R&S SMU200A Test Procedures
1007.9845.82 1.17
E-12
Wideband Noise
Test assembly Connect spectrum analyzer to RF socket of the DUT.
Test method Operate the spectrum analyzer in the CHANNEL POWER mode.
Set SPAN to 400 kHz, channel bandwidth to 350 kHz, detector to RMS
and
MIXER LVL MANUAL to –5 dBm. For reference level measurement,
set the center frequency of the analyzer to the test frequency, set the
reference level to the test level, and read the channel power. Then
measure the noise level by shifting the center frequency of the analyzer
by the desired offset (for example 5 MHz). The channel power now
displayed must be converted to 1 Hz bandwidth and referred to the
reference level. By performing a measurement without input signal, input
terminated with 50 Ohm, the analyzer’s inherent noise can be determined
and subtracted if necessary.
Measurement
Determine the channel power with the center frequency of th
e
anal
yzer set to the test frequency and note it down as P
ref
.
Increase the center frequency by the offset (e.g. 5 MHz).
Inhibit the switching of the attenuator with AMPT RF ATTEN
MANUAL without entering a value so that the input mixer is not
overdriv
en.
Low
er the reference level of the analyzer by 20 dB, read the new
channel power P
noise
and note it down.
Minimize the output level on the DUT by means of RF OFF, r
ead the
c
hannel power P
res
and note it down.
Evaluation
If the power P
res
is lower than P
noise
by more than 0.41 dB and less
than 10 dB, the inherent noise power of the analyzer c
an be
s
ubtracted. To this effect, convert the two power values into mW
according to the formula PmW = 10
(PdBm/10)
, then subtract P
res
from
P
noise
. Reconvert the corrected power into dBm according to the
f
ormula PdBm = 10*log10(PmW).
If the power P
res
is less than 0.41 dB below the power P
noise
, the
anal
yzer resolution is not sufficient for a precise measurement. Th
e
t
rue result is in such case certainly more than 10 dB below t
he
m
easured value. If P
res
is more than 10 dB below P
noise
, P
noise
need
not
be corrected since the noise component of the analyzer is
negligible (<0.41 dB).
Convert the power P
noise
to 1 Hz bandwidth according to the formula
P
1Hz = Pch - (10*log10(channel bandwidth))
.
T
he difference between the (possibly corrected) power P
noise
in dBm
and the power P
ref
in dBm is the broadband noise floor in dBc.
Recommended test
frequencies
See Table 1-4.
Recommended test levels P
min
with unmodulated carrier and P
min
with I/Q modulation at full-
scale level (0.5 V DC at I or Q input)
To Next Page
Loading...
