Rohde & Schwarz R&S FSP Series, Руководство по эксплуатации

Страница: 45 / 796

Operating Manual 1164.4556.12 - 06 2.21
R&S FSP Getting Started
Noise Measurements
Fig. 2.17 Repeatability of channel power measurements as a function of the number of samples
used for power calculation
The curves in Fig. 2.17 indicates the repeatability obtained with a probability of 95%
and 99% depending on the number of samples used.
The repeatability with 600 samples is
±
0.5 dB. This means that – if the sample
detector and a channel bandwidth over the whole diagram (channel bandwidth =
span) is used – the measured value lies within
±
0.5 dB of the true value with a prob-
ability of 99%.
If the RMS detector is used, the number of samples can be estimated as follows:
Since only uncorrelated samples contribute to the RMS value, the number of sam-
ples can be calculated from the sweep time and the resolution bandwidth.
Samples can be assumed to be uncorrelated if sampling is performed at intervals of
1/RBW. The number of uncorrelated samples (N
decorr
) is calculated as follows:
N
decorr
= SWT × RBW
The number of uncorrelated samples per trace pixel is obtained by dividing N
decorr
by 501 (= pixels per trace).
Example:
At a resolution bandwidth of 30 kHz and a sweep time of 100 ms, 3000 uncorrelated
samples are obtained. If the channel bandwidth is equal to the frequency display
range, i.e. all trace pixels are used for the channel power measurement, a repeat-
ability of 0.2 dB with a confidence level of 99% is the estimate that can be derived
from
Fig. 2.17 .
10 1000 100 10000 100000
0
0.5
1
1.5
2
2.5
3
Number of samples
max. error/dB
95 % Confidence
level
99 % Confidence
level