PRIVATE AND CONFIDENTIAL
RESEARCH DEPARTMENT
A.K.G. ELECTROSTATIC MICROPHONES TYPES C26 AND C28
Report No. L-048
( 196 1/12)
THE BRITISH BROADCASTING CORPORATION
ENGINEERING DIVISION
RESEARCH DEPARTMENT
A.K.G, ELECTROSTATIC MICROPHONES TYPES C26 AND C28
Report No. L-048
( 196 1/12)
H.D. Harwood, B.Sc, A.Inst.P,, A.M. I.E. E, / X^^ 4 ^'
J.R. Chew, B.Sc. (Eng. ), Grad. I.E.E, *~t^
R.L. Deane
R.J, Packer (T. Somerville)
This
Report
is the prop
e r t y
of
the
Brltis
h
Broad
cas
ting
Corporation
and
may
not b
e
r ep r o du
c e d
or disclose
d t o a
third
P
arty
in
any
f o rn
without
the
written
p erm!
s s
ion of the
Corpor at J
on .
Report No. L-048
A.KoG„ ELECTROSTATIC MICROPHONES TYPES C26 AND C28
Section Title Page
SUMMARY ............................... 1
1 INTRODUCTION 1
2 DESCRIPTION 2
<j a JL. a OcHSrSX • •eaaaoaoaooeaeseasesaaaasas <j
2.3. Type CK26 Capsule ....................... 3
2.4. Type CK28 Capsule ....................... 3
2. 5. Amplifier and Cables ... ............. 4
3 PERFORMANCE .............................. 5
3.1. Method of Measurement ..................... 5
3.2. Frequency Characteristics ................... 6
3.3. Impedance ................ ...... 7
3.4. Sensitivity .................. 9
4 NOISE ........... ........ ........ ..... 9
4.1. Internally Generated Noise .... ....... 9
4. 2. Interference from Magnetic Fields ........ 10
4.3. Interference from Radio-Frequency Signals 10
4.4. Wind Noise ...... ......... . 10
4. 5. Interpretation of Noise Measurements ..... 11
4.6. Listening Tests ........ ... 11
5 CONCLUSIONS . 11
o REFERENCE ••••••••#••••••»•••••**•■»*•••* -L^
PRIVATE AND CONFIDENTIAL
July 1961
Report No. L-048
( 1961/12)
A.K.G. ELECTROSTATIC MICROPHONES TYPES C26 AND C28
SUMMARY
This report describes tests on two electrostatic microphones manufactured by
the Akustische und Kino-Gerate Gesellschaft (A.K.G. ). The first, type C26, has
omnidirectional and the second, type C28, cardioid characteristics. The microphones
are of small dimensions and provision is made in each case for introducing an extension
piece between the capsule and the head amplifier. Two types of windshield are
provided; one can be fitted over the end of the microphone itself, whilst the other
is designed to protect the microphone capsule when an extension piece is used.
Measurements have been made of the microphone frequency characteristics with
and without the various fittings, and the degree of interference from wind and from
magnetic fields has also been determined. The quality of reproduction obtained,
particularly from the type C28, is good and the degree of interference is low. The
susceptibility to radio-frequency fields has also been measured and a method of
reducing it has been indicated.
1. INTRODUCTION
The Akustische und Kino-Gerate Gesellschaft of Vienna has introduced two new
electrostatic microphones, type C86 which has a nominally omnidirectional character-
istic and type C28 with a cardioid characteristic. The same head amplifiers, mains
units and interconnecting leads are used in both types, the only differences being in
the internal construction of the capsules types CK26 and CK28, which have identical
external dimensions. The capsule in each case is surrounded by a layer of fine wire
gauze which forms part of the microphone case and affords an appreciable measure of
protection against interference from wind. For more severe conditions a windshield,*
type W28, is provided, which fits over the microphone head. To make the instrument
less conspicuous, the portion of the case surrounding the capsule can be removed and
slender extension pieces, types C29 or C30 of differing lengths, can be inserted
between the capsule and head amplifier so that the latter may be concealed; when
either of these extension pieces is used, a different windshield, type W17, is
normally employed to protect the capsule.
The type C28 has potential applications in places such as footlights where
an inconspicuous microphone having a high degree of suppression of sound from the rear
This windshield, which bears no type number, is described by the maker as type W26 or type
W28 , according to the microphone with which it is supplied. Throughout this report,
however, it will be referred to as the type V28 .
is required. When the type C30 extension piece and the type W17 windshield are
employed, the height is suitable and the size small enough to enable the microphone to
be used on the stage by crooners.
The type C26 microphone might be of interest as a replacement for the
Philips type EL3921/00 microphone which is now obsolete.
Two experimental models of the type C28 microphone were examined in June
1957 and another pair, incorporating improvements, in March 1958. Some minor
modifications were suggested and production units were received in May 1958. At the
request of S.B.S.B. 's Department, two type CK26 capsules were tested in May 1959.
The price to the B.B.C. is £90 for either the type C26 or the type C28.
The approximate cost of the type G29 extension piece is £4, of the type C30 £4. 10s. Od. ,
and of the type W17 and type W28 windshields £4. 5s. Od. and £7. 13s. Od. respectively.
A case is available to house all these items and costs an additional £8.
2. DESCRIPTION
2. 1. General
Fig. 1 gives an external view and dimensions of the microphone, a capsule
type CK28, the two extension pieces types C29 and C30, and the windshields types W17
and W28; as previously mentioned, the type C26 instrument is externally identical.
The windshields are designed to enclose the entire capsule; they are constructed of
two layers of wire gauze sprayed with flock on the inside and spaced about 1/8 in.
(3 mm) apart. This form of construction is quite effective but if the flock became
wet it is doubtful whether the flow resistance would return, on drying, to its
previous value.
Fig. I-A.K.G. microphone type C28 with extension pieces types C29 and
C30 and windshields types WI7 and W28. External view and dimensions
2.2. Vfeight
The weight of either microphone without cable is 4 lb (0*2 kg) and of the
mains unit 9 lb (4 kg).
2.3. Type CK26 Capsule
The construction of the pressure type capsule CK26 is on conventional lines.
The size is similar to that of the Philips microphone type EL3921/00 and the direc-
tional properties are therefore similar.
2.4. Type CK28 Capsule
The design of the type CK28 capsule, which is the subject of a patent, is
novel in that two apertures spaced along the capsule length are provided instead of
the usual one; these communicate with different acoustic networks, each designed to
operate over a different part of the frequency range. Fig. 2 shows the main con-
structional details. Sound can reach the rear of the diaphragm either through
aperture 1 and network la, lb, lc and Id, or through apertures 2 and network 2a, 2b,
2c, 2d, 2e and 2f . It is claimed that each of the elements forming these networks
may be adjusted separately to give the desired directional characteristics. This
|C lb la APERTURES /GAUZE
V
Fig. 2 - A.K.G. microphone type C28. Constructional details of capsule type CK28
arrangement has advantages over the usual construction since if only one rear sound
aperture is employed it is difficult to design a microphone which maintains a constant
cardioid characteristic over the vhole audio-frequency range.
In the prototype CK28 capsules, holes were formed in the plate P and these,
together with the enclosed volume of air, V, affected the phase of the sound pressure
at apertures 1 and 2 and therefore the directional characteristics. As this fact had
not been appreciated by the designer, and as the volume V in the different extension
pieces had not been closely controlled in production, the directional characteristic
was dependent on the particular fitting used. In the production models, however,
this variation was prevented by eliminating the holes in the plate P. The type C28
response curves which are given later in the report all refer to capsules modified in
this way.
2.5. Amplifier and Cables
Fig. 3 shows the electrical circuit diagram of the head amplifier. The
transformer may be connected to give a nominal output impedance of 50 or 200 ohms;
the manufacturer suggests that the change from one output impedance to the other
should only be undertaken in the factory.
The microphone is connected by a plug and 65 ft (20 m) of cable to the mains
unit type N28K, from which the audio-frequency output is taken through a 3-pin plug
and socket. In an emergency a faulty mains unit may be replaced by one of the type
N12K normally used for the A.K.G. type C12 microphone.
h.t.+ I •>
I I
LT +
OUTPUT ' •)^- L -
TAG
STRIP
lOOk
-AAAA-
:sook
8M
"-WV-f
S00k<
0-005
500V
607?
2
f^R
I80M
l-8k
JIS
1-5
125V
CONNECTIONS
FOR son
OUTPUT
IMPEDANCE
TI4/I
lOO p
SOOV
Fig. 3 - A.K.G. microphone type C28. Circuit diagram of head amplifier
3. PERFORMANCE
3. 1. Method of Measurement
All the frequency characteristics of the type C26 microphone and those of
the type C28 microphone above 200 c/s were measured by comparison with a pressure
standard in a dead room. The characteristics of the type C28 microphone at frequen-
cies below 200 c/s were measured in a travelling-wave duct by comparison with the same
standard; the characteristics of the type C28 microphone with the type C29 extension
piece were obtained in a similar manner. Generally the accuracy of comparison is
+ i 63; errors of ± 1 dB are, however, possible in the measurements of the types C28
and C29 instruments for sound incident at angles greater than 90 .
All the characteristics given in this section were measured with the
transformer connected to give a nominal output impedance of 200 ohms.
4l»0»
=S<
.•
<&•_
}\\
4-90
• +135'
i
-•• i
/
+ l»0"
\\\+90"
C26
CAPSULE 311
1 1 1
+ ,/s'\
10* 10 s 10*
Fig. H - A.K.G. microphone type C26. Frequency characteristics measured
200-ohm connection for sound incident at various angles
-50 §
o
a.
u
■55 |
o
-60 £
>•
a
H
■65 5
-70 a
x
i-
*
-75 v
at
,
r
~*
L
s
/
N
i
"
— -
\
\
\
\
\
CAP!
>ULE
l 1
. 3
O
1—
-50
■55
-60
z
s
H t-
J —
O =
> o
- a
g -o
u; z
tu UJ
a.
-65 f
S
CD*
■70
FREOUENCY, c/t
Fig. 5 - A.K.G. microphone type C26. Axial frequency characteristic
measured at 200-ohm connection
3.2. Frequency Characteristics
Pig. 4 shows the open-circuit frequency characteristics of a type C26
microphone for sound incident at various angles, and Fig. 5 the axial response when
fitted with another type CK26 capsule. For comparison, Fig. 6, taken from Research
Department Technical Memorandum M. 1008, shows the corresponding curves for a Philips
microphone type EL3921/00. If the windshield type W28 is fitted to the type C26
microphone the overall frequency response is not changed by more than ± 1 dB. When
the microphone capsule is mounted on either the type C29 or the type C30 extension
piece so that the smaller windshield type W17 has to be used, the change in response,
shown in Fig. 7, is slightly greater.
Fig. 8 shows the open— circuit frequency characteristics of a type C28
microphone for sound incident at various angles, and Fig. 9 the difference in axial
frequency response of three capsules. The change in response caused by the type W28
windshield is remarkably small for a microphone with a cardioid polar characteristic
and is not more than ± i dB for sound incident at 0° and 90 and 2 dB at 180 .
Fig. 10 shows the response of the capsule referred to in Fig. 8 when mounted on the
type C29 extension piece; when the longer extension piece type C30 is employed, the
same frequency response is obtained. Fig. 11 shows the response of a microphone with
and without the type W17 windshield; it will be noted that although the shield does
-60
-65
-70
CO
-80
' ^*\ JQ°
S
s.
*
*£
•-'
N\+'0«
-^- _
*«*J
I —
'">
+l(
^
+ 90'
n' + pun"
X J
+I55« \
CAPSULE 129
EL
3921/C
>o
Fig. 6 - Philips microphone type EL392I/00. Frequency characteristics
measured at 200-ohm connection for sound incident at various angles
+ 180*
5
_^ ;
'->
t
1
+90*
V
^
s.
i/\~v
1
180'
V
/ + 90°
(_) .a 4
\J
WI7 ON
:26/3C
)
-
10* 10" 10'
FREOUENCY,c/s
Fig. 7 - A.K.G. microphone type C26/C30. Change in response produced by
windshield type WI7 for sound incident at various angles
not greatly change the axial frequency response, it does appreciably degrade the
directional characteristics at low frequencies.
The above figures refer to early production models. However, measurements
made on current microphones show a loss at low frequencies, varying from —5 dB to
-10 dB at 50 c/s, much greater than that shown in Fig. 9. In addition there is some
evidence that under service conditions the bass response decreases still further with
time. This matter is being further investigated.
3.3. Impedance
The mid— band output impedance of the microphone tested, when using the
200 ohms transformer connection, lies between 160 ohms and 240 ohms and is thus
s
•
^
#
7
-N —
\
V
+ 90*
^>
•
i
V
i
\
y
» *
y
i
•""
\
•-
"-•.
V
"*
— ■ —
, m
+ l»0«
,*>
-~Z__.
— ■--
wm ^^-
N
/
i
i
i
+90"
1
\
t
,
r
1
o. 1
+ 180°
\
\ r
CAPSULE 2
1 1 1
C28
v
10'
-50
-35
-60
-65
-70
-75
-80
s
o
a.
UJ
a
x
Fig. 8 - A.K.6. microphone type C28. Frequency characteristics measured at
200-ohm connection for sound incident at various angles
/
\
■ *
7
** *
^jS'_
\CAPSUIE
0. 24 '
*
*
"
s.
*>""
CAPSULE
2'
^A
f
s
'
CAPSULI
241 -^1
/
/
/
-5
-10
I0 2 |o* I0"
FREQUENCY, c/s
Fig. 9 - A.K.G. microphone type C28. Axial frequency measured at 200-ohm
connection for three capsules
0°
i
+ 90'
_.
.-
-
--L-.
_- -""
»^""~ "
+ 90°
C28/C29
\
'
X
\
' '
"' — .
..
-*""
~\
i
1
\
"
— ..
■■• ,,,
+ 180°
*
\
1
1
1
\
\
V
CAPSULE 2
1 1 1
'
-50
3
-55
-«0 E
>
-70 i
u-
UJ
tc
X
1-
-75 i
I0 2 10" 10*
Fig. 10 - A.K.G. microphone type C28/C29. Frequency characteristics
measured at 200-ohm connection for sound incident at various angles
-»o
*l
90*
A
/
**\
^
'
■*•;
" s «
\ t
S'
■ ■*;~j2'-~^ > '
■— »»^
■-^
+-
-■T
^
#
'V
u
S
<""
+90°
1 +180°
C28/C29
if
'W
■ Tfc
»-'
oi
i
*
J
f\
*'
^
V
i *
X
/,
(
•
t
I
i
X.
+ 180"
V'
*
*-».
•
<-..,
*-/*'•
,
J
-^...
„.,_
>
<'
CAPSULE 249
- ** —
- —
--"
I0 1
■50
-55
■<o £
o
"e
U
-« 5
o
-70 >
-75
-80
FREQUENCY, e/»
Fig. II - A.K.G. microphone type C28/C29. Frequency characteristics
measured at 200-ohm connection for sound incident at various angles
showing effect of windshield
Wi thout wi ndsh i el d
90 ^
1S0 C
With W17 windshield
I I I I I
outside the maker's tolerance of + 10$; the value of this impedance is affected to
some extent by feedback through the grid-anode capacitance of the valve, which forms a
potential divider with the capacitance of the particular capsule and extension piece
in use. The impedance is, however, substantially constant over the whole audio-
frequency band so that the on-load frequency characteristics do not differ signifi-
cantly from those shown.
3.4. Sensitivity
The high tension supply from the mains unit is unstabilised and as a result
the capsule polarising potential, and therefore the microphone sensitivity, changes
with the voltage of the mains supply. For nominal mains voltage the open— circuit
sensitivity of the type C26 microphone in the mid-band region is -56 dB with reference
to 1 volt/dyne/cm 2 when using capsule No. 310 and -57 dB for capsule No. 311; the
maker's figure is stated to be "about 1-3 mV", i.e. -57«5 dB.
Four type CK28 capsules were tested in the same head amplifier; the
microphone open-circuit sensitivity lay between -54 dB and -55 dB with reference to
1 volt/dyne/cm 2 , the maker's figure again specifying approximately -57*5 dB. When
the type C29 and C30 extension pieces are used, the capacitance-to-earth of the
additional length of lead involved shunts the signal from the capsule, and the
sensitivity of the microphone is thereby reduced. The loss for the type C29
extension piece is 1 dB and for the type C30 approximately 3 dB; the maker does not
appear to have appreciated this effect as the same sensitivity is quoted for all
three conditions.
4. NOISI
4.1. Internally Generated Noise
The internally generated noise appearing at the output of the microphone is
a combination of flicker effect in the valve and of thermal agitation in the resistive
component of the grid circuit impedance.
The open-circuit noise when weighted by an aural sensitivity network type
ASN/3 is -105 dB with reference to 1 volt both for the type G86 and for the type C28.
The mid-band sound pressures required to give the same output levels are +25 dB and
+23 dB respectively with reference to 0*0002 dyne/cm 2 ; if the rise in axial response
of the type C26 microphone at high frequencies were equalised, the noise level would
be somewhat reduced. For comparison, the weighted noise figure for the Philips type
EL3921/00 microphone is +27 dB.
It appears that the noise figure for the types C26 and C28 microphones is
not as good as it could be, because the head amplifier employs a triode operating
under conditions of high gain producing a large effective input capacitance — about
65 I4J&; as the type CK28 capsule has a capacitance of about 25 jj/JF, the signal at the
valve grid is reduced to some 11 dB below the open-circuit output, resulting in an
appreciable reduction in signal-to-noise ratio. The level delivered by the amplifier
is too high to mix directly with that from most other types of microphone used in the
Corporation; hence an attenuator of about 15 dB, designated AT. 2/6 and provided by
10
Equipment Department, is normally fitted in series with the output. It seems
probable that if, instead of using external attenuation, the amplifier gain, and thus
the input capacitance, were reduced, the same effect would be achieved but with a
better signal— to-noise ratio.
4.2. Interference from Magnetic Fields
Measurements were made of the maximum open— circuit voltage induced in the
microphone by a uniform magnetic field. The unweighted mid-band sound levels, with
reference to 0»0002 dyne/ cm 2 , required to give an output equivalent to that produced
by uniform fields of 1 milligauss at 50 c/s, 1 kc/s and 10 kc/s are +1 dB, +13 dB and
+23 dB for the type C26 microphone using capsule No. 311, and -2 dB, +10 dB and +20 dB
for the type C28 using capsule No. 2. These levels are regarded as extremely low and
should cause no trouble under normal studio conditions.
4.3. Interference from Radio-Frequency Signals
As with many electrostatic microphones, radio— frequency signals may be
picked up and demodulated in the head amplifier. Although the manufacturer has
endeavoured to reduce this effect to a minimum, considerable interference has been
experienced on some occasions. An investigation carried out by Research Department 1
revealed the means by which the radio— frequency currents enter the amplifier screening
and by paying careful attention to the method of earthing the microphone case the
interference has been very considerably reduced. Details of the modification have
already been issued as a memorandum by S.E.S.B. on 9th June 1960. It is carried out
by connecting together the two tags shown in Fig. 12. However, when these micro-
phones are to be used in places where appreciable radio— frequency fields are known to
exist, tests should still be made in advance to ensure that interference is not likely
to be troublesome.
Fig. 12 - Wiring modification to
reduce radio-frequency
interference
OLD CONNECTIONS
■ NEW CONNECTIONS
4.4. Wind Noise
Measurements were made of the wind noise generated when the microphone was
placed at various angles to a streamlined air flow of 10 m.p.h. both for the normal
condition and when equalised so that the axial response was uniform from 1 kc/s down
to 50 c/s. The open-circuit noise was weighted by the standard A.S.A.* network and
measured by a V.U. meter; the results are given in the following table in terms of
the level, with respect to 0*0002 dyne/cm 2 , of a 1 kc/s tone calculated to give an
equal signal.
American Standards issoeiation, Standard Z.24.3
Noise and Other Soands".
1944, "8onnd Level Meters for Measurement or
11
0°
45°
90°
135°
180°
dB
dB
dB
dB
dB
92
85
85
96
83
102
96
97
107
96
77
61
59
79
64
84
77
75
98
81
106
100
98
96
87
117
112
111
108
101
70
69
70
71
76
85
86
88
91
90
TABLE 1
Microphone condition Angle
Normal C28: Capsule No. 216. Amplifier No. 352
IMequalised
Equalised
With W28 windshield, unequalised
With W28 windshield, equalised
Bare on G89 extension, unequalised
Bare on C29 extension, equalised
On C29 extension with W17 windshield, unequalised
On C29 extension with W17 windshield, equalised
Normal C26: Capsule No. 311. Amplifier No. 352
Unequalised
Equalised
With W28 windshield, unequalised
With W28 windshield, equalised
Bare on C29 extension, unequalised
Bare on C29 extension, equalised
On C29 extension with W17 windshield, unequalised
On C29 extension with W17 windshield, equalised
The wind noise levels with the types W17 and W28 windshields measured without bass
equalisation are low for both types of microphone.
4.5. Interpretation of Noise Measurements
In applying these results it should be remembered that the aural sensitivity
weighting where used is intended to give an indication only of the loudness of the
noise. The subjective assessment of the annoyance caused depends on such factors as
the degree to which it may blend with the studio "atmosphere" and other background
noises.
4.6. Listening Tests
Listening tests were carried out on speech from non— reverberant surroun-
dings. The results obtained were in agreement with the objective tests.
5. CONCLUSIONS
The type C26 microphone appears to be a reasonable replacement for the
Philips type EL3921/00. The electrical noise is a little lower and the microphone is
slightly less directional at high frequencies; as with the Philips microphone, a
certain amount of equalisation is desirable.
50
47
45
50
51
49
43
43
46
49
52
38
38
37
37
50
38
38
37
37
76
63
66
60
53
71
58
57
60
51
55
40
40
42
48
51
38
36
39
45
12
The axial frequency response of the type C28 microphone is free from serious
irregularities and is well maintained over the frequency band. The directional
characteristic is also nearly independent of frequency. The electrical noise is low
but could probably be made even lower by a modification of the head amplifier.
The facility of mounting the type CK28 capsule on one of the extension
pieces types C29 or C30 has proved useful on occasion. The windshields provided are
effective with both the types CK26 and CK28 capsules and produce appreciably less
degradation in the frequency characteristics than is usual with commercial shields.
6. REFIRENCI
1. B.B.C. Research Department Report No. L-047, "Radio-Frequency Interference
in Electrostatic Microphones".
MM
Printed by B.B.C. Research Department, Kings wo od Warr en, Tadwortn, Surrey
