How to Read a PCB Laminate Datasheet: Dk, Df, Tg, CTE, Td – All Key Parameters Explained

Dielectric Constant (Dk / εr): What It Controls and Why Stability Matters More Than the Value

Dk — also written as εr or relative permittivity — measures the material’s ability to store electrical energy relative to a vacuum. In practical PCB terms, Dk controls two things: the propagation velocity of signals through the board, and trace width for a given characteristic impedance.

Higher Dk means slower signal propagation and narrower traces for a given target impedance. FR-4’s Dk of approximately 4.2–4.8 produces 50Ω microstrip traces that are familiar to most layout engineers. Rogers RO4350B at Dk 3.48 produces wider traces for the same impedance. The exact Dk value matters for impedance calculation — but what matters more for real-world board performance is Dk stability.

A material quoted as Dk 4.5 at 1 MHz may behave as Dk 4.2 at 1 GHz and Dk 4.0 at 10 GHz. Standard FR-4 Dk variation from nominal can reach ±10% depending on resin content, frequency, temperature, and moisture content. That ±10% translates directly into ±5% impedance variation, which is the entire tolerance budget for a controlled-impedance design. Premium laminates like Rogers RO4350B are specified with ±0.05 Dk tolerance — a number that makes production impedance control actually achievable.

Critical datasheet reading points for Dk:

• Note the test frequency. Dk quoted at 1 MHz is almost useless for a 5 GHz design. Request frequency-swept data from the supplier if the datasheet only shows one frequency point.
• Note the resin content. Dk increases with glass content and decreases with higher resin content. Prepreg Dk differs from core Dk for the same product family. Always ask your fabricator for “press-out” Dk values matching your actual stackup construction.
• Never use a generic “FR-4” Dk assumption (often quoted as 4.2 or 4.3) for impedance calculations on production boards. Get the actual datasheet value for the specific product your fab house is using.

Dissipation Factor (Df / Loss Tangent): The Signal Integrity Number Above All Others

Df — also written as tan δ or loss tangent — measures the fraction of electromagnetic energy dissipated as heat within the dielectric per cycle. In plain engineering terms: higher Df means more signal energy turned into heat, which means higher insertion loss in your traces, filters, and antenna feeds.

At low frequencies (below about 500 MHz), Df is rarely the binding constraint even for standard FR-4 (Df ~0.018–0.025). Above 1 GHz, dielectric loss becomes a first-order concern. At 10 GHz and above, Df is the dominant loss mechanism in most PCB materials, and the difference between a Df of 0.020 (standard FR-4) and 0.0017 (Isola Astra MT77) is the difference between a functional and a non-functional mmWave design.

Critical datasheet reading points for Df:

• Like Dk, Df is frequency-dependent and typically worsens at higher frequencies. A Df of 0.003 at 1 GHz may become 0.005 at 10 GHz for the same material. Always use Df values measured at or above your operating frequency.
• Compare values measured by the same test method. Df measured by the Bereskin Stripline method may differ from Df measured by Split Post Cavity resonator on the same physical material. Mixing test methods in a comparison across suppliers produces misleading conclusions.
• Resin content affects Df. Higher resin content generally gives lower Df (the resin has lower Df than glass fibre). Datasheets quoting at 50% resin content are following IPC guidelines for standardised reporting — your actual board stackup may use different glass styles with different resin percentages, shifting the effective Df.

Electrical Parameter Benchmarks by Application

Application	Max Acceptable Df	Dk Range Target	Critical Parameter
Digital logic / power (<1 GHz)	~0.020	3.5–5.0	Tg, Td (thermal priority)
Sub-6 GHz RF / 5G	≤0.005	3.0–3.8	Dk stability, Df
High-speed digital (10–56 Gbps)	≤0.004	3.0–4.0	Df, copper roughness
mmWave 5G / automotive radar	≤0.002	2.5–3.5	Df, TCDk, Dk tolerance
Sub-1 mm wave / satellite	≤0.001	2.2–2.8	Df, Dk, moisture absorption

Thermal Parameters: What Determines Assembly Survival and Field Reliability

Tg — Glass Transition Temperature: The Thermal Mechanical Limit

Tg is the temperature at which the resin matrix transitions from its normal rigid, glassy state to a softer, more rubbery condition. This transition is reversible — the material returns to its original mechanical properties when it cools below Tg — but operating above Tg causes Z-axis expansion rates to increase dramatically, placing stress on via barrels and through-hole plating.

Standard FR-4 has Tg typically around 130–140°C. Lead-free soldering peaks reach 245–260°C. Any board going through lead-free assembly must be fabricated from material with Tg well above the peak reflow temperature — not because the board operates above Tg in service, but because it must survive assembly without structural damage. High-Tg FR-4 (Tg ≥ 170°C) is the baseline for lead-free production. For under-hood automotive and industrial applications where ambient exceeds 100°C, Tg ≥ 180°C provides adequate margin.

Three test methods are used to measure Tg, and they produce different numerical results for the same material:

• DSC (Differential Scanning Calorimetry, IPC-TM-650 2.4.25): Most commonly reported; good reproducibility.
• TMA (Thermomechanical Analysis, IPC-TM-650 2.4.24): Measures the dimensional change inflection point; typically gives lower Tg values than DSC.
• DMA (Dynamic Mechanical Analysis): Measures stiffness change; typically gives higher values.

When comparing Tg across datasheets, confirm which method was used. A “Tg 175°C by TMA” and “Tg 185°C by DSC” may describe the same material. If the datasheet doesn’t specify the test method next to the Tg value, request clarification before using the number for specification purposes.

Td — Decomposition Temperature: The Absolute Chemical Ceiling

Td is the temperature at which the resin undergoes irreversible chemical decomposition, typically defined as the temperature at which the material loses 5% of its weight as measured by TGA (Thermogravimetric Analysis, IPC-TM-650 Method 2.4.24.6). Unlike Tg, this transition is permanent — a material that has exceeded Td has chemically broken down and cannot be restored.

Td sets the absolute ceiling for assembly temperatures. Lead-free reflow peaks at 245–260°C. Rework operations may reach 280°C locally. Td of ≥340°C provides comfortable margin through these processes. Standard commodity FR-4 sometimes has Td values as low as 300–310°C — technically above peak reflow but with minimal margin for rework cycles. High-reliability designs specify Td ≥ 340°C as a minimum.

A common mistake is selecting material based on Tg alone, assuming adequate Td. Some materials have excellent Tg but marginal Td. Both must be checked independently.

T260 and T288 — Time to Delamination: The Assembly Process Reality Check

T260 and T288 measure how long a laminate can withstand 260°C and 288°C respectively before delaminating (separating of laminate layers), measured by TMA per IPC-TM-650 Method 2.4.24.1. These time values are more directly useful for production planning than either Tg or Td alone because they simulate cumulative thermal exposure.

A board with T260 of 60 minutes and T288 of 30 minutes can survive many lead-free assembly operations before delamination risk becomes significant. A board with T260 of 15 minutes will not survive rework scenarios that require extended exposure to high temperatures. For designs requiring multiple reflow passes (common in double-sided SMT assembly) or field rework, specify T260 > 30 minutes minimum and T288 > 10 minutes.

CTE — Coefficient of Thermal Expansion: Via Barrel Survival

CTE measures how much a material expands per degree of temperature change, expressed in ppm/°C (parts per million per degree Celsius). In PCB materials, CTE must be considered separately across three axes: X, Y, and Z.

X/Y-axis CTE is constrained by the glass reinforcement weave, keeping lateral expansion relatively low (typically 12–20 ppm/°C for FR-4). This is generally not a critical concern for most designs because copper traces expand compatibly with the laminate in these directions.

Z-axis CTE is the reliability-critical parameter. When a board heats up during soldering or operation, the laminate expands in the Z-direction (through the thickness). The copper plating inside via barrels, however, has a much lower CTE (~17 ppm/°C for copper). If the laminate’s Z-axis CTE is significantly higher than copper’s, the laminate stretches during heating while the copper barrel resists — generating tensile stress in the via barrel. Over thousands of thermal cycles, this stress causes fatigue cracking and via barrel failure.

Standard FR-4 Z-axis CTE typically runs 60–80 ppm/°C — roughly 3–4× higher than copper. High-performance filled FR-4 materials reduce Z-axis CTE to 40–55 ppm/°C. Rogers RO4350B achieves Z-axis CTE of 46 ppm/°C, meaningfully better than standard FR-4. The total Z-axis expansion from 50°C to 260°C is the number IPC specifications use — look for materials where this total is below 3.0–3.5% for high-layer-count boards with tight via aspect ratios.

Note: CTE changes above Tg. All PCB laminates expand at a higher rate above their Tg. This is why the combined effect of a low Tg and high CTE is particularly damaging — the material expands most aggressively at exactly the temperatures it encounters during reflow.

Thermal Parameter Reference for Common Design Scenarios

Design Scenario	Min Tg (°C)	Min Td (°C)	T260 Minimum	Z-Axis CTE Target
Consumer electronics, lead-free	150	320	20 min	<4.0% (50–260°C)
Industrial / commercial lead-free	170	340	30 min	<3.5%
Automotive body / chassis	170–180	340	45 min	<3.0%
Automotive under-hood	180+	360+	60 min	<3.0%
High-layer-count (>12L) / high aspect ratio	175+	340	60 min	<2.5%
Polyimide / extreme temp	>250	>400	N/A (robust)	<3.0%

Mechanical Parameters: Reliability Through the Physical Lifetime

Peel Strength: Copper Adhesion to the Substrate

Peel strength measures the force required to separate the copper foil from the laminate surface, reported in pounds per inch or N/mm (per IPC-TM-650 Method 2.4.8). Minimum acceptable peel strength for most applications is 1.0 lb/in (0.17 N/mm) on cured material. After thermal stress exposure, peel strength should remain above 0.70 lb/in.

For low-loss laminates using PTFE or ceramic-filled composites, peel strength can be lower than FR-4 materials because the smooth, non-polar dielectric surface provides less mechanical adhesion for copper. Copper surface treatment (VLP, HVLP foil) further reduces mechanical roughness, which is why peel strength specification is a counterbalancing concern when specifying very low-profile foils.

Moisture Absorption: The Hidden Electrical and Reliability Variable

Moisture absorption, measured by IPC-TM-650 Method 2.6.2.1A (percentage weight gain after 24 hours of water immersion), affects both electrical performance and mechanical reliability. Water in the dielectric raises Dk and Df, shifts impedance, and can contribute to delamination under thermal stress if moisture is present during high-temperature exposure.

Standard FR-4 absorbs 0.10–0.20% by weight. Rogers RO4350B absorbs 0.06%. Rogers RO3003 absorbs 0.04%. PTFE-based materials absorb <0.02%. For outdoor infrastructure and automotive applications where the PCB is exposed to humidity cycling throughout its service life, moisture absorption is a first-class specification criterion, not a footnote.

Moisture in the laminate before soldering is particularly dangerous. Hot laminate plus absorbed moisture creates steam that can cause internal delamination — the source of the well-known “popcorning” failure mode. Pre-baking boards before soldering is standard practice for moisture-sensitive laminates; the required bake time and temperature depend on moisture absorption characteristics stated in the datasheet.

CAF Resistance: The Failure Mode Dense Boards Create

Conductive Anodic Filament (CAF) is the electrochemical migration of copper ions along the glass fibre-resin interface under sustained voltage bias in the presence of moisture. CAF results in conductive filaments growing between adjacent vias or between planes, eventually causing insulation resistance failure. CAF is a primary concern in high-density designs with via-to-via spacings below 0.4 mm, operating in humid environments under DC bias.

CAF resistance is tested per IPC-TM-650 Method 2.6.25 — a 500-hour accelerated test at 85°C/85% RH under voltage bias between adjacent test structures. The result is reported as a pass/fail percentage at specific via-to-via spacings, or as the minimum spacing at which the material passes at a given failure rate.

Not all datasheets include CAF resistance data. For designs with tight via pitches, high-voltage bias, or outdoor/humid operating environments, request explicit CAF resistance data from the laminate supplier. Some material families (particularly those with specific CAF-resistant glass-resin interface treatments) are marketed specifically for their CAF performance.

Additional Parameters That Appear on Datasheets and What They Mean

Thermal Conductivity (W/m·K): How efficiently the laminate transfers heat. Standard FR-4 runs ~0.3 W/m·K. High-performance laminates may reach 0.6–0.8 W/m·K. This parameter is relevant for power-dense designs but rarely the primary driver for standard digital or RF boards where conduction cooling through copper planes dominates heat transfer.

Volume Resistivity (MΩ·cm): The bulk electrical resistance of the laminate material. High values (>10⁸ MΩ·cm) indicate good insulating quality. This parameter matters for high-voltage designs (HV power supplies, EV battery management) where leakage current through the substrate must be minimised.

Dielectric Strength (V/mil): The maximum electric field strength before breakdown. Standard FR-4 runs ~1000–1500 V/mil. For high-voltage designs, verify this parameter against your maximum working voltage with appropriate safety margins.

UL94 Flammability Rating: V-0 is the highest rating (self-extinguishing within 10 seconds, no dripping flaming material). V-1 and V-2 have progressively weaker flame resistance. Most production PCBs specify UL94 V-0. Some specialty materials (notably Rogers RO4003C) do not carry V-0 rating — its companion material RO4350B does, which is why RO4350B is the default specification for most commercial applications despite RO4003C’s slightly better Df.

The Test Method Problem: Why You Can’t Always Compare Datasheets Directly

This is the practical issue that experienced engineers run into when making laminate selection decisions from datasheets alone. Two materials from two manufacturers may list identical Dk values at the same frequency, but those values were measured by different methods on samples with different resin contents, tested at different laboratories, and may not actually represent identical electrical performance in your specific stackup.

The test methods defined in IPC-TM-650 set the procedures but don’t eliminate all sources of variation. Resin content variation alone can shift Dk by 0.2–0.3 units within a single product family (for example, a 1080 prepreg at 65% resin versus a 7628 prepreg at 45% resin for the same base material). Some manufacturers include resin-content-dependent Dk/Df data in their datasheets or application notes; many provide only a single number at a nominal 50% resin content.

The reliable approach: use datasheet values to narrow down your candidate materials to two or three options, then request frequency-dependent characterisation data at resin contents matching your actual stackup construction. Many laminate suppliers’ application engineering teams can provide this data upon request, or can refer you to a stackup tool (such as Isola’s PCB Stackup Designer or Siemens Z-planner) that uses their material models directly.

Master Reference: PCB Laminate Datasheet Parameters at a Glance

Parameter	Symbol	Units	What It Controls	Critical Threshold
Dielectric Constant	Dk (εr)	Dimensionless	Impedance, signal velocity	Stability over freq/temp
Dissipation Factor	Df (tan δ)	Dimensionless	Insertion loss, signal attenuation	<0.004 for GHz-range RF
Glass Transition Temp	Tg	°C	Dimensional stability; assembly survival	≥170°C for lead-free
Decomposition Temp	Td	°C	Chemical stability ceiling	≥340°C for lead-free
Time to Delamination	T260 / T288	Minutes	Assembly thermal endurance	T260 ≥30 min for production
CTE (Z-axis)	αz	ppm/°C	Via barrel fatigue under thermal cycling	<3.5% total (50–260°C)
CTE (X/Y-axis)	αx, αy	ppm/°C	Dimensional stability, component alignment	12–20 ppm/°C typical
Peel Strength	—	lb/in or N/mm	Copper adhesion, trace reliability	>1.0 lb/in (uncured)
Moisture Absorption	—	% weight	Electrical stability, delamination risk	<0.2% for most; <0.1% for outdoor
Thermal Conductivity	k	W/m·K	Heat dissipation through substrate	>0.5 W/m·K for thermal management
CAF Resistance	—	Pass/fail %	Insulation reliability; high-density designs	Pass at design via pitch
Breakdown Voltage	—	V/mil or kV	High-voltage insulation integrity	>1000 V/mil (standard)
Flammability	UL94	Rating	Safety, regulatory compliance	V-0 for most commercial

Common Mistakes When Interpreting PCB Laminate Datasheets

Using 1 MHz Dk for a multi-GHz design. FR-4 Dk at 1 MHz is frequently quoted and irrelevant for any design operating above 1 GHz. Always verify the frequency at which electrical properties were measured, and extrapolate or request data at your actual operating frequency.

Assuming high Tg means high Td. These parameters are related but independent. A material can have excellent Tg and mediocre Td. Both must be verified for lead-free and high-reliability applications.

Ignoring the test method footnote. When comparing Tg across materials from different suppliers, the method used (DSC vs. TMA vs. DMA) produces different numbers for the same material. DSC and TMA values for the same laminate can differ by 10–15°C.

Treating datasheet Dk as the production Dk. Stackup Dk depends on the actual glass style used in each layer and its resin content after pressing. The datasheet nominal value is a starting point, not an end-point. Request fabricator-specific press-out data for each glass style in your stackup.

Not checking Z-axis CTE separately from X/Y CTE. Many datasheets report all three axes but make X/Y the prominent value. Z-axis CTE is the via reliability parameter — never substitute an X/Y value for it.

Useful Resources for Reading and Comparing PCB Laminate Datasheets

• IPC-4101 Standard (Base Materials for Rigid PCBs): Available from ipc.org — the industry reference that defines slash sheets classifying laminate materials by their parameter combinations. Understanding IPC-4101 slash sheet notation is the fastest way to specify a material unambiguously on a fabrication drawing.
• Isola PCB Stackup Designer: Free tool at isola-group.com — uses frequency-dependent, resin-content-dependent material models rather than datasheet nominal values, giving more accurate impedance and loss predictions.
• Rogers Laminate Properties Tool: Interactive database at rogerscorp.com — filterable by Dk, Df, Tg, and other parameters, with frequency-swept data for all Rogers products.
• Sierra Circuits PCB Material Selector: Free comparison tool at protoexpress.com/tools/pcb-material-selector — lets you filter by Tg, Dk, Df, T260, CAF resistance, and other parameters simultaneously across hundreds of materials from multiple suppliers.
• Isola “Making Sense of Laminate Dielectric Properties” (PDF): Available from isola-group.com — an authoritative technical paper explaining how resin content and test method affect Dk/Df values, with measurement methodology comparison. Essential reading for anyone comparing datasheets across manufacturers.
• IPC-TM-650 Test Methods Manual: Free access at ipc.org/test-methods — the source document for all test method numbers referenced on laminate datasheets. Looking up the actual test procedure for a parameter you don’t understand is more reliable than any secondary summary.
• Siemens Z-planner Enterprise: Advanced stackup tool with a 200+ laminate family database containing frequency-dependent material models for accurate loss and impedance calculation. Supported by a free trial.
• Ventec PCB Material Datasheets: Ventec’s material library includes the VT-901 polyimide, VT-4B thermal IMS series, and a range of high-Tg FR-4 products, each with complete IPC-TM-650 referenced data tables — a useful cross-reference for verifying parameter ranges across a supplier with strong application engineering support.

5 FAQs on How to Read a PCB Laminate Datasheet

Q1: Why does my fabricator quote a different Dk from what the laminate datasheet says?

Because the datasheet Dk is typically measured on a flat pressed sample at a specific resin content (often 50% by IPC convention), and the Dk of your actual stackup layer depends on the glass style (1080, 2116, 7628) and the pressed resin content after lamination at your fabricator’s specific press cycle. A 2116 prepreg at your fabricator may press to 48% resin content, giving a Dk slightly different from the 50% datasheet value. This is normal and expected — ask your fabricator for their measured “press-out” Dk values for each glass style they plan to use. Any fabricator doing controlled-impedance work should have this data from their own characterisation boards.

Q2: What’s the difference between Tg, Td, T260, and T288? Why do I need to check all four?

These are four distinct measurements of different failure mechanisms. Tg is where the material softens (reversible, affects Z-axis expansion). Td is where it chemically decomposes (irreversible, sets the process ceiling). T260 and T288 measure time to delamination — how long the laminate survives at reflow-relevant temperatures before internal layers separate. A material could have a good Tg and still have a short T260 if its adhesion chemistry is weak at elevated temperatures. For lead-free production, all four parameters matter: Tg ≥ 170°C keeps the material structurally sound during reflow; Td ≥ 340°C prevents chemical degradation; T260 ≥ 30 min ensures survival through standard assembly operations. Checking only Tg and assuming the others are adequate is the most common thermal specification error.

Q3: My design operates at 3.5 GHz. The datasheet only shows Dk and Df at 1 MHz. Is that good enough?

No. For any design above 1 GHz, you need Dk and Df measured at or near your operating frequency. Df particularly worsens with frequency for most materials — standard FR-4 Df of ~0.010 at 1 MHz becomes 0.020–0.025 at 5 GHz. Using the 1 MHz number for a 3.5 GHz impedance or loss calculation will underestimate both Dk (affecting your trace width calculation) and Df (underestimating insertion loss). Contact the laminate supplier’s application engineering team and request frequency-swept data up to at least 5–10 GHz. Many suppliers provide this data on request even when it’s not in the standard datasheet.

Q4: How do I verify that a laminate my fabricator procured actually matches the datasheet I specified?

The safest approach is to require incoming material certification from your fabricator — a laminate supplier’s certificate of conformance stating the specific product name, lot number, and key measured properties (Tg, Td, Dk at specified frequency). For high-reliability or safety-critical programs, incoming inspection can include DSC measurement of a sample to verify Tg and TGA measurement to verify Td, using IPC-TM-650 methods. For controlled-impedance designs, the most practical verification is impedance test coupon measurement (TDR) on every production panel — if Dk were significantly off from specification, the impedance results would flag it. The combination of material certification plus TDR impedance verification is the standard practice for automotive, aerospace, and high-reliability programs.

Q5: When a datasheet lists both “design Dk” and “process Dk,” which do I use?

Use “design Dk” (also sometimes called “electrical Dk”) for impedance calculations and signal integrity simulation. Use “process Dk” (manufacturing Dk) as the target your fabricator uses when calculating trace widths to hit the target impedance. Rogers Corporation datasheets distinguish between these explicitly for their RO4000 series materials — the design Dk for RO4350B is 3.48 while the process Dk used for manufacturing is 3.66. Using the wrong value in your simulation models introduces systematic error. If your SI tool doesn’t allow you to distinguish between the two, use the design Dk and ensure your fabricator’s impedance model uses their process-specific value.

Conclusion: The Datasheet Is a Starting Point, Not a Final Answer

Learning how to read a PCB laminate datasheet properly is genuinely useful engineering competence — not in the academic sense, but in the sense of catching problems before they become expensive prototypes that fail for reasons that were documented in plain numbers on a page you didn’t read carefully enough.

The most important shift in mindset is understanding that datasheet values are conditions-dependent measurements, not fixed material constants. Dk changes with frequency, temperature, and resin content. Df worsens at higher frequencies. CTE changes above Tg. Tg measurement varies with test method. None of these dependencies are hidden — they’re all derivable from the datasheet and the referenced test methods. The engineer’s job is to apply each parameter to the actual operating conditions, not the standard measurement conditions.

Know which parameters matter most for your specific design — thermal for assembly reliability, Dk/Df for signal integrity, Z-axis CTE for via reliability — and read those sections of every datasheet with the depth they deserve. The parameters you skip when selecting a material have a way of becoming the root cause analysis when a product fails.