What is self-resonant frequency and why does it matter?

Self-resonant frequency (SRF) is the frequency at which the inductor's parasitic capacitance resonates with its inductance, creating a parallel resonance. Below SRF, the component behaves as an inductor (impedance increases with frequency). At SRF, the impedance reaches a maximum (parallel resonance). Above SRF, the parasitic capacitance dominates and the component behaves as a capacitor (impedance decreases with frequency). The practical rule is to use an inductor only below one-third of its SRF to ensure it behaves predictably as an inductor. For example, a 10 nH chip inductor with SRF of 6 GHz should only be used below 2 GHz. Higher inductance values have lower SRF because more turns create more parasitic capacitance. A 100 nH inductor might have SRF of only 800 MHz, limiting its use to below 250 MHz for predictable inductive behavior.

How does Q factor affect circuit performance?

Q factor measures the ratio of energy stored to energy dissipated per cycle: Q = omega*L / R_series = X_L / R_series. Higher Q means lower loss. In a matching network, inductor Q directly determines the insertion loss: for an L-network, IL approximately equals 1 / (1 + Q_circuit/Q_inductor) in linear terms, or about 4.34 * Q_circuit / Q_inductor dB for high Q_inductor. For a narrowband match with Q_circuit = 5 using inductors with Q = 50, the insertion loss is approximately 0.43 dB. With Q = 20 inductors, the loss increases to 1.09 dB. In filters, inductor Q determines the passband insertion loss and the achievable selectivity. In VCO tank circuits, higher Q reduces phase noise according to Leeson's model (phase noise is proportional to 1/Q^2). Typical Q values at 1 GHz: wirewound chip inductors achieve Q of 40 to 80, multilayer chip Q of 15 to 30, and thin-film Q of 30 to 60.

What inductor types are used at RF?

Wirewound chip inductors (0402 to 0805 packages) use a coil of wire around a ceramic or ferrite core, achieving the highest Q (40 to 100) but with lower SRF due to more parasitic capacitance. They are used for matching networks, bias tees, and filters below 3 GHz. Multilayer chip inductors use printed conductor layers in a ceramic body, achieving moderate Q (15 to 30) with smaller size and lower cost. They are common for decoupling and bias circuits. Thin-film inductors use photolithographic patterning on alumina or silicon substrates, achieving tight tolerance (plus or minus 2 percent) and moderate Q (30 to 60) with high SRF. They are used in precision matching networks. Air-core and printed spiral inductors on MMIC substrates eliminate core losses entirely, achieving usable performance at millimeter-wave frequencies where all other types fail. MMIC spiral inductors of 0.1 to 2 nH are standard for 20 to 100 GHz circuits.

Passive Components

RF Inductor

/in-duk-ter/

Stores energy in a magnetic field with impedance X_L = 2πfL. Parasitic capacitance creates SRF (self-resonant frequency): above SRF, component behaves as a capacitor. Use below SRF/3 for reliable inductive behavior. Q = X_L/R_series determines loss in matching networks, filters, and VCO tanks. Types: wirewound (Q 40-100, low SRF), multilayer chip (Q 15-30), thin-film (Q 30-60, tight tolerance), air-core/MMIC spiral (mmWave).

Use below: SRF/3

Q: 15-100

X_L: 2πfL

Understanding RF Inductors

At low frequencies, an inductor is simply a coil of wire with a well-defined inductance value. At RF frequencies, everything changes. The inter-turn capacitance, the skin effect, the proximity effect, and the substrate losses all become significant, transforming the simple inductor into a complex distributed element. Understanding these high-frequency effects is essential for selecting the right inductor and predicting circuit performance.

The most common mistake in RF design is using an inductor above its SRF. A 100 nH inductor with SRF of 800 MHz used at 1 GHz does not provide 100 nH of inductance; it behaves as a capacitor and the circuit will not function as designed. Checking the SRF against the operating frequency is the first step in any inductor selection for RF applications.

Inductor Equations

Inductive reactance:
X_L = 2πfL = ωL (Ω)
10 nH at 2 GHz: X_L = 125.7 Ω

Self-resonant frequency:
SRF = 1/(2π√(LC_para))
Use rule: f_max < SRF/3
10 nH, SRF=6 GHz: use <2 GHz

Quality factor:
Q = X_L/R_series = ωL/R_s
R_s includes DC R, skin effect,
proximity effect, core loss

Matching network loss:
IL ≈ 4.34 × Q_circuit/Q_inductor dB
Q_ckt=5, Q_L=50: IL≈0.43 dB
Q_ckt=5, Q_L=20: IL≈1.09 dB

RF Inductor Technology Comparison

Type	Q @ 1 GHz	SRF Range	Tolerance	Application
Wirewound chip	40-100	1-10 GHz	±5%	Matching, bias tee
Multilayer chip	15-30	1-15 GHz	±5-10%	Decoupling, bias
Thin-film	30-60	2-20 GHz	±2%	Precision matching
MMIC spiral	5-15	10-100+ GHz	±5%	mmWave MMIC
Conical	20-50	DC-18 GHz	±10%	Wideband bias choke

Common Questions

Frequently Asked Questions

What is SRF?

SRF = 1/(2π√(LC_para)). Parasitic capacitance resonates with inductance. Below SRF: inductive. At SRF: max impedance (parallel resonance). Above SRF: capacitive. Use below SRF/3. 10 nH with SRF=6 GHz: use below 2 GHz. 100 nH: SRF~800 MHz, use below 250 MHz. Higher L = lower SRF (more turns = more C_para).

How does Q affect performance?

Q = ωL/R_s. Higher Q = lower loss. Matching network IL ≈ 4.34×Q_circuit/Q_inductor dB. Q_ckt=5, Q_L=50: 0.43 dB loss. Q_L=20: 1.09 dB. Filters: Q determines IL and selectivity. VCO: phase noise ∝ 1/Q². Typical Q@1 GHz: wirewound 40-100, multilayer 15-30, thin-film 30-60.

What types for RF?

Wirewound: highest Q but lower SRF, for matching/filters <3 GHz. Multilayer: moderate Q, small size, for decoupling/bias. Thin-film: tight tolerance (±2%), precision matching. MMIC spiral: 0.1-2 nH for 20-100 GHz on-chip. Conical: wideband bias chokes (DC-18 GHz). Air-core eliminates core losses for highest-frequency use.

Related Terms

← Impedance Matching Insertion Loss →