CircuiTree - July 2008 - (Page 26)

Toward a PCB Production Floor Metric for Go/No Go Testing of Lossy High-Speed Transmission Lines tions around the world will alter the loss properties exhibited by transmission lines as differing amounts of water are taken into the dielectric. As expected, the RIE measurement on the surface microstrip agrees with this. When the boards are subjected to different environmental conditions, the effect can be seen. Results were obtained using TDR1 only. The RIE results from the embedded microstrip (Figure 16) show that this structure is less susceptible to immediate changes of humidity than the surface microstrip (Figure 15). Figure 16 Variations in RIE Due to Environmental Conditions (Stripline) Figure 14 RIE Values Versus VNA Readings Environmental Considerations microstrip layers do not show a greater loss characterization for the IS410 (1080) because of the impedance mismatch induced undulations of the measurements. It is important that the impedance mismatches be limited to ensure accurate S21 measurements. These results from the VNA measurements were also indicated in the RIE values. Table 2 contains direct correlation between the VNA measurements and the computed RIE values. An RMS line was ﬁtted to the S21 measurements between 0.5 GHz and 1.6 GHz, shown in Figure 13,2 to counteract the impedance mismatches. A single frequency point of 1.4 GHz from this ﬁtted line was chosen for correlation. The value was determined using the equation 0.35/250 ps. The results recorded in Table 2 and shown in Figure 14 show a high degree of correlation between the loss measurement from the VNA and TDR2. FR-4 is fundamentally hygroscopic (i.e., absorbs moisture from the environment). Inevitably this means that different condiTrace/material L1 12in IS410 (2116) L3 12in IS410 (2116) L6 12in IS410 (2116) L8 12in IS410 (2116) L1 12in IS410 (1080) L3 12in IS410 (1080) L6 12in IS410 (1080) L8 12in IS410 (1080) L1 12in IS408 L3 12in IS408 L6 12in IS408 L8 12in IS408 L1 12in N4000-13 L3 12in N4000-13 L6 12in N4000-13 L8 12in N4000-13 26 July 2008 • circuitree.com Figure 15 Variations in RIE Due to Environmental Conditions (Surface Microstrip) Discussion When testing the characteristic impedance of PCB traces, designers and fabricators have often taken an informal approach to impedance coupon design. The effect of this has been that perhaps not as much attention was paid to the launch characteristic of the test probe/coupon interface as would be desirable. Launch of the test signal into a coupon for trace loss measurement takes on a more critical aspect, with a much higher chance RIE (TDR2) 2.82 4.18 4.18 2.78 3.20 4.07 4.05 3.07 2.62 2.80 2.81 3.05 2.55 2.73 2.77 2.59 VNA 2.869 4.052 3.995 2.823 3.074 3.924 3.798 2.923 2.290 2.325 2.318 2.655 2.223 2.252 2.313 2.246 Table 2 Table 2 RIE Versus VNA of inaccurate or misleading measurements if the coupon design is not fully considered along with the test probe interconnect. This paper proposes that where possible a standard approach to coupon design should be adopted—especially where product is being sourced from multiple vendors. Good progress has been made in deﬁning a candidate layout for this standard. A dual-line probe using the back to back pin layout improves R&R. These results show that the RIE ﬁgures for the embedded stripline structure have the greatest differentiation between materials. It is also known that inner layer traces can be manufactured more reliably and precisely, so a more consistent test structure can be created. Currently most manufacturing PCB transmission line testing is done at frequencies and geometries where the environment causes little variation. Ideally, localized RH should only minimally affect the proposed production ﬂoor metric. Previously suggestions have included baking of boards prior to testing and/or soaking of boards in a controlled RH environment. We believe that in a production environment these options may prove uneconomic due to additional process time rendering them suitable only for batch testing. Inner layer traces are far less responsive to immediate changes of environment so are more suitable for this testing. It is important to realize that the test methodology outlined in this paper is aimed at material validation and fabrication, not validation of the design; validation of the design should have been done elsewhere. It may be concluded that the best structure for material validation will be a stripline structure—even if this is not representative of the structures used on the board. Although the absolute RIE numbers are different when obtained by different TDRs, with the corresponding methods of ﬁltering, the RIE numbers show the same relationship and identify similar trends. These numbers also show a similar relationship to the loss ﬁgures measured by VNA. The difference in 28 http://circuitree.com

Table of Contents Feed for the Digital Edition of CircuiTree - July 2008

CircuiTree - July 2008 Contents My Line Industry Review Tech Talk Flexible Thinking Toward a PCB Production Floor Metric for Go/No Go Testing of Lossy High-Speed Transmission Lines Intelligent Design 20-Year Retrospective Ask the Flexperts Environmentally Speaking BPA Growth Curves Considering Design Variants to Maximize Process Efficiency Market Outlook Technical Product Spotlights Classified Ads Upcoming Events Ad Index

CircuiTree - July 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.