Eight substrate families compared on dielectric constant, loss, thermal behaviour and cost — and the order of decisions that picks the right one.
Microwave boards, by definition, work above 1 GHz, and at those frequencies the substrate stops being packaging and becomes part of the circuit. The materials used fall into a handful of families, and each family trades electrical performance against mechanical behaviour, processability and cost in a different way. Knowing those trade-offs is most of the selection work — the underlying physics is the same microwave and RF material property set that applies to every board in this category. What follows is a material-by-material comparison.
Pure PTFE is the reference point for low loss. Dk sits between 2.1 and 2.2 and remains stable over a wide frequency range, while Df is very low, typically 0.0005 to 0.002, which keeps attenuation minimal in high-frequency transmission. The disadvantages are physical rather than electrical: low mechanical strength, a tendency to deform, and processing that demands specialised techniques and tight control. It is also expensive.
Ceramic-filled PTFE addresses the mechanical weakness without giving up the electrical advantage. Adding ceramic filler pushes Dk into the 2.5 to 10 range depending on filler type and content, which gives a designer a way to tune impedance, while Df stays low at roughly 0.001 to 0.005. The result is a substrate that suits radar and wireless systems needing mechanical stability as well as low loss. The cost is process difficulty and some risk to material uniformity. Our notes on PTFE glass fibre cloth laminate cover a related approach to reinforcing the same chemistry.
These blend ceramic particles into a polymer matrix rather than using PTFE as the base. Dk is adjustable between roughly 3.0 and 10, which makes them useful across a broad band of designs, and the electrical and mechanical performance is balanced enough for a wide range of frequency applications. Thermal handling is good. Cost is moderate to high, and while they process more easily than PTFE, they still need process control to hold consistency.
Polyimide is chosen for thermal stability rather than for the lowest possible loss. Dk typically falls between 3.5 and 4.5 with a relatively low dissipation factor, and the material resists chemical attack, which suits harsh high-frequency environments. The higher Dk is the trade-off: it limits performance in some applications compared with PTFE and ceramic materials.
Cyanate ester sits between epoxy and polyimide in character. Dk ranges around 3.0 to 3.8 and Df between 0.002 and 0.006, giving good electrical performance alongside high thermal stability and low moisture absorption. Its low coefficient of thermal expansion matches copper well, which reduces delamination risk during thermal cycling — a meaningful advantage in high-reliability builds. The cost is high, and manufacturing needs special handling to keep performance consistent.
The RO4000 series covers Dk between 3.2 and 3.8 with a low loss factor of 0.002 to 0.004, and it is widely used in microwave and RF applications because it processes much like FR-4. The RO3000 series goes further on loss and stability, which makes it the choice for higher-frequency work where insertion loss is the binding constraint. The range of options is the main advantage: a designer can match frequency and loss requirements without leaving the family. The trade-offs are price relative to standard materials (though below PTFE), and the need to choose carefully, because performance and processing requirements differ between series.
PPE substrates offer a Dk between roughly 2.8 and 3.5 with Df in the 0.002 to 0.003 region, and they combine good thermal stability with mechanical properties that are easy to work with. They suit medium to high-frequency microwave applications where cost and manufacturability matter. Electrical performance falls slightly short of PTFE, but the gap in processability is larger than the gap in loss.
Modified epoxy is the cost-conscious option for lower frequency work. Dk sits between 3.0 and 4.0 with low loss factors, and thermal and mechanical performance are considerably better than conventional epoxy. Process compatibility is the selling point: the material runs close to standard FR-4 practice. Its limitations appear at higher frequencies, where stability does not match PTFE or ceramic-enhanced laminates.
| Material | Typical Dk | Typical Df | Best fit |
|---|---|---|---|
| PTFE | 2.1–2.2 | 0.0005–0.002 | Lowest-loss microwave and millimeter-wave circuits |
| Ceramic-filled PTFE | 2.5–10 | 0.001–0.005 | Radar and wireless systems needing mechanical stability |
| Ceramic-enhanced composite | 3.0–10 | Low | Broadband designs with mixed electrical and mechanical needs |
| Polyimide | 3.5–4.5 | Relatively low | Harsh, high-temperature environments |
| Cyanate ester | 3.0–3.8 | 0.002–0.006 | High-reliability, low-loss, low-moisture builds |
| Rogers RO4000 | 3.2–3.8 | 0.002–0.004 | Microwave and RF work with FR-4-like processing |
| Rogers RO3000 | Lower Dk grades available | Very low | Higher-frequency designs where loss is the constraint |
| PPE | 2.8–3.5 | 0.002–0.003 | Medium to high frequency with cost sensitivity |
| Modified epoxy | 3.0–4.0 | Low | Low to medium frequency, FR-4-like manufacturing |
The decision usually resolves in this order: loss budget first, then thermal environment, then mechanical and process requirements, then cost. A millimeter-wave link that cannot tolerate attenuation goes to PTFE or a low-loss series such as the Rogers 3000 family; a radar board that must survive vibration and heat goes to ceramic-filled PTFE; a cost-driven design at a few GHz can stay with modified epoxy. The complete guide to laminate selection covers how these families sit against each other in practice, and it is worth reading before a specification is frozen rather than after.
Frequently Asked Questions
Pure PTFE, with Dk around 2.1 to 2.2 and Df between 0.0005 and 0.002. Its weakness is mechanical, not electrical: low strength, easy deformation and a demanding process route.
To fix the mechanical weakness. The filler raises Dk into a tunable 2.5 to 10 range and improves thermal and dimensional stability while keeping loss low, which is why it dominates radar and wireless hardware.
When thermal and chemical survival outrank minimum loss. Polyimide works in harsh environments, but its higher Dk of 3.5 to 4.5 limits it in the most loss-sensitive designs.
Its low coefficient of thermal expansion matches copper well, which reduces delamination during thermal cycling. That makes it attractive in high-reliability builds despite its cost and handling requirements.
At low to medium frequencies, yes. Dk falls between 3.0 and 4.0 with low loss, and it processes much like FR-4. Above that range its stability lags PTFE and ceramic-enhanced materials.
Send us the frequency range, loss budget and thermal requirements and we will shortlist the materials that fit and quote the build.
Get a Fast Quote Fast Reply on WhatsApp