where:
uted by the transformers,
the
results
obtained
using
a
transformer hardly resemble
the
ideal
theoretical results
as
f2
is the setting
of
the HF
Rolloff
Control,
and
to
use noise
figure contours
and
amplitude transfer
curves
ENB
is the equivalent
noise
bandwidth
in
Hz,
predicted in
the example. When ug
a transformer, it
is best Q
ft
is
the setting
of
the LF Rolloff
Control.
obtained empirically for
the
individual
transformer.
Sub
section
3.11
contains such data
for
the transformers.
If
the
two frequency
settings
are
far
apart,
the
equivalent
noise
bandwidth
to a
close approximation
is simply
1r/2
33
GROUNDiNG
times
the HF rolloff
frequency.
'
_
The Model
113
is specifically
desi ned to have
a hi
h
de ree
Sup.hOSe_
one
_ihtehded
tO_ Operate
the Modei
its
ih
of
immunity
to the effects
of
gtiound-loop
curretnts.
itfhe
con
unction
with
t h
.
.
-
. .
tm ted
t
50 in expegmeh
t att htisehts
ahsoutlge
signal ground
side of
the
input
connectors
is connected
to
. p a _
O .0
ms’ ah
Suppose i‘
'S_ h_OWh
t at t e
chassis
ground
through a
10 ohm
resistor.
To maintain
information
re
u edf
th
d b l
. . . .
.
q
ir rom
e
experiment
is situate
e ow
_
_
ground-loop
immunity,
avoid
shorting the input
grounds to
3
S.ihCe.
tits soutge
tthetmet)hotgsegorhttiisutihe
te_tt_he
chassis ground at
the amplifier
end of the cables.
To
further
bo ad
_dOtirie. is,
eheh deg
Oh, ahh
W|i_Ft 'nt
ti amp iliet
reduce
the effects
of ground-loop
currents,
use an input
an Wt
is minimize
ysettihet
e to O
Cohtto
to cable
that
is short
and
has
a
high
conductivity outer
dc and
the HF toiiott
‘?°"‘F'°'
to 3 kH,Z'
DC Coupling
is used‘
conductor. (For
a
discussion of the circuit
desi
n
with
9
h
: . .
.
.
T e source
thetmai
noise ih
this Case
is
respect
to grounding
considerations,
see
Section
IV).
E
=\/4x1.38
>< 10-” X 2.9x
102x 50>< 3
X103X1.57
=
6 i X to-s
V
rms
CAUTION:
The
power
rating of the
10 ohm resistor is
%
i
watt.
Avoid excessive ground-to-ground
current
to avoid
From
Figure
lll-1,
the
noise
figure
for
the Model
113
at
a
burning
out this
reststtm
center frequency
of 1500 Hz
and
a source
resistance
of 50
ohms
is
18
dB. Substituting
these values
into
equation
(3):
when
the Signet
is aepned
to the
input via
a transformer,
totai
equivaient
input
noise
operate the
amplifier single-ended.
That
is, place
one input
Z 61
X to-s
X ibis/20
selector
to the
ground
position, and connect the
trans-
eieev
former
to the other input
(ac couple); connect
one
transformer lead
to the outer
shell of
the BNC connector
Occasionally
the
dc
component
is
not
required,
so that
the
and
eenneet the
other
transformer
ieed to
the Center
hest
heise Dettetmehee
eeh he
8Chie\/ed
With
the
help
0t
eh
conductor
of
the
connector.
Connecting this way
avoids the
input
transformer.
This
is
possible
by taking
advantage
of
nrebiem
Qt etatie
energe
bui|d-up
at tne inputs
due te
no
the
property
of
a transformer
to
multiply
the
source greund
return_
resistance
by
the
square
of
the
turns
ratio
and the voltage
by the ratio.
With atransformer
inserted
between
the
signal
3_4
$|G|\|AL
\/()|_TAGE
AND
GAIN
source
and
the amplifier
input
we can
increase
the
effective
source resistance
to
a
value
that reduces
the noise figure
to
less
than
0.05
dB.
From
Figure
lll-1,
the source resistance
With
the Vetiehie
9eih
eehttei
ih the eeiihtated
hesitieh
should
be
about
1 megohm.
The
transformer
turns
ratio,
ieewii
the hesitieh
et
the geih Seieetet
eeetiteteiy
Sets
the
tor
this inibedanee
inerease, is
W:
t40_ -|-herntai
gain to the indicated
level. Intermediate
levels
of gain
may
source
noise
at the
amplifier
input
is
equal
to the noise be
ehteihed
ieV use
et the
tiheeiibteteei
Vetiehie
Qeih
generated
by
the 50 bhrn
sbtiree
inuitibiied
by the turns
control. If
it
is desired
to accurately
set
an intermediate
ratio. With
a noise
figure
near
zero,
the total equivalent
rms
ievei et
geihl ephiy
eh ihhtit
Sighei to
the ethpiitiet
ihptit
input nbise
is aisb
edtiai to
the nbise
generated
by the 50
and,
using
an ac voltmeter,
proportionally
adjust the output
dhrn sbtiree
nitiitibiied
by
the turns
ratio! I 70
tiv rrns_
signal
level
to
a corresponding
value
between the levels
of
Although
the
numerical
value
of equivalent
input
noise
is
the two
tixed geih
5ettih95 hteeketihg
the
desired
9aih-
much
larger
than
before,
the signal-to-noise
ratio
isisub-
stantially
increased.
This
can
be
seen by
considering
all
of
the transformed
source
signal
voltage
as
appearing
at the The
maximum
Output that
the aihphtiet
Cah
provide is
to
amplifier
input
terminals,
because
the Model
113
input
t/hits pea_k'tO'heak
(through 600
Ohms ihto
ah Opeh
Circuit)"
teeietenee
is mueh
ieteet than
the 1
meeehm
etesenteei
by For
maximum
input voltages,
refer
to
the specifications
and
the
transformer;
the signal-to-noise
ratio
is
equal to the to Suhsectioh
3'i0'
maximum
possible
value:
esig/E
x
antilog NF/20) E esie/E.
In
this
example
the transformer
increases
the
signal-to-noise
ratio
by a
factor
of 8.
3.5 DC
ZERO
ADJUSTMENT
Princeton
Applied
Research
Corporation
manufactures
transformers
for
use in
low level,
low
frequency
applica-
The dc zero adjustment may need to
be
touched Cip as the
tions.
Characteristics
and
details regarding
the use of some
Qeih
is
ehehgeti
ih Dettietliet, it the ediustmeht
is
titst made
of
these
transformers
are
discussed
in Subsection 3.11.
Wheh
0Detetih9
at e i0W
Qeih,
t0ii0Wed
ietet hV ttehstet te
Because
the
transformers
have so
much influence
in
high-Qeih
Opetetiehi it is impetteht
that
the de
Zero be
determining
bandwidth, and because
of the noise contrib- Cheeked,
ahd,
it heeessetV.
teedlusted
as required-
lll-2
where:
ENB
is
the
equivalent
noise
bandwidth
in
Hz,
f2
is
the
setting
of
the
HF
Rolloff
Control,
and
fl
is
the
setting
of
the
LF
Rolloff
Control.
If
the
two
frequency
settings
are far
apart,
the
equivalent
noise
bandwidth
to
a close
approximation
is
simply
1T/2
times
the
HF
rolloff
frequency.
Suppose
one
intended
to
operate
the
Model 113
in
conjunction
with
an
experiment
that
presents
a
source
impedance
of
50
ohms,
and
suppose
it
is
known
that
the
information
required
from
the
experiment
is
situated
below
1 kHz. Since
the
source
thermal
noise
contributing
to
the
total
noise
is
dependent
on
bandwidth,
the
amplifier
bandwidth
is
minimized
by
setting
the
LF
rolloff
control
to
dc
and
the
HF
rolloff
cuntrol
to
3 kHz. Dc
coupling
is
used.
The
source
thermal
noise in
this
case is:
E
=)4
x 1.38 x 10
23
x
2.9
x 10
2
X
50 x 3 x 10
3
x 1.57
=
6.1
x 10-
8
V rms
From
Figure
111-1,
the
noise figure
for
the
Model 113
at
a
center
frequency
of
1500 Hz
and
a
source
resistance
of
50
ohms
is
18 dB.
Substituting
these
values
into
equation
(3):
total
equivalent
rms
input
noise
= 6.1 X 10-
8
X
10
1
8 /
20
= 0.5
pV
rms
Occasionally
the
dc
component
is
not
required,
so
that
the
best noise
performance
can
be
achieved
with
the
help
of
an
input
transformer.
This
is
possible
by
taking
advantage
of
th e
property
of
a
transforme
r
to
multi ply
the
sou rce
resistance
by
the
square
of
the
turns
ratio
and
the
voltage
by
the
ratio.
With a
transformer
inserted
between
the
signal
source
and
the
amplifier
input
we
can
increase
the
effective
source
resistance
to
a val ue
that
reduces
the
noise figure
to
less
than
0.05 dB.
From
Figure
111-1,
the
source
resistance
should
be
about
1
megohm.
The
transformer
turns
ratio,
for
this
impedance
increase,
is
VR2/R
1 = 140.
Thermal
source
noise
at
the
amplifier
input
is
equal
to
the
noise
generated
by
the
50
ohm
source
multiplied
by
the
turns
ratio.
With a noise figure near
zero,
the
total
equivalent
rms
input
noise
is
also
equal
to
the
noise
generated
by
the
50
ohm
source
multiplied
by
the
turns
ratio, = 70
pV
rms.
Although
the
numerical
value
of
equivalent
input
noise
is
much
larger
than
before,
the
signal-to-noise
ratio
is
sub-
stantially
increased.
This
can
be
seen
by
considering
all
of
the
transformed
source
signal voltage as
appearing
at
the
amplifier
input
terminals,
because
the
Model 113
input
resistance
is
much
larger
than
the
1
megohm
presented
by
the
transformer;
the
signal-to-noise
ratio
is
equal
to
the
maxi~um
possible value: esig/E x
antilog
N F
/20)
.::=
esig/E.
In
thiS
example
the
transformer
increases
the
signal-to-noise
ratio
by a
factor
of
8.
Princeton
Applied
Research
Corporation
manufactures
transformers
for
use in low level, low
frequency
applica-
tions.
Characteristics
and
details
regarding
the
use
of
some
of
these
transformers
are
discussed in
Subsection
3.11.
Because
the
transformers
have
so
much
influence
in
determining
bandwidth,
and
because
of
the
noise
contrib-
111-2
uted
by
the
transformers,
the
results
obtained
using a
transformer
hardly
resemble
the
ideal
theoretical
results as
predicted
in
the
example.
When ug a
transformer,
it
is
best
to
use noise figure
contours
and
amplitude
transfer
curves
obtained
empirically
for
the
individual
transformer.
Sub-
section
3.11
contains
such
data
for
the
transformers.
3.3 GROUNDING
The
Model 113
is
specifically
designed
to
have a high
degree
of
immunity
to
the
effects
of
ground-loop
currents.
The
signal
ground
side
of
the
input
connectors
is
connected
to
chassis
ground
through
a 10
ohm
resistor.
To
maintain
ground-loop
immunity,
avoid
shorting
the
input
grounds
to
chassis
ground
at
the
ampl
ifier
end
of
the
cables.
To
further
reduce
the
effects
of
ground-loop
currents,
use
an
input
cable
that
is
short
and
has a high
conductivity
outer
conductor.
(For
a discussion
of
the
circuit
design
with
respect
to
grounding
considerations,
see
Section
IV).
CAUTION:
The
power
rating
of
the
10
ohm
resistor
is
'h
watt.
Avoid excessive
ground-to-ground
current
to
avoid
burning
out
this
resistor.
When
the
signal
is
applied
to
the
input
via a
transformer,
operate
the
ampl ifier single-ended.
That
is, place
one
input
selector
to
the
ground
position,
and
connect
the
trans-
former
to
the
other
input
(ac
couple);
connect
one
transformer
lead
to
the
outer
shell
of
the
BNC
connector
and
connect
the
other
transformer
lead
to
the
center
conductor
of
the
connector.
Connecting
this
way
avoids
the
problem
of
static
charge
build-up
at
the
inputs
due
to
no
ground
return.
3.4 SIGNAL VOLTAGE
AND
GAIN
With
the
variable gain
control
in
the
calibrated
position
(ccw),
the
position
of
the
gain
selector
accurately
sets
the
gain
to
the
indicated
level.
Intermediate
levels
of
gain
may
be
obtained
by use
of
the
uncalibrated
variable gain
control.
If
it
is
desired
to
accurately
set
an
intermediate
level
of
gain,
apply
an
input
signal
to
the
amplifier
input
and,
using an
ac
voltmeter,
proportionally
adjust
the
output
signal level
to
a
corresponding
value
between
the
levels
of
the
two
fixed gain
settings
bracketing
the
desired
gain.
The
maximum
output
that
the
amplifier
can
provide
is
10
volts
peak-to-peak
(through
600
ohms
into
an
open
circuit).
For
maximum
input
Voltages, refer
to
the
specifications
and
to
Subsection
3.10.
3.5
DC
ZERO ADJUSTMENT
The
dc
zero
adjustment
may
need
to
be
touched
up
as
the
gain
is
changed.
In
particular,
if
the
adjustment
is
first
made
when
operating
at
a
low
gain,
followed
later
by
transfer
to
high-gain
operation,
it
is
important
that
the
dc
zero
be
checked,
and,
if
necessary,
readjusted
as
required.
)
To Next Page
Loading...
