There are several challenges associated with accurately measuring high-speed differential devices. Among these are that non-coaxial calibration standards are not sufficiently precise, calibration techniques are commonly misapplied, and direct test fixture measurement is very difficult. This paper has illustrated a method of removing the effects of test fixtures in a way that is accurate, relatively simple to perform, and employs a minimal number of assumptions.
