Rohde & Schwarz R&S FSP Series, Operating Manual

Page: 55 / 796

Operating Manual 1164.4556.12 - 06 2.31
R&S FSP Getting Started
Measurements on Modulated Signals
5. Set the optimal reference level and RF attenuation for the applied signal
level.
➢
Press the ADJUST REF LVL softkey.
The R&S FSP sets the optimal RF attenuation and the reference level based
on the transmission channel power to obtain the maximum dynamic range.
The following figure shows the result of the measurement.
Fig. 2-1 Adjacent channel power measurement on a CDMA IS95 signal
The repeatability of the results, especially in the narrow adjacent channels,
strongly depends on the measurement time since the dwell time within the 30
kHz channels is only a fraction of the complete sweep time. A longer sweep
time may increase the probability that the measured value converges to the
true value of the adjacent channel power, but this increases measurement
time.
To avoid long measurement times, the R&S FSP measures the adjacent
channel power in the time domain (FAST ACP). In the FAST ACP mode, the
R&S FSP measures the power of each channel at the defined channel
bandwidth, while being tuned to the center frequency of the channel in
question. The digital implementation of the resolution bandwidths makes it
possible to select a filter characteristics that is precisely tailored to the signal.
In case of CDMA IS95, the power in the useful channel is measured with a
bandwidth of 1.23 MHz and that of the adjacent channels with a bandwidth of
30 kHz. Therefore the R&S FSP jumps from one channel to the other and
measures the power at a bandwidth of 1.23 MHz or 30 kHz using the RMS
detector. The measurement time per channel is set with the sweep time. It is
equal to the selected measurement time divided by the selected number of
channels. The five channels from the above example and the sweep time of
100 ms give a measurement time per channel of 20 ms.
Compared to the measurement time per channel given by the span (= 5 MHz)
and sweep time (= 100 ms, equal to 0.600 ms per 30 kHz channel) used in the
example, this is a far longer dwell time on the adjacent channels (factor of 12).
In terms of the number of uncorrelated samples this means 20000/33 µs = 606
samples per channel measurement compared to 600/33µs = 12.5 samples per
channel measurement.
Repeatability with a confidence level of 95% is increased from
±
1.4 dB to
±
0.38 dB as shown in Fig. 2.17 . For the same repeatability, the sweep time
would have to be set to 1.2 s with the integration method. The following figure
shows the standard deviation of the results as a function of the sweep time.