Issue link: https://iconnect007.uberflip.com/i/1540184
OCTOBER 2025 I DESIGN007 MAGAZINE 11 F i g u re 4 : E M s i m u l ato r 3 D v i ew of t h e P i c ote st S 5 0 P C B a n d t h e re s u l t i n g e l e ct ro m a g n et i c f i e l d s w h e n exc i te d by a 5 0 - a m p d y n a m i c ste p l o a d at 1 M H z w i t h a 1 0 n S e d g e rate o n t h e a g g re s s o r p owe r ra i l . T h e re i s a re s o n a n c e at 1 1 M H z t h at re s u l t s i n t h e wo rst- c a s e c ro s st a l k a n d g ro u n d b o u n c e f ro m t h e a g g re s s o r p owe r ra i l to t h e v i ct i m p owe r ra i l . ▼ a PCB design. This micro-ohms of impedance must also have a DC resistance component to minimize IR voltage drops and electro-thermal heating. The question is not if, but rather why a power integrity engineer wouldn't simulate. Full FEM simu- lators are easily accessible and provide invaluable insights through pre-/post-layout validation before expensive hardware is built and before electronic hardware makes its way to the customer. Free data- sheets, advice, and tools all sound good, but what downstream risk is being traded for a lack of real engineering? A full 3D EM simulator, like Keysight's PIPro in ADS, can create a behavioral model of the PCB interconnect 4 . Unlike measurements, probes in simulation can be placed anywhere, even on the ground net to look at the same net voltage varia- tion, aka ground bounce. Or probes can be placed on adjacent power rails to look at net-to-net cross- talk (Figure 3). In this example, the aggressor power rail has a resonance around 11 MHz. When excited by a fast 50-amp 10-nS step load cycling at 1 MHz, the aggres- sor power rail noise can couple to the victim power rail, causing voltage ripple due to net-to-net crosstalk and ground bounce voltage variations on the shared ground net. EM simulators not only enable the inclu- sion of the PCB parasitics when simulating dynamic power rail ripple but also provide valuable EM field plots for assessing where EMI coupling occurs. Fig- ure 4 shows the EM simulator results for the E-fields on the Picotest S50 PCB layers. The fields on the vic- tim power rail at 1 MHz compared to 11 MHz clearly show that more coupling is happening at 11 MHz and where it is happening. Figure 5 illustrates how a quick modification to the victim power rail routing, which removes some areas " " The question is not if, but rather why a power integrity engineer wouldn't simulate.