In the app notes for spectrum analyzers there is discussion about the 2.3dB correction that is needed when measuring noise to account for the different statistics of sinusoids vs. noise due to the log amp, iIF filter, and something else which I've forgotten.
Is there a similar issue when using a power meter with standard diode sensor heads, like the E series. If there is, is there an app note on that or is it the same as with spectrum analyzers.
Thanks.
Chuck
Is there a similar issue when using a power meter with standard diode sensor heads, like the E series. If there is, is there an app note on that or is it the same as with spectrum analyzers.
Thanks.
Chuck
a noise measurement has a lot in common with a digitally modulated RF signal (like CDMA). Newer sensors like the E9300-series were designed to measure these types of signals correctly, with no correction factor required.
This note has a good explanation:
http://cp.literature.agilent.com/litweb ... -7150E.pdf
A diode operated in the square-law region will produce an output proportional to the average power as long as the peak amplitude of the measured signal is kept inside the square-law region of the diode (roughly <-20dBm). This is true for the E9300 since they have a design with high power and low power diode stacks, so they are always using the diodes inside the square law region. The E4412A/E4413A CW sensors do not have this provision, so the operator must use care to keep the peak power <-20dBm if you want to use make this type of measurement.
A thermocouple sensor like the 8480 or N8480-series will always be OK for this measurement.