1 1 - 1
125 -16.1 -4.2 -0.2 -
6.0
1 -1 1 - 1
160 -13.4 -3.0 -0.1 -4.4 1 -1 1 - 1
200
-10.9 -2.0 0 -3.1 1 -1 1 - 1
250 - 8.6 -1.3 0 -1.9 1 -1 1 - 1
315
- 6.6 -0.8 0 -1.0 1 -1 1 - 1
400 - 4.8
-0.5 0 -0.3 1 -1 1 - 1
500
-3.2 -0.3 0 0 1 - 1 1 - 1
630
-1.9 -
0.1
0 -0.1 1 -1 1 - 1
800 - 0.8 0 0 -
0.4 1 - 1 1 - 1
1000 0 0 0 0 1 -1 1 - 1
1250 0.6 0 0 1.9 1 -1 1 - 1
1600 1.0 0 -0.1 5.4 1 - 1 1 - 1
2000 1.2 -0.1 -0.2
8.0 1 -1 1 - 1
2500 1.3 -0.2 -0.3 10.0 1 -1 1 - 1
3150
1.2 -0.4 -0.5
11 .0 1 -1 1 - 1
4000 1.0 - 0.7 -0.8 10.9 1 -1 1 - 1
5000
0.5 -1.2 -1.3 10.0 1.5 -1.5 1.5 -1.5
6300
-0.1 -1.9 -2.0
8.5 1.5 -2 1.5 - 2
8000
-1.1 -2.9 -3.0 6.0 1.5 -3 1.5 - 3
10000 -2.5 -4.3 -
4.4
3.0 2
-4
2 - 4
12500 -
4.3 - 6.1
-6.2 -0.4 3
-6
3 - 6
16000 -6.6 -
8.4
-8.5 -
4.4
3
-00
3 -
00
20000 -9.3 -
11 .1 -
11 .2 -8.1 3
-00
3
-00
Table 2. Weighting network tolerances.
18