PCB return current path problem
I know that the return current will use the plane closest to the signal trace. Then can you give me the answer to these two specific conditions?
1) When we say 4 layers, layer 1 2 3 is the signal layer, and the plane is continuously on the 4th layer. For all 3-layer signals, the return current path will be on the 4th plane because there are no other planes.
2) Now for the second case, the 4-layer overlay is the signal ground plane signal. Suppose there is a signal trace on the third layer, but I also inject copper into some of the vcc tracks on some areas of the third layer (incomplete). Then, if I draw the signal trace on the 4th layer, the return current will be injected on the ground plane of the 2nd layer or part of the copper on the 3rd layer, so that the trace will pass under part of the copper casting and will not be poured in part of the copper. Passed under.
I hope that I am very clear.
It would be clearer if you post some photos. In addition: What is the PCB for high speed signals? If not, remember that the return current (actually any current) always follows the path with the least impedance. -
For DC and low frequency fills (<several MHz /> several hundred ns long switch edges, don't worry too much about RF magic), the ground plane will be the return path. After entering the VHF range, the return path is usually the path with the least impedance. If your VCC power plane has a lot of ceramic decoupling, then the return current will follow the nearest laminar flow (ie either have good capacitance coupled to ground or ground)
@Sam I don't know how. The capacitor has an effect on the return current because the return path is a huge road and the capacitor is near the pin. I only know about dec. cap. Yes they have frequency. Therefore, by connecting some of them in parallel, we get low impedance and they provide a more constant voltage level we want.
So in this case, since continuous grounding may have the lowest impedance, it will be our return current path because some of the copper casting on any layer cannot have a lower impedance than the fully cast ground plane -
But how to reach each plane (via) will have a greater impact on the impedance of copper dumping on the plane.
Good ceramic decoupling capacitors are very close to the short circuit of the high frequency signal, which is why they allow the return current to pass between the layers at the higher end of the frequency range. DC / LF may pass through the ground plane, but if there is a close decoupling capacitor, then multiple frequencies will want to keep as close as possible to the signal trace, which allows HF to run closer to the signal trace along the VCC layer, then That's it. I don't worry if nothing runs on a few mega frequencies and nothing produces a switching edge below 100ns. But there is no actual contact. I mean, if we connect a trace from the source to the receiver, for example from the mcu pin to the i2c sensor pin, the electromagnetic field will do this and the current will return along the same path but on the plane. In fact, it seems to me that we don't seem to be physically connected to anywhere. I'm confused:) -
Therefore, it is important to summarize all of these details, such as impedance ground plane stacking, at high speeds or high frequencies. I don't need to worry too much at low frequencies. Mcu sensor and other circuits -
The current will be separated according to its impedance in all possible ways, so the shortest and more conductive way will get more current, rather than longer and less conductive. But you can consider each wire/through hole/connector/equal resistance and (usually very small) combination of capacitance and inductance. Then you have to solve the resulting network to get the current and voltage values everywhere. Usually this is not important because there is a wide path, almost all currents are such, and there is such a small impedance on the path, in any case the voltage is almost zero. -
Why do you care about the return current? Are you concerned about radiated emissions or signal integrity? Or something else? In any case, you should always route the fastest edge signal to a position immediately adjacent to a continuous plane. If there is any break on the aircraft, do not route the fast edge signal at the break. This applies to both stacking options. The clock and any signal with fast rise and fall times should be in close proximity to GND. If you place GND on the internal layer, you can place two tracking layers next to GND. Things to consider.