CW normalization and source amplitude



In order to interpret the results from power monitors correctly it is important to understand (1) the normalization used and (2) the effect of the source amplitude.

There are two normalization states: “CW normalization” and “No normalization”. You can select this in the “Setting” menu as shown below:

The difference between these two settings is explained in detail in this link. When CW normalization is chosen, the results for the fields in the power monitors are normalized by the Fourier transform of the source signal s(t), this returns the impulse response:

The source signal is shown in the “Frequency/Wavelength” tab of the source settings:

By default, the CW normalization is used. To disable it you have to select “No normalization” in the “Normalization state”.

The second issue, is the effect of changing the source amplitude. This can be done in the “General” tab of the Edit source window. The source signal s(omega) does not depend on the source amplitude, but the fields in the simulation do depend on it. Therefore, if the source amplitude is 100 times larger the E field shown by the monitor will be 100 times larger. This scaling happens for both “CW normalization” and “No normalization”

We can look at a simple example to understand the points above. You can run the simulation file usr_CW_freespace.fsp (11.1 KB) and compare the output from “Monitor1” for the different possible settings. In the simulation a plane wave is propagating in free space. The figure below shows the calculated electric field (real part) for different settings.

Note that when CW normalization is used the electric field curve is flat, even though the source signal is not; this is expected because the source signal is used for the normalization. The small deviations from the flat behavior are the result of numerical grid dispersion as explained here. The value of the field (1 or 100 in the example) depends on the source amplitude. On the other hand, if “no normalization” is used then the E field curve follows the shape of the source signal, as expected.

It is also a good idea to look at this example in the Knowledge Base, where the results for transmission are also discussed.

Source power
Frequency domain field and power 모니터에서 electric field에 대한 질문
How to pump a ring resonator
Solar cell 관련해서 문의드립니다
Scale bar in surface plot units
FDTD source amplitude unit
How can I get mode source poynting vector?