Good point, yes, the Switch Node has the most dv/dt. A better drawing would be to draw the parasitic cap from the switch node to the test ground plane, rather than the 5V rail! Definitely agree. I guess the main point i wanted to make in the video, was, from the energy transportation point of view, the L-C filter is just some energy storage devices, taking the energy from the input capacitance and transport the energy to the load, eventually, but it takes time, therefore, all the parasitic component close to the load will also contribute to make the energy transition process, and this is the main cause of common mode noise.

Hi, 0V is just a name, isn't it! We all know that we name it as 0V, simply because we reference it to the same point of battery negative. But as it is demonstrated here, the battery negative is not the same as the PCB 0V, even for DC, they are not the same, because the voltage drop caused by the long wire, not to mention the RF signals. So there will be a voltage drop between battery negative and the 0V on the PCB. Same applies between the test ground plane and the PCB 0V, there will certainly be a voltage difference between the 0V on the PCB and the test ground plane, as a result, you will see current induced (via the parasitic capacitance between the 0V and the test ground plane), therefore, induced RF current will flow on the 0V line. Hope this helps

Hi, good question. Think about in a test house, the chamber ground is connected to the wall, so the whole Faraday shield is "earthed". Inside the chamber, the antenna is measured with reference to the chamber ground. The key to your question is to understand that the RF reference ground is not earth ground, though for safety reasons, we often earth the RF ground.

"Earth" ground is -only- for safety or lightning protection. It has no bearing on EMC measurements.