Issue link: https://iconnect007.uberflip.com/i/1028393
SEPTEMBER 2018 I DESIGN007 MAGAZINE 63 Subsequently, transient simulation was per- formed for the abovementioned three test cases to observe the phenomenon of FEXT in time domain. In this transient simulation, a square wave signal with 500MHz Nyquist frequency, 1.2Vpp amplitude and 3V/ns slew rate is injected into port 1 of each test case's 3DEM model, with port 3 being pulled low (i.e., serves as near end point of victim line), followed by probing at port 4 (i.e., serves as far end point of victim line). Referring to Figure 5, the noise induced at the far end-point of the victim line in time domain for test case 1, 2 and 3 is 60mVpp, 90mVpp and 170mVpp, respectively. The least severe crosstalk is encountered in test case 1 because each signal has its own return path, and the ground trace between the two signal traces minimizes the crosstalk between them. This structure provides a per- fect and sufficient return path for the signal. In test case 2, each signal has its own return wire or path too. However, ground trace removal between the two signal traces worsens the for- ward signal crosstalk vs. test case 1. On the other hand, for test case 3, there is only a single ground wire connecting the reference planes of two boards. This sole ground wire becomes the return path bottleneck of the two signal traces and results in more critical crosstalk for the return current vs. test case 2, contributed by the sharing of a common return path by dif- ferent signals that intensifies the interference or coupling among the signals due to higher mutual inductance. Lastly, the surface return current density on ground wire for these three test cases is depicted in Figure 6. When a square wave at 500 MHz Nyquist frequency is injected at both Figure 5: Noise induced at victim's far end-point due to crosstalk for three test cases.