2016 Best Paper Winner

Round Robin of High Frequency Test Methods by IPC-D24C Task Group

Round Robin of High Frequency Test Methods by IPC-D24C Task Group

Glenn Oliver
Jonathan Weldon
Research Triangle Park, NC
Chudy Nwachukwu
Chandler, AZ
John Andresakis
Park Electrochemical
Melville, NY
John Coonrod
Rogers Corporation
Chandler, AZ
David L. Wynants, Sr.
Taconic Advanced Dielectric Division
Petersburgh, NY


Currently there is no industry standard test method for measuring dielectric properties of circuit board materials at frequencies greater than about 10 GHz.  Various materials vendors and test labs take different approaches to determine these properties.  It is common for these different approaches to yield varying values of key properties like permittivity and loss tangent.  The D-24C Task Group of IPC has developed this round robin program to assess these various methods from the “bottom up” to determine if standardized methods can be agreed upon to provide the industry with more accurate and valid characteristics of dielectrics used in high-frequency and high-speed applications.

Problem Statement

Accurate values of relative permittivity (ɛr) and loss tangent (tan δ) are important characteristics for designers and fabricators in predicting electrical performance of circuits at high frequencies [1].  The most common method for evaluating these parameters at frequencies up to 10 GHz is described in IPC-TM-650- [2].  This method is equivalent to ASTM-D-3380. [3] This method excites a stripline resonator at both ends with the dielectric under test comprising most of the volume.  The stripline is created by establishing intimate contact using a constant clamp force.  This method is highly repeatable and is optimized for QA testing at a specific frequency.  This method is not well suited for characterizing at frequencies higher than 10 GHz.

Both analog and digital applications now commonly excite dielectric materials at frequencies well above 10 GHz.  Measurements at higher frequencies are especially challenging for many reasons.  For instance, the wavelength of radiation at 30 GHz is < 10 mm in air and < 5 mm in FR4.  This makes it more challenging to isolate the interactions of the waves with the material under test from any parasitics introduced by the test fixture.  Another significant challenge at these high frequencies is that current is concentrated at the “skin” of metal surfaces.  As frequencies increase, the microstructure of metal surfaces contributes more significantly to overall loss or degradation, and makes it nearly impossible to isolate the impact of the dielectric losses separate from the metal.


In an effort to potentially determine standardized test methods at these frequencies, seven members of IPC D-24C Task Group developed a round-robin to measure ɛr and tan δ for various printed circuit board (PCB) materials using different methods of their choosing and compare results. 

First, this paper details the problem followed by a description of the various evaluation methods being considered; each method is described with sufficient information to allow for third party replication.  Next, the results from each labs independent dielectric property characterizations are presented and subsequently compared.  Finally, this paper will discuss each methods pros and cons and any conclusions or next steps.

Continue reading - IPC members only