July 27, 2026
As HDI density keeps climbing and wearables and robotics modules fight for every square millimeter, embedded passives — moving components from the surface into the board's inner layers — are shifting from a niche option to a standard weapon for high-density design. They free up surface area, eliminate solder joints, and cut parasitics. The price: wider tolerances, longer lead times, and several extra design checkpoints. This article organizes the design essentials, selection logic, and manufacturing realities of embedded resistors and capacitors.
Embedded components are resistors and capacitors (occasionally inductors and ICs) fabricated directly into the PCB substrate with conductive and dielectric materials rather than soldered on the surface. Versus SMT, they offer three structural advantages: space savings (surface area freed for actives and routing), no solder joints (better resistance to vibration and thermal cycling), and lower parasitics (smaller decoupling loop inductance, better signal integrity). RF performance holds up under test — an attenuator built from embedded resistors performed well up to 60 GHz, and resistive material under traces added only minor loss at 40–50 GHz.
The pragmatic rule: choose SMT when budget dominates; accept embedded's cost and process complexity only when space, reliability, thermal, or signal quality are hard requirements.
The stackup is the first battleground. Mainstream layout tools (Altium, Allegro, CADSTAR) all support embedded components, and a few rules apply broadly:
Resistors and capacitors differ in practice: resistors can sit directly on an existing signal layer — no window layer, no extra vias — while capacitors are best placed in a window layer near the reference plane. Thickness impact is minimal: a 10 Ω/square resistive material adds only about 1 μm, and embedded capacitance can actually reduce total thickness by letting power and ground planes sit closer together.
Impedance continuity is where embedded designs most often go wrong on high‑speed boards. Key points:
Loss from the resistive layer itself is only fractions of a dB per inch — acceptable, which is why embedded resistors are widely used on PTFE‑based high‑frequency boards.
Embedded resistance follows R = (ρ × L) / (W × T) , so value tuning is a matter of adjusting length and width — but tolerance is unavoidable: base tolerance runs about ±5%, and etching plus registration variation pushes the total to a typical 15–20%.
Three design‑side countermeasures:
Two more selection parameters: choose TCR below 100 ppm/°C for temperature stability, and after computing dissipation with P = I²R, derate power capacity by at least 25%. Thick film for high power, thin film for precision, plated resistors for high‑current and high‑frequency work.
The biggest value of embedded capacitance is decoupling: a planar capacitor directly under an IC's power pins has far lower loop inductance than a surface SMT part. For selection:
Layout strategy: prioritise low ESR, place capacitors adjacent to IC power pins, distribute them across multiple layers to cover a wider noise spectrum, combine embedded and discrete SMT capacitors into a complementary decoupling network, and maintain solid, continuous power planes.
Four things to align with your fabricator before design freeze:
When we help customers evaluate embedded designs, three misjudgements come up most often:
The right use of embedded passives is trading tolerance headroom for space and reliability where the circuit can afford it — not replacing SMT wholesale.
eCloud provides prototyping and DFM pre‑review services from standard multilayer boards through HDI and high‑frequency designs, with hands‑on familiarity with embedded‑passive stackup planning and manufacturing constraints. Evaluating embedded design feasibility? Talk to our engineering team at the layout stage to confirm tolerances, stackup, and fab‑note completeness — before the issues surface at prototyping.