Advertisement
may be operattd is small
,
being
determined
solely by variations (with
frequency)
in the signal so
rce
output
power.
Since
the range of variation
in the level of
power into the ncident detector is
comparatively small (approximately
10 db when
the signal source is a Model 670HM oscillator)
,
it
is practical
to attenuate the
incident voltage ahead
.
of
the
detector"
To provide this desirable attenuation,
generally a 20-db directional
coupler is used in
the incident channel.
Because of the
wide
dynami
c
range
over
which
the reflected detector may
be operated (a possible
range
of 40 db
,
with
100% reflection
at one end of the range
.and the possibility of a
reflection
as low as
1
% at the other end), much attenuation
ahead of the reflected detector is not p
r
a
c
tical
,
since the minimum voltage into
the Model 416A must be at
least
3 mic rovolts.
By choosing
a signal source with
sufficiently high output power
,
howeve
r,
a directional coupler which offers 10 db
attenuation can be tolerated.
NOTE: With a 20
-
db
coupler
in
the incident channel
and a 10
-
db coupler
in the
reflected channel,
the voltage in the incident channel will
be down 10 db from
the level it
should have to maintain a true
ratio.
By setting
the
Model 416A EXCESS COUPLER LOSS switch
at 10 db
,
a 10-db gain
is
provided
in the incident input
circuit to
bring
the
incident voltage
up
to the
proper level.
Po
s
sible
Errors
-
a.
If the
level
of
power
into
the
detector
c
annot be kept
low
enough to
operate the detecto
r
in
the linear
regi
on of
its
input/output character-
istic, the
resulting
departur
e fr
orn a
true
squa
re
-law
characteristic
will be the sour
ce
of
two types of possible
error.
(1) In order to
calibrate
a
r
a
t
io mete
r,
it is
necessary to assume that
the
detector
follows
some
fixed
detec
tion
law.
The
Model 416A
has been calibrated on
a
square
-law
basis.
Where
the
voltages
out
of
the detectors do not
r etain a
square-law
r
elation to
theil-
in-
put
y
oltag
es,,,
the ratio n:1
e
t
e.r c
annot
indicate
the true
ratio
of re
-
fleeted
-
to
-
incident power.
The magnitude
of
this error,
of course
,
is a
function
of
the amount
of
deviation from a true
square
-
law
characteristic• and at inputs up
to -
5 dbm
will
not be great.
(2)
There
is another facto
r
with
re
spe
ct to
dev
i
ation
from
square
-law
that
c
an
c
ontribute gr
e
ate
r
error
than lack of tracking
between
meter
calibration and the proportional character
of the input volt-
ages. This fa cto
r
is
lac
k of
tracking
betwe
e
n
the
two detectors be-
cause
t
he
ir
d eviations
from square -
law are different.
If the deviation-
fro
rn-square
-
law
c
ha
r
acteristic of the
two detectors were identical,
a
true ratio would
be
maintained
,
and
no appre
c
iable error would be
intro
duced by
this factor.
Diffe
re
nces in
crystal characteristics
exist,
howeve
r,
and
lac
k of
tracking
between the two detectors in
the nonl
i
near
region of
the
ir input/output characteristic is to be
expected.
=3
1
-
Advertisement
