PCB dielectric loss testing quantifies the dissipation factor (Df, tan δ) of copper-clad laminates and finished printed circuit boards, determining how much electromagnetic energy the dielectric material converts into heat at high frequencies. Accurate Df characterization is essential for material selection, insertion-loss budgeting, and signal-integrity simulation in 5G, millimeter-wave radar, and high-speed digital systems. This article explains dielectric loss fundamentals, recognized test methods under IPC, IEC, ASTM, GB, and JIS frameworks, sample requirements, and typical industrial applications.
Technical Overview
Dielectric loss is the portion of electromagnetic energy dissipated as heat within an insulating material when it is subjected to an alternating electric field. It is expressed through the dissipation factor (Df), also known as the loss tangent (tan δ), which is defined as the ratio of the imaginary part (ε″, loss factor) to the real part (ε′, dielectric constant) of complex permittivity. A higher Df means greater attenuation of high-frequency signals as they propagate along or through the dielectric.
For printed circuit boards, dielectric loss directly governs insertion loss, channel attenuation, thermal buildup, and resonator quality factor. Standard FR-4 laminates typically exhibit Df values of approximately 0.015–0.022 at 1–10 GHz, mid-loss materials range from 0.008–0.012, and low-loss thermoset or PTFE-ceramic systems can reach 0.0010–0.0045. As data rates advance from 56 Gbps toward 224 Gbps per lane and wireless platforms extend into millimeter-wave bands, verifying Df with repeatable, traceable test methods has become a core qualification requirement for laminate suppliers, PCB fabricators, and system designers.
Df is strongly frequency-dependent and also varies with temperature, moisture absorption, resin content, glass-fabric style, and measurement direction due to the anisotropy of woven glass reinforcement. Professional dielectric loss testing therefore always reports the measurement frequency, conditioning environment, and specific test method, because results obtained by different techniques can differ systematically.
Testing Samples
Copper-Clad Laminate (CCL) Sheets
Flat laminate panels with the copper fully etched away are the primary sample type for resonator-based Df testing. Depending on the method, panel sizes from approximately 30 mm × 30 mm up to 300 mm × 300 mm and thicknesses between 0.1 mm and 3.2 mm are accepted. Testing multiple thickness points is recommended because resin-to-glass ratio changes the effective dissipation factor.
Finished Multilayer PCBs and Test Coupons
Production boards or dedicated coupons can be evaluated to confirm that lamination cycles, post-processing, and moisture uptake during fabrication have not degraded the dielectric. These measurements help correlate raw material datasheets with actual channel-level signal-integrity performance.
Prepreg and Cured Resin Plates
B-stage prepreg can be pressed and cured into neat flat plates under controlled conditions and then tested, enabling characterization of the resin system itself. This supports new material development, incoming inspection programs, and qualification against specification sheets such as IPC-4101.
Testing Items
A comprehensive dielectric loss characterization program typically includes the following measured and derived parameters:
- Dissipation factor (Df / tan δ) at discrete frequencies, commonly 1 MHz, 1 GHz, 2.5 GHz, 5 GHz, 10 GHz, 20 GHz, 40 GHz, and 77–79 GHz
- Dielectric constant (Dk / εr), which most resonator methods determine simultaneously with Df
- Frequency-sweep loss curves across an application band to capture dispersion behavior
- Temperature-dependent Df, for example from −40 °C to +150 °C, for automotive and aerospace environments
- Df after humidity conditioning, such as 85 °C / 85 %RH exposure, to quantify moisture sensitivity
- Df before and after thermal stress, reflow simulation, or accelerated aging to assess long-term stability
Testing Methods
Parallel-Plate Capacitance Method (ASTM D150 / IEC 62631-2-1)
A guarded-electrode technique applied at low frequencies, typically from power frequency up to about 1 MHz. It provides good accuracy for Df values above roughly 0.001 but is not representative for gigahertz-class designs and is mainly used for baseline quality control.
Stripline Resonator Method (IPC-TM-650 Method 2.5.5.5)
A resonant stripline circuit is etched on the sample material and sandwiched between ground planes. The quality factor and resonant frequency shift yield both Dk and Df, typically in the 1–10 GHz range. Because the measurement replicates an actual circuit configuration, this method is widely used for incoming laminate inspection, although conductor roughness contributes to the measured loss.
Split-Cylinder and Split-Post Dielectric Resonator Methods (IPC-TM-650 Method 2.5.5.13)
A high-Q cylindrical or post-pair resonator is split so that a flat, copper-free sheet can be inserted between the halves. The shift in resonant frequency and the change in loaded versus unloaded Q factor provide precise Dk and Df values, generally from 1 GHz to 20 GHz. These techniques offer excellent repeatability for low-loss materials down to Df of approximately 0.002 and require minimal sample preparation.
Cavity Perturbation Method (ASTM D2520)
A small specimen is inserted into a high-Q metallic cavity, and the resulting perturbation of the resonant frequency and Q factor is analyzed. The method extends into the tens of gigahertz and supports elevated-temperature measurement, but it demands accurate sample dimensions and careful fixture characterization.
Balanced-Type Circular Disk Resonator Method (IEC 63138 Series)
A conductor-backed coplanar waveguide excites a balanced circular disk resonator patterned directly on the sample, enabling Dk and Df determination from roughly 10 GHz up to 110 GHz. It is a reference method for millimeter-wave PCB materials used in 77 GHz automotive radar, backhaul radios, and beyond-5G modules.
Open Resonator and Transmission Line Methods
Fabry–Pérot open resonators and free-space techniques measure thin, low-loss sheets at very high frequencies with minimal fixture loss, while vector-network-analyzer transmission methods extract complex permittivity from S-parameters of transmission-line structures. These approaches are useful for broadband sweeps and design-correlation studies, provided that conductor and radiation losses are properly de-embedded.
Applications
Dielectric loss testing supports qualification and design activities across every high-performance electronics segment:
- 5G and 6G wireless infrastructure, including sub-6 GHz massive MIMO antennas and millimeter-wave front-end modules where Df bounds array efficiency and thermal load
- Automotive electronics, especially 77–79 GHz corner radar and in-cabin sensing boards that demand stable Df across temperature
- High-speed digital systems such as servers, switches, and artificial-intelligence accelerators, where Df defines the insertion-loss budget for 112G and 224G serial links
- RF and microwave circuits including filters, couplers, power amplifiers, and oscillators, whose Q factors depend directly on dielectric loss
- Aerospace, satellite, and defense platforms requiring loss stability under humidity, thermal cycling, and radiation environments
- Advanced packaging substrates and IC carriers, where thin low-loss dielectrics enable dense high-frequency interconnect
Accredited third-party laboratories routinely perform these measurements to support supplier qualification, datasheet verification, failure analysis, and incoming inspection for both domestic and export-oriented manufacturers.
FAQ
At which frequency should dielectric loss be tested?
The test frequency should match the application band whenever possible. Low-frequency values from ASTM D150-style methods do not extrapolate reliably to gigahertz designs because Df rises with frequency for most laminate systems. For general RF materials, 10 GHz is the industry reference point, while radar and millimeter-wave products should be verified at 40 GHz or higher using resonator methods suited to those bands.
What is the difference between dielectric constant and dissipation factor?
Dielectric constant (Dk, εr) describes how much electric energy the material stores and determines signal velocity and characteristic impedance. Dissipation factor (Df, tan δ) describes how much of that energy is lost as heat and determines attenuation. Both are components of complex permittivity, and most resonator test methods report them together.
How are samples prepared for Df testing?
Typical preparation includes completely etching away the copper foil, cleaning residues, baking the specimen, for example at 105 °C for one to two hours, to remove absorbed moisture, and conditioning it in a standard laboratory atmosphere of approximately 23 °C and 50 % relative humidity. Flatness and thickness uniformity are critical because resonator methods calculate permittivity from the sample geometry.
What Df level qualifies as a low-loss material?
Industry practice generally classifies materials as follows at 10 GHz: ultra-low loss below 0.002, typical of PTFE-ceramic systems; low loss from 0.002 to 0.005, covering many hydrocarbon and enhanced-epoxy laminates; mid-loss from 0.005 to 0.010; and standard FR-4 from 0.015 to 0.022. The required class depends on link length, frequency, and the allowed insertion-loss budget.
Why do Df results differ between test methods?
Differences arise from electric-field orientation relative to the glass weave, fixture and conductor losses included in the measured Q, test frequency spacing, and sample thickness. Woven-glass laminates are anisotropic, so the through-thickness value measured by a parallel-plate method can differ from the in-plane value seen by a stripline circuit. Data should only be compared when obtained with the same method, frequency, and conditioning history.