A microwave PCB substrate determines more than the material listed on a fabrication drawing. Its dielectric properties, thickness, copper construction, and interaction with the manufacturing process influence whether an RF circuit meets its loss, impedance, and phase requirements. A laminate that works well in one stack-up may be unsuitable for another.
For RF engineers and technical sourcing teams, the useful question is which material and construction can meet the application’s requirements at an acceptable production cost. Selecting the lowest loss tangent alone does not answer that question. This guide compares practical substrate options, explains which specifications need careful interpretation, and shows how to connect material selection with fabrication, prototype validation, and an accurate request for quotation.
Select the substrate against a defined circuit requirement: operating band, permitted transmission loss, impedance, phase or resonance stability, power dissipation, and environmental conditions. Frequency narrows the options, but frequency alone does not determine the material.
A short interconnect and a long antenna feed network can need different loss budgets at the same frequency. A compact resonator may place different demands on dielectric constant than a broadband transmission line. Establish the critical electrical structures before comparing laminate names.
Also distinguish board laminates from semiconductor packaging. An IC substrate connects and supports a semiconductor die within a package; selecting that structure is a separate task from selecting the microwave PCB carrying the packaged device. When both are involved, define the electrical interface between them.
Use material families to create a shortlist, then compare exact grades and constructions. The following table provides qualitative screening guidance, not guaranteed performance or universal frequency limits.
| Material family or construction | When to investigate it | Main trade-off | What to confirm before selection |
|---|---|---|---|
| Conventional FR-4 glass/epoxy | Short RF paths where modeled and measured performance can meet the budget | Accessible processing, but “FR-4” alone does not define adequate RF properties | Exact grade, frequency-dependent data, construction, and repeatability |
| Hydrocarbon/ceramic RF laminate | Designs needing controlled dielectric properties with familiar PCB processing | Practical fabrication balance; performance still depends on foil and stack-up | Exact laminate, bonding system, dielectric thickness, and available copper |
| Reinforced PTFE laminate | Applications where dielectric loss warrants a specialized construction | Attractive electrical properties with additional processing considerations | Fabricator experience, hole preparation, handling, and dimensional control |
| Ceramic-filled PTFE | Designs needing a particular combination of permittivity and stability | Formulations differ substantially; one grade cannot represent the family | Grade-specific electrical, thermal, and fabrication data |
| Hybrid RF/digital stack-up | Assemblies combining RF paths with digital control and power distribution | Potential material savings with additional integration and qualification work | RF field boundaries, bonding compatibility, stack symmetry, and transition design |
For a concrete example, the official Rogers RO4350B laminate specifications list a typical dissipation factor of 0.0037 at 10 GHz. Rogers also states that RO4350B uses standard epoxy/glass processing methods without the special through-hole treatments associated with PTFE materials. These are grade-specific statements, not guarantees for every hydrocarbon/ceramic laminate. rogerscorp.com
The RT/duroid 5870/5880 data sheet identifies RT/duroid 5880 as a glass-microfiber-reinforced PTFE composite, with process Dk specified as 2.20 ± 0.02 and typical dissipation factor of 0.0009 at 10 GHz. The sheet distinguishes specification values from typical properties. These figures describe the laminate under stated test conditions; they do not predict finished-board insertion loss by themselves. rogerscorp.com
When discussing the required PCB products, identify the material family, exact grade, and proposed construction separately. A supplier recognizing a brand name has not yet confirmed the requested thickness, foil, or multilayer configuration.
Use dielectric data appropriate to the model, frequency, and field orientation. Process Dk, design Dk, and effective permittivity describe different things and should not be substituted without justification.
Process Dk supports material characterization and quality control under a defined test method. Design Dk is supplied to help represent circuit behavior under stated characterization conditions. For example, Rogers lists RO4350B process Dk as 3.48 ± 0.05 and typical design Dk as 3.66. The two entries do not indicate inconsistent material quality. rogerscorp.com
Rogers’ technical paper, Understanding Dielectric Constant for Microwave PCB Materials, explains why a measured dielectric constant depends on the test method and its relationship to the application. Record the source, frequency, and measurement basis alongside every Dk used in a simulation. rogerscorp.com
Effective permittivity describes propagation in the actual transmission-line structure. In microstrip, fields occupy both the substrate and the surrounding medium, so effective permittivity is not simply the bulk material Dk. Coatings, adjacent conductors, and geometry can change the field distribution.
A higher Dk can help reduce the physical size of some resonant structures; a lower Dk can suit other performance and geometry objectives. Neither direction is universally better. During PCB design, compare the resulting dimensions, manufacturable spacing, and sensitivity to tolerances rather than selecting the smallest Dk on a chart.
Loss tangent, also called dissipation factor or Df, describes dielectric dissipation. Finished RF performance also depends on conductor loss, mismatch, radiation, and the transitions included in the measurement.
John Coonrod of Rogers demonstrates in Selecting Circuit Materials for Microwave Power Amps that the relative importance of dielectric and conductor loss changes with circuit thickness. The article also compares the influence of copper roughness. This supports evaluating the complete construction instead of ranking materials solely by Df. Microwave Journal
Before paying for a lower-loss dielectric, ask which mechanism consumes the loss budget. If connector launches or conductor construction dominate, a laminate change may deliver less benefit than expected. If long RF paths dominate dielectric loss, that same change may be worthwhile.
For copper roughness and RF loss, request the actual foil option and the roughness data used by the model. Avoid treating different roughness metrics as interchangeable. Keep surface finish and plating assumptions consistent between candidate stack-ups so the comparison isolates the intended variable.
The relevant dielectric thickness is the separation between the signal conductor and its reference conductor, not automatically the total finished-board thickness. State both values in the stack-up.
For a selected transmission-line geometry, changing dielectric height changes the dimensions needed to achieve a target impedance. Very narrow conductors can make a fixed etching variation more significant relative to nominal width. Increasing thickness changes other electrical and mechanical constraints, so “thinner is always better” is not a reliable purchasing rule.
Ask the fabricator to review finished conductor width, copper thickness, dielectric thickness, and achievable variation together. A nominal impedance calculation cannot demonstrate production capability. Request a tolerance analysis for sensitive structures rather than assigning every dimension the tightest available limit.
Separate the release into material requirements, finished-geometry requirements, and electrical acceptance requirements. That distinction lets engineering and purchasing identify whether an apparent price difference comes from laminate choice, process control, or verification scope.
Review thermal conductivity, thermal expansion, temperature-dependent dielectric behavior, and moisture exposure as separate considerations. They address different risks and cannot be replaced by a single “high-temperature material” description.
For a power amplifier, evaluate the heat path from the device through its attachment, copper, vias or other thermal structures, and enclosure. A low-loss laminate can reduce one source of dissipation without necessarily providing the required device-to-chassis thermal path.
For a phase-sensitive network or resonator, request dielectric behavior across the actual operating temperature range. For multilayers and plated holes, ask how the selected construction will be qualified for the specified thermal exposure. Do not turn a laminate’s nominal property into a guarantee for the assembled product.
Define storage, humidity, coating, and enclosure conditions before validation. Where environmental stability matters, agree which electrical measurements will be repeated after conditioning. The acceptance plan should describe the exposure and the result required, rather than using an undefined requirement such as “suitable for outdoor RF.”
A hybrid construction is worth evaluating when premium RF material can be confined to the electromagnetic structures that need it. Savings depend on the complete fabrication and validation route, not simply the area of expensive laminate removed.
Confirm where the fields propagate. For embedded transmission lines, more than one surrounding dielectric layer can matter. Replacing a bonding layer or nearby core may change the electrical model even if the signal conductor remains on the original RF laminate.
Consider an illustrative controller board with a localized RF front end. Compare an all-RF construction against a hybrid design using the same RF acceptance criteria, quantities, and delivery assumptions. Include bonding materials, extra process steps, test coupons, and qualification effort. The hybrid is useful only if the quoted and validated result supports the saving.
Compare costs per accepted board. A lower bare-board price can be offset by added tuning, testing, or requalification. Conversely, specifying premium material throughout a board can add cost where the electrical design does not need it.
A prototype should test the proposed production construction and the assumptions most likely to affect RF performance. Specify its purpose before ordering a PCB prototype.
Useful validation may include:
Define calibration, measurement reference planes, and any de-embedding method with the test provider. Otherwise, fixture or connector behavior can be mistaken for substrate performance. An impedance coupon alone does not verify the insertion-loss budget or antenna performance.
Preserve the exact laminate, copper option, bonding system, finish, and stack-up revision when moving to production. If a supplier proposes a substitution, identify which model inputs and acceptance results must be reviewed before approval.
Coordinate PCB assembly requirements before releasing the fabrication package. Include the intended reflow route, sensitive component attachment, handling, and permitted rework. Bare-board validation and assembled-product validation answer different questions.
Provide enough information for suppliers to quote the same construction and verification scope. If the material remains undecided, request clearly separated alternatives rather than allowing unspecified substitutions.
| RFQ information | What to provide | Why it affects the quotation |
|---|---|---|
| Electrical requirements | Operating band, impedance, loss budget, and relevant phase or resonance limits | Defines material screening and validation scope |
| Material definition | Manufacturer, grade, thickness, foil option, and substitution policy | Controls procurement and electrical-model assumptions |
| Stack-up | Layer sequence, RF reference planes, bonding materials, and finished thickness | Determines lamination and transmission-line construction |
| Fabrication data | Released CAD/manufacturing files, drill data, geometry tolerances, and finish | Establishes process feasibility and inspection work |
| Thermal and environmental scope | Power conditions, temperature exposure, humidity, and relevant qualification plan | Identifies application-specific risks and tests |
| Acceptance evidence | Coupons, inspection reports, RF measurements, fixtures, and responsibility for testing | Prevents different test scopes from appearing price-equivalent |
| Commercial scope | Prototype and production quantities, delivery destination, required date, and assembly scope | Separates setup, material procurement, testing, and delivery costs |
If the project also includes package-substrate development, request a separate IC substrate capability review for the die-to-package structure. Board-level RF laminate approval should not be treated as approval of package-substrate manufacturing capability.
Before selecting a supplier, request written confirmation of material availability, finished-construction tolerances, and the agreed evidence of acceptance. Use that confirmed scope to compare quotations and authorize the validation build.
FR-4 can be considered when the exact grade, geometry, path length, and operating conditions meet the electrical budget. Validate the proposed construction. A generic FR-4 designation does not establish loss, phase stability, or repeatability at the application frequency.
No. A PTFE grade may offer useful dielectric performance, while a hydrocarbon/ceramic grade may simplify fabrication. Compare exact materials, copper options, and finished stack-ups against the electrical requirements and total procurement cost.
Process Dk is associated with material characterization under a defined test method. Design Dk supports circuit modeling under stated conditions. Use the laminate manufacturer’s guidance, and distinguish both from the effective permittivity of the actual transmission-line structure.
No universal thickness follows from frequency alone. The appropriate dielectric height depends on transmission-line geometry, conductor dimensions, loss, radiation, coupling, and manufacturing constraints. Evaluate candidate constructions with a suitable electromagnetic model and prototype measurements.
Matching a nominal Dk is insufficient. Compare measurement conditions, loss, thickness, foil, thermal behavior, and bonding compatibility. Review the electrical model and qualification requirements before approving any substitution in a released RF design.
Choose measurements tied to the design risks: impedance, insertion loss, return loss, phase, or resonance. Define fixtures and reference planes. Where assembly or environmental exposure matters, include the corresponding conditions in the validation plan.
Lorem ipsum dolor sit amet consectetur adipiscing elit dolor
Learn how PCB impedance coupons, TDR measurements, tolerance limits, and traceable reports support fabrication acceptance and more accurate supplier RFQs.
Compare flying probe and bed-of-nails for bare multilayer PCBs, including electrical coverage, netlist control, probe access, fixture economics, and RFQ requirements.
Learn the differences between PCB surface preparation, wet processing, and final finishes, with practical guidance on plating terminology, inspection evidence, and RFQ requirements.
