Issue link: https://iconnect007.uberflip.com/i/1467185
56 DESIGN007 MAGAZINE I MAY 2022 support that a major supplier like NCAB can offer to its customers during the design, quot- ing and production stages of a circuit board. In this instance the particular focus was on signal integrity. Reflecting upon the emergence of demand for controlled impedance product, he observed that during the 1990s only 10-20% of designs had that requirement, whereas today's trans- mission rates increased the percentage to the 80-90% region and upcoming 5G networks operating at 28 GHz would require ever more critical PCB properties. His schematic balanc- ing PCB design challenges against design tools and process over the last five decades indicated a rapidly growing area of risk as complexity increased. What could be done to minimise this risk and to avoid over-specifying or under-spec- ifying the design? e answer was to consult with the PCB supplier's applications engineer- ing specialists as early as possible in the project and allow them to calculate and recommend optimum materials and stack-ups to achieve cost-effective performance and reliability and shortest time-to-market. He stated that NCAB set standards stricter than IPC, both in tech- nical requirements and in quality assurance. And in the case of special requirements, they offered application-specific consultations with field engineering experts. Op den Camp gave examples of designs with data greater than 2.3 Gbps and impedance tolerances of 5%. Another significant point he made was that many of the designs offered for volume pro- duction were based on the capabilities of pro- totype suppliers whose yields might not be economic in a production environment. Again, consultation at an early stage could improve the suitability of the design for volume manu- facture and offer yield improvements and cost savings. He gave examples of products designed for operation at high analog frequencies and high- speed digital applications and discussed hybrid constructions and low-loss materials. His graphs of relative permittivity versus frequency indicated that for FR-4 materials, dielectric constant was "not really constant" as frequency exceeded 2 GHz, so there was a Michiel Op den Camp