Simulation is the simulation of PCB design, PCB design is used for the design of processing and production. As more and more error factors are brought to design by processing, is the simulation of PCB design or the simulation after production and processing taken into account? In his previous articles, Mr. Gao has shared a lot about the impact of machining errors on PCB design in the domain of high-speed serial links, and the difference between simulation results with and without machining. Of course, we all know that with higher and higher rates, machining factors, such as stub through hole, hole aperture, layer deviation, etching and other factors, will bring more and more impact on high-speed signal performance. We have introduced and shared cases in the previous articles.
Since we have talked a lot about high-speed processing factors, we are going to talk about a simpler one today, that is, the simulation design of power supply voltage drop, which is often referred to as DC simulation. As for its simulation principle, it is probably the simplest one of the simulation theory, the following picture can be clear.
Therefore, as long as we get the voltage value and the maximum current value of the power supply network, we can simulate the voltage drop of the PCB power link. Just like the example below, the power link from VRM to the sink end has a voltage of 0.85V and a current of 5A. After knowing these parameters, we can do it easily.
The following is the simulation results of the power link. It seems that both the voltage at the load end and the current density of the whole link can meet the requirements. That is to say, the PCB power supply design after simulation is OK!
This is of course a very objective result, which objectively and accurately reflects that the PCB design of the power link meets the requirements. And that’s basically the end of the simulation.
Suddenly one day, Mr. High Speed thought that the PCB link of the power supply might also be affected by some processing factors like the high speed signal? Then Mr. Gao Fei consulted some IPC standards to see if there were any machining tolerances that would affect the design of power supply. Finally, he found the standard of copper thickness machining in a section, which is the following picture.
As can be seen from the figure, the copper thickness of 0.5oz is generally made at 0.6mil, but the most extreme limit is 0.449mil. Since it is listed in the IPC standard, that is to say, even if 0.5oz copper is really made after processing and only 0.449mil thick, it still meets the requirements!
Mr. Gao Fei suddenly thought, if the copper thickness of the case mentioned above is really only 0.449mil after processing, will the pressure drop be…
When everyone was still worried, Mr. High Speed had directly changed the copper thickness from 0.6mil to 0.449mil in the original case for simulation verification! It turns out that this change has a significant effect on both the voltage drop and the current density.
According to the simulation results of 0.449mil, the voltage drop increased by about 7mV, and the current density deteriorated by 33A/mm², which could not be ignored.
After, of course, aware of the problem of “gravity”, Mr High-speed and also our DFM engineers and board factory colleagues, want to know whether to make probability limit copper thick high is not high, well they said basically have not seen such a situation, or to 0.6 mil is given priority to, the following are some of the PCB they provide copper thick slice figure, It’s consistent with what they said.
Although this is just a whim of Mr. High Speed’s ideas and simulation verification, there is no actual board with the ultimate copper thickness. But from the general direction, more to understand some processing knowledge, especially the impact of processing on design and simulation whether it is for hardware engineers, PCB engineers or SI engineers are very helpful Oh!
