| FORUM

FEDEVEL
Platform forum

Reference for HS signals theoretical question

adamS90 , 10-22-2023, 01:22 PM
Hi,
In traditional transmission line (ex coax, or in PCB fig 1) we have source physically connected between positive line and negative line. In that case the positive conductor and negative conductor that are PHYSICALLY connected to the source are forming guides for the energy to travel between them (for DC current we have physical low velocity drift current on both lines physically connected, for AC we have electrons oscillating on that lines and we say electrons oscillate together with EM energy (regardless of whether the energy prompts the electron oscillation or vice versa)).

Now, in case of figure 2 the energy (wave) is guided by SIG and PWR layers and only in dielectric between them the energy is traveling, but the GND layer is physically connected to the negative terminal - so we should expect oscillating electrons on that conductor, therefore if we say electron oscillation=EM propagation we should also have some Poynting vector along the negative (here GND) terminal. So does on the fig 2 we have energy flowing between PWR and GND as well (orange poiting vector)?

And, if we stick to the convention the current is flowing, what is the path of that current - fig 2 or fig 3 blue line?
I think it is fig 3 for AC signals, but.... let’s consider DC current, cause for DC we also have fields and Poynting vectors, so for DC nobody negate our drift current loop is the one from picture 2, but the field that is transmitting energy for sure goes between SIG and PWR as can not go over PWR conductor, so we have different drift current loop and different area where the energy is flowing, i can not explain this, anybody can?


Lastly on picture 4 when we have cut in the PWR then in that place the E-field goes between SIG and GND and we have impedance mismatch, this is right understanding, right?

qdrives , 10-23-2023, 01:16 PM
For figure 2 that depends on the frequency. At 10MHz and below, yes there will be current on subsequent layers. This may be due to the lack of 'skin effect'.
According to Heidy Barns that made a simulation, the current was in the same direction (PWR and Gnd).


I am trying to understand what your question is for figure 3, but it is best to have a capacitor are the receiver end that couples Pwr to Gnd so that the energy can return to Gnd.

In figure it is not only the impedance mismatch, but also the EMC noise caused by the slot.
Running high speed clock traces over a gap or slot in the ground return (or reference) plane can cause radiated emissions and crosstalk issues. This short vi...
Use our interactive Discord forum to reply or ask new questions.
Discord invite
Discord forum link (after invitation)

Didn't find what you were looking for?