PCB laminate Tg, Td, CTE, Dk, Df explained with Nanya material examples. Understand how each property affects reliability, signal integrity, and laminate selection.

Every PCB laminate datasheet throws the same wall of numbers at you — Tg, Td, CTE, Dk, Df — and a lot of engineers either memorize the rules of thumb without fully understanding them or skip past the details entirely and trust that whatever the fab recommends is good enough. That works until your board fails a qualification test, delaminates during rework, or shows unexpected signal loss at 10 GHz.

This article is a practical explanation of each of those properties, grounded in how they actually affect design outcomes on real boards. The context is Nanya PCB laminates, which give us a concrete product family to anchor the discussion — but the underlying physics applies to any laminate family you’re evaluating.

Why These Five Properties Are the Ones That Actually Matter

Laminate datasheets list dozens of properties. Flexural strength, peel strength, arc resistance, volume resistivity — these all matter in specific contexts, but they’re rarely the reason a design succeeds or fails. The five properties in the title of this article — Tg, Td, CTE, Dk, and Df — show up on the critical path of almost every PCB reliability and performance problem I’ve seen in engineering practice.

Tg and Td tell you how the material behaves under thermal stress. CTE tells you what happens to your board dimensionally as temperature changes. Dk determines your impedance and signal velocity. Df determines how much signal power gets converted to heat in your dielectric as frequency increases. Understanding all five, and understanding how they interact, gives you the foundation for laminate selection that actually holds up through fabrication, assembly, and field deployment.

Tg: Glass Transition Temperature — The Property Most Engineers Misunderstand

What Tg Actually Means Physically

Tg, the glass transition temperature, is the temperature at which the epoxy resin in a PCB laminate transitions from a rigid, glassy state to a softer, rubbery state. Below Tg, the resin is hard and dimensionally stable. Above Tg, it becomes viscoelastic — it can deform more easily, and its coefficient of thermal expansion in the Z-axis increases sharply, often by a factor of three to five.

The critical thing to understand is that crossing Tg is not destructive by itself. A board that temporarily exceeds its laminate’s Tg during lead-free reflow doesn’t immediately fail. But operating repeatedly or continuously above Tg accelerates via fatigue, promotes delamination, and degrades long-term reliability. The resin doesn’t decompose at Tg — that’s a different property (Td) — it just softens.

How Tg Is Measured (and Why the Measurement Method Matters)

Here’s something that trips up a lot of purchasing and engineering teams: the same laminate can show significantly different Tg values depending on which test method is used. Three methods are in common use:

Test Method	Standard	Typical Result vs. DSC	Notes
DSC (Differential Scanning Calorimetry)	IPC-TM-650 2.4.25	Baseline reference	Measures heat capacity change at transition
TMA (Thermomechanical Analysis)	IPC-TM-650 2.4.24	~10–15°C lower than DSC	Measures dimensional change vs. temperature
DMA (Dynamic Mechanical Analysis)	IPC-TM-650 2.4.24.2	~10–20°C higher than DSC	Measures stiffness change vs. temperature

When you’re comparing Nanya laminates to each other or to competitors, confirm that you’re comparing Tg values from the same test method. Nanya typically reports both DSC and TMA values on their datasheets, and the difference between them — usually 10–15°C — is not an error; it reflects the physics of the different measurement approaches.

Nanya Laminate Tg Values: What the Numbers Mean in Practice

Laminate Grade	Tg (DSC, °C)	Operating Margin at 85°C	Lead-Free Reflow Suitability
Standard FR-4	130–140	45–55°C	Marginal (multiple cycles risky)
NPG	155–175	70–90°C	Good
NPGN	155–175	70–90°C	Good
NP-822	200+	115°C+	Excellent
NP-930 (PTFE base)	N/A (no Tg)	N/A	Requires special process

PTFE-based materials like the NP-930 don’t have a Tg in the conventional sense — PTFE is semi-crystalline, not amorphous, so the glass transition concept doesn’t apply. That’s actually an advantage for extreme temperature applications.

The practical takeaway: if your board operates above 85°C continuously, or if it goes through more than two reflow passes in assembly, standard FR-4 at Tg 130–140°C is starting to look thin on margin. The upgrade to NPG or NPGN (Tg 155–175°C) costs a modest premium and gives you substantially more headroom.

Td: Thermal Decomposition Temperature — The Limit You Really Don’t Want to Reach

Td vs. Tg: Understanding the Difference

Where Tg is a reversible softening transition, Td is the point at which the resin begins to chemically decompose — breaking down polymer chains, releasing volatile gases, and permanently degrading the laminate structure. Td is measured by thermogravimetric analysis (TGA) and is typically defined as the temperature at which the material loses 5% of its initial mass.

Once a laminate exceeds Td, the damage is permanent and non-recoverable. In practice, decomposition shows up as measurable blistering, delamination within the laminate, and outgassing that can cause voids in plated through-holes during reflow.

T260 and T288: The More Practical Thermal Stress Tests

While Td tells you where decomposition begins, the industry uses two related tests that are more directly meaningful for manufacturing process evaluation:

T260 measures how long a laminate can survive at 260°C before delamination occurs. This is directly relevant to lead-free solder reflow, which peaks at approximately 255–260°C.

T288 uses 288°C, which is relevant for rework operations and certain military/aerospace qualification tests.

Property	Nanya FR-4	Nanya NPG	Nanya NPGN
Td (°C, TGA 5% loss)	~300–315	~330–345	~335–355
T260 (minutes to delamination)	>30	>60	>60
T288 (minutes to delamination)	~5–10	>15	>20

The significance here is straightforward. A board that goes through multiple lead-free reflow cycles, or that requires rework with a soldering iron applied for extended periods, is being stressed toward or above 260°C repeatedly. A laminate with T260 of 30 minutes might survive a single reflow cycle but accumulate damage over multiple cycles. The NPG and NPGN laminates with T260 of 60+ minutes give you meaningful process margin.

When Td Becomes a Critical Design Constraint

The Td becomes a first-order constraint in high-density assemblies where rework is likely, in military and aerospace hardware where extended solder process times are sometimes required, and in any board design that includes multiple heavy copper pours that act as heat sinks and extend local temperature dwell times during reflow. For standard consumer electronics assembled once and never reworked, Td margin is less of a daily concern.

CTE: Coefficient of Thermal Expansion — The Silent Killer of Via Reliability

Understanding Z-Axis CTE and Why It’s Different from X/Y

CTE describes how a material expands and contracts with temperature change. For PCB laminates, there are actually three relevant CTE values — X-axis, Y-axis, and Z-axis — and they behave very differently because the fiberglass reinforcement constrains expansion in the X/Y plane while the Z-axis is largely controlled by the resin alone.

X/Y CTE for FR-4 type laminates is typically 12–18 ppm/°C, which is reasonably matched to the 17 ppm/°C of copper. This is why trace and plane reliability in the X/Y plane is rarely the failure mode.

Z-axis CTE is the problem. Below Tg, standard FR-4 runs about 55–70 ppm/°C in Z. Above Tg, that number jumps to 200–300 ppm/°C. Copper’s CTE remains at 17 ppm/°C through all of this. The via barrel — which is copper bonded to the drilled hole wall in a Z-axis column through the board — is being stretched and compressed dramatically every thermal cycle, relative to the laminate. Over thousands of cycles, the copper fatigues and cracks.

CTE Impact on Via Reliability: A Quantitative Look

Consider a typical multilayer board with 1.6mm thickness and plated through-holes:

Material	Z-CTE below Tg	Z expansion per 100°C ΔT across 1.6mm
Copper barrel	17 ppm/°C	2.7 µm
Standard FR-4	60 ppm/°C	9.6 µm
Nanya NPG	50 ppm/°C	8.0 µm
Nanya NPGN	45 ppm/°C	7.2 µm
Differential strain (FR-4 vs Cu)	—	6.9 µm per 100°C
Differential strain (NPGN vs Cu)	—	4.5 µm per 100°C

Over thousands of thermal cycles in a product with a 10-year service life, the reduced differential expansion of NPGN versus standard FR-4 accumulates into a significant via fatigue margin improvement. For vias at high aspect ratio (board thickness / drill diameter greater than 8:1), this becomes critical.

Why CTE Mismatch Also Affects BGA and Fine-Pitch Component Reliability

Beyond via reliability, Z-axis CTE affects solder joint fatigue under BGA packages. When the laminate expands more than the chip package during thermal cycling, the solder balls at the outer corners of the BGA are strained diagonally — both in X/Y from the package-substrate CTE mismatch and in Z from the laminate’s Z-axis expansion. This is why PCB-level CTE characterization is part of the component qualification process for automotive and aerospace-grade assemblies.

Dk: Dielectric Constant — Your Impedance and Signal Velocity Control

What Dk Controls in PCB Signal Behavior

Dk, the dielectric constant (also called relative permittivity, εr), quantifies how much the substrate material slows down electromagnetic wave propagation compared to free space. Higher Dk means slower signal propagation velocity and shorter wavelengths at a given frequency. This directly affects two critical design parameters:

Controlled impedance: The characteristic impedance of a microstrip or stripline trace depends on the Dk of the surrounding dielectric. A higher Dk makes a given trace width produce a lower impedance. When you specify a 50-ohm microstrip on a 0.5mm dielectric, you’re solving an equation that includes Dk as a primary variable. If the Dk varies across a panel or between production lots, your impedance varies with it.

Signal propagation velocity: The speed of a signal in a dielectric is c / √Dk, where c is the speed of light. In FR-4 at Dk = 4.4, signals travel at approximately 48% of the speed of light. This affects timing, especially in length-matched differential pairs and clock distribution networks.

Dk Frequency Dependence and Why the Test Frequency on the Datasheet Matters

Dk is not constant with frequency — it decreases as frequency increases, because at higher frequencies the polarization mechanisms in the resin can’t fully respond to the oscillating field. Standard FR-4 Dk measured at 1 MHz is typically 4.4–4.8. At 1 GHz it’s closer to 4.2–4.5. At 10 GHz it might be 4.0–4.3.

This dispersion means that when you’re comparing laminates for high-frequency applications, the Dk value at your operating frequency is what matters — not the 1 MHz value that’s often listed most prominently on datasheets.

Nanya Laminate	Dk @ 1 MHz	Dk @ 1 GHz	Dk @ 10 GHz
Standard FR-4	4.4–4.8	4.2–4.5	4.0–4.3
NPG	4.2–4.6	4.0–4.3	3.8–4.1
NPGN	4.0–4.5	3.9–4.2	3.7–4.0
NP-530	3.7–3.9	3.6–3.8	3.4–3.6
NP-822	3.3–3.5	3.2–3.4	3.0–3.2
NP-930	3.0–3.1	2.97–3.0	2.94–2.97

How Dk Uniformity Affects Production Impedance Yield

Dk uniformity across a laminate panel and between production lots is often more important than the absolute Dk value, particularly for high-volume production of controlled-impedance boards. A Dk variation of ±0.2 across a panel translates to roughly ±2–3 ohms on a 50-ohm nominal impedance line — which is the difference between shipping parts and scrapping them.

Nanya specifies Dk tolerance on their product datasheets, and this tolerance narrows as you move up the product family from standard FR-4 toward RF-grade laminates. When qualifying a laminate for high-volume controlled impedance production, it’s worth requesting process capability data (Cpk on Dk) from the laminate manufacturer, not just the nominal specification.

Df: Dissipation Factor — The Property That Sets Your High-Frequency Loss Budget

What Df Physically Represents

Df, the dissipation factor (also called loss tangent, tan δ), measures the fraction of electrical energy that gets converted to heat as an electromagnetic wave propagates through the dielectric. In physical terms, it reflects the imperfect ability of polar molecules in the resin to follow an oscillating electric field — some energy lags behind and dissipates as heat.

The loss contribution from dielectric dissipation (dielectric loss) scales linearly with both frequency and Df. Double the frequency, double the dielectric loss. Double the Df, double the dielectric loss. At low frequencies (below ~100 MHz), dielectric loss is small compared to conductor loss from skin effect resistance. As frequency increases, dielectric loss grows until it dominates the total insertion loss budget.

Calculating Dielectric Loss: A Practical Example

For a rough estimate of dielectric loss in dB/100mm of transmission line:

Dielectric loss (dB/100mm) ≈ 27.3 × Df × √Dk × (f_GHz / c_mm/ns)

Let’s compare standard FR-4 and Nanya NPGN at 10 GHz on a 100mm trace:

Standard FR-4: Df = 0.022, Dk = 4.2 → Dielectric loss ≈ 4.1 dB/100mm

Nanya NPGN: Df = 0.015, Dk = 3.9 → Dielectric loss ≈ 2.6 dB/100mm

Nanya NP-822: Df = 0.0025, Dk = 3.1 → Dielectric loss ≈ 0.42 dB/100mm

The difference between standard FR-4 and NPGN at 10 GHz is 1.5 dB per 100mm of trace. That is a significant margin in a system where your total insertion loss budget might be 6–8 dB including connectors, vias, and length.

Df Comparison Across Nanya Laminates

Laminate	Df @ 1 GHz	Df @ 10 GHz	Suitable Frequency Range
Standard FR-4	0.018–0.025	0.022–0.030	Below ~1 GHz (digital/analog)
NPG	0.015–0.022	0.018–0.025	Below ~3 GHz
NPGN	0.012–0.018	0.015–0.022	Below ~10 GHz (high-speed digital)
NP-530	0.003–0.006	0.004–0.007	Up to ~10 GHz
NP-822	0.002–0.003	0.0025–0.0035	Up to ~30 GHz
NP-930	0.0008–0.0013	0.001–0.0015	Up to 77 GHz+

The jump from NPGN (Df ~0.015–0.022 at 10 GHz) to NP-530 (Df ~0.004–0.007) is where you cross from “digital laminate used at high speed” territory into “purpose-built RF laminate” territory. For signal integrity work in the 1–10 Gbps range, NPGN is viable. For microwave RF applications above 3 GHz, you want to be on the NP-series materials.

How These Properties Interact: The Real-World Picture

These five properties don’t exist in isolation. Here’s how they work together in practice:

Tg affects CTE behavior: Tg sets the knee where Z-axis CTE jumps. A material with higher Tg keeps you in the lower-CTE regime through more of your operating and processing temperature range.

Td and Tg together define your process window: High Tg gives you operating margin. High Td gives you manufacturing margin during assembly. You need both.

Dk and Df both vary with temperature and moisture: A board that absorbs moisture shows elevated Dk and Df — both trace impedance and insertion loss shift. This is why moisture absorption is an indirect indicator of electrical stability in deployed products.

Copper roughness interacts with Df at high frequency: At mmWave frequencies, conductor loss from rough copper surfaces can exceed dielectric loss even with very low-Df materials. When evaluating NP-822 or NP-930, always check what copper foil type (standard, RTF, HVLP, or VLP) the datasheet values were measured with.

Useful Resources for PCB Laminate Property Evaluation

IPC-TM-650 Test Methods Manual (ipc.org) — The authoritative source for how every laminate property in this article is measured; essential for interpreting and comparing datasheet values

IPC-4101D: Specification for Base Materials for Rigid and Multilayer Printed Boards — Defines which slash sheet classification corresponds to which minimum property requirements

Nanya New Material Technology Datasheet Downloads — Always access current revision datasheets directly from Nanya’s official product pages before design sign-off

Polar Instruments Si9000e (polarinstruments.com) — Transmission line field solver that accepts real Dk/Df values and calculates impedance and insertion loss; free trial available

UL Yellow Card Database (ul.com) — Verify UL 94 recognition for specific Nanya laminate grades; useful for compliance documentation

Rogers Corporation PCB Material Selection Guide (rogerscorp.com) — Useful as an industry reference benchmark; their published Dk/Df data across temperature and frequency is well-characterized and helps calibrate expectations for other manufacturers’ materials

NIST Dielectric Properties Database (nvlabs.nist.gov) — Reference data for dielectric properties of base materials; useful for cross-checking laminate datasheets

IPC-9151 PCQR2 Database — Industry benchmark data for fabrication process capability including via reliability across laminate types

Frequently Asked Questions

Q1: My laminate datasheet shows Dk at 1 MHz. Can I use that value for my 5 GHz design?

Not reliably. Dk decreases with frequency, so the 1 MHz value will overestimate the Dk at 5 GHz, which means your impedance calculations will be off and your signal velocity estimate will be pessimistic. For RF and high-speed digital work, you need Dk measured at or near your operating frequency. The IPC-TM-650 2.5.5.5 (clamped stripline resonator) and 2.5.5.9 (full sheet resonance) test methods provide Dk values at specific test frequencies — check which method and frequency your datasheet uses. If only 1 MHz data is available, use it as a rough starting point and plan for empirical validation with impedance coupons.

Q2: We’re using standard FR-4 and experiencing delamination during rework. Is upgrading Tg alone enough to fix this?

Tg upgrade helps, but it depends on where the delamination is occurring. If it’s at the copper-laminate interface, the issue may be inadequate surface preparation or pre-bake before rework rather than bulk Tg. If delamination is happening within the laminate body or at a prepreg-core interface, upgrading to NPG or NPGN with higher Td and better thermal stability is likely to help. Get the delaminated board cross-sectioned and identify exactly where the failure plane is before spending money on a material upgrade.

Q3: How does moisture absorption relate to Dk and Df performance?

Moisture is a polar molecule with a very high dielectric constant (approximately 80). When it absorbs into a PCB laminate, it increases the effective Dk of the composite material — which shifts controlled impedance lines away from their design target — and also increases Df, which raises insertion loss. For boards in outdoor, marine, or industrial environments, moisture absorption is a significant source of in-service electrical performance drift. The NPGN’s ~0.06–0.08% moisture absorption versus standard FR-4’s 0.15–0.20% represents a meaningful electrical stability advantage in humid deployments.

Q4: What CTE value should I target for a board with 0.5mm pitch BGA components?

The CTE mismatch between the laminate and the BGA package is what drives solder joint fatigue. Most fine-pitch BGA packages on silicon or ceramic substrates have X/Y CTE values in the 4–12 ppm/°C range. Standard FR-4’s X/Y CTE of 14–18 ppm/°C creates a mismatch that is managed at 0.5mm pitch mostly through underfill in demanding thermal cycling applications. The laminate Z-axis CTE affects via reliability but not BGA joint fatigue directly. For extreme reliability requirements, controlled-CTE laminates or copper-invar-copper (CIC) cores can be used to reduce X/Y CTE, but these are specialty solutions beyond the standard Nanya product range.

Q5: What’s the most common mistake engineers make when reading laminate datasheets?

Comparing values measured with different test methods, at different frequencies, and with different copper types as if they’re directly comparable. A Dk value measured at 1 MHz by capacitance method is not comparable to a Dk value measured at 10 GHz by stripline resonator. A Df measured on a sample with low-profile (LP) copper foil will look lower than the same material tested with standard electrodeposited copper, because copper surface roughness affects measured Df at high frequency. When evaluating Nanya laminates against competitors, read the test conditions for each value — not just the number — before drawing conclusions.

Putting It All Together: A Property Selection Guide

Design Requirement	Most Critical Property	Secondary Property	Recommended Nanya Grade
Lead-free assembly reliability	Tg + Td	Z-axis CTE	NPG or NPGN
Automotive thermal cycling	CTE + Tg	Td	NPGN
High-speed digital (>10 Gbps)	Df	Dk uniformity	NPGN
RF/microwave (3–30 GHz)	Df + Dk	Dk stability vs. temp	NP-530, NP-822
mmWave (>30 GHz)	Df + Dk	CTE, moisture absorption	NP-930
Halogen-free compliance	Material chemistry	All of the above	NPGN
Cost-optimized consumer product	Cost / performance balance	Tg adequacy	Standard FR-4

These five properties — Tg, Td, CTE, Dk, and Df — tell most of the story of how a laminate will perform in your design. Read the datasheet with all five in mind, understand the test methods behind each number, and you’ll make better laminate decisions than most engineers who are just going off habit or what the last person in their seat specified.

Suggested Meta Descriptions

Option 1 (158 characters):
Learn what Tg, Td, CTE, Dk, and Df mean for Nanya PCB laminates. This engineer’s guide explains each property with tables and real design impact examples.

Option 2 (160 characters):
PCB laminate Tg, Td, CTE, Dk, Df explained with Nanya material examples. Understand how each property affects reliability, signal integrity, and laminate selection.

Option 3 (155 characters):
Confused by Nanya laminate datasheets? This guide explains Tg, Td, CTE, Dk, and Df in plain engineering terms with comparison tables and design guidance.

Recommended: Option 2 — it leads with the target keyword phrase “PCB laminate Tg Td CTE Dk Df explained,” references Nanya directly for brand relevance, and covers the key search intent (understanding what these properties mean and how to use them) within the 160-character limit.