Issue link: https://iconnect007.uberflip.com/i/241303
article influence of via stub length and antipad size on the insertion loss profile continues Figure 3: Influence of stub length on insertion loss profile. Measurement Results that Prompted the Investigation During a routine measurement of insertion loss over frequency on a 0.220" board (Figure 2), significant differences were found, depending on the measured layer. Testing the same layers with backdrilled vias eliminated the differences and resulted in a straight insertion loss curve without the deep resonances (Figure 3). This finding initiated a thoroughly investigation into the influence of via stubs on electrical performance. The Test Vehicle An 8-layer stackup was used for the test vehicle, with one offset stripline on layer 3 (referencing to ground layers 2 and 4). Layer 6 was an unused layer and layers 5 and 7 were also ground layers. The outer layers provided the landing patterns for probing. The probing was performed from the top side, which in turn generated maximum via stubs for the layer 3 features. A mid-loss material has been applied for the DOE, as many designs in the range of 3.125 – 10Gbs+ are using them. Similar glass styles and thicknesses were used for the cores and prepregs to get a relatively balanced stripline design. A rather wide line width in combination with 1 oz. copper delivered minimum DC resistance. Together with a smooth copper foil, these design attributes were resulting in a relatively small insertion loss. The complete stackup details can be found in Figure 4. The design consisted of single-ended and differential transmission lines on layer 3, with a via connecting the lines to the outside at each end of the traces. In addition, single-ended and differential impedance test coupons were placed on the panel. The size of the antipads on the plane layJanuary 2014 • The PCB Design Magazine 45