Choose an RF inductor at the circuit’s operating frequency. Compare Q and effective inductance there, check self-resonance and the board model, and verify the exact package. Matching a low-frequency inductance value alone cannot qualify a replacement.
1. Define the RF circuit first
Record operating band, bandwidth, impedance, signal level and whether the part provides matching, tuning, filtering or RF isolation. Each use has different loss and resonance constraints. A higher current rating alone says little about RF suitability.
2. Compare Q at the same frequency
Q relates reactive to resistive impedance. Its value depends on frequency, instrument and fixture. Compare Q curves over the operating band rather than ranking parts by an isolated catalogue maximum. Low DC resistance does not describe all RF losses. [1]
3. Treat SRF as a circuit boundary
Parasitic capacitance resonates with the coil. Near the first self-resonance, impedance and phase change sharply; above it, the part may behave capacitively. A simple approximation is f₀ = 1 / (2π√(LC)), but a model or measurement is needed for real layouts. [2]
4. Include the package and PCB
The ground plane, substrate and mounting affect capacitance and measured SRF. Compare manufacturer fixtures and use the relevant board model. A smaller package is not automatically the better match. [2]
5. Validate the complete network
Use current manufacturer S-parameters or a suitable model, then measure the populated circuit over frequency, temperature and component tolerance. Record insertion loss, matching or tuning response and measurement setup. Do not approve a part from a single nominal L value.
Parameter review table
| Review item | Required evidence | Decision |
|---|---|---|
| Q and effective L | Curves at operating frequencies | Compare the entire band |
| SRF and parasitics | Original test fixture and model | Check phase and resonance behavior |
| DCR and current | Exact-part rating conditions | Check heating separately from RF loss |
| Package and PCB | Land pattern, ground plane and substrate | Validate the assembled network |
Put the review into practice
Illustrative review: two parts both marked 10 nH may have different Q and effective L in your band. Plot their original curves on the same frequency axis and compare circuit simulations before selecting a prototype. This example supplies no invented vendor Q or SRF figures.
Frequently asked questions
Is higher Q always better?
No. The desired network response also depends on bandwidth, loss, matching and tolerance. Use Q at the actual frequency together with the full circuit response.
What happens near SRF?
Impedance and phase change with parasitic resonance. A part selected for inductive behavior needs frequency-dependent verification; do not assume its low-frequency L remains valid.
Can I compare Q values measured at different frequencies?
Not directly. Obtain common-frequency curves and compatible test conditions, or measure both parts with the same setup.
Can Q at 250 MHz represent performance at 900 MHz?
No. Q and effective inductance vary with frequency. Use a curve or model at 900 MHz with compatible measurement conditions; neither copying the 250 MHz value nor linear extrapolation establishes it.
Is SRF greater than three times the operating frequency always safe?
No. In an ideal parallel-LC model, at f = SRF/3 the apparent inductance is 1/(1 − 1/9) = 1.125 times the low-frequency value. This hypothetical 12.5% shift is not a universal real-part tolerance; validate the actual network and PCB.
Can two 10 nH inductors be directly interchanged?
Matching nominal L is insufficient. Compare frequency-dependent Q/L, SRF, tolerance, current conditions, pad geometry and the assembled circuit.
Exact-part examples and evidence
0402HP-10NXJR
The catalogue record is a 10 nH 0402HP part. Open Coilcraft’s 0402HP-10N family page for Q/L curves, test notes and ordering-code definitions. Match the full suffix to the current drawing before ordering; family-level information does not establish the lifecycle or availability of every historical suffix. The current 0402HP-10N family table at 25°C lists typical Q 62 at 900 MHz and 90 at 1.7 GHz, typical SRF 4.7 GHz and maximum DCR 0.11 Ω. These family-source values are not a new characterization of every historical ordering suffix.
Open catalogue recordOpen cited documentation0402HP-12NXJR
Checked same-series catalogue record: 12 nH. The current original 0402HP-12N family table at 25°C lists typical Q 66 at 900 MHz and 100 at 1.7 GHz, typical SRF 4.4 GHz and maximum DCR 0.10 Ω. Original pages include L/Q curves, SPICE/model links and measurement fixtures. This 12 nH reference is not a 10 nH substitute; verify the exact ordering suffix.
Open catalogue recordOpen cited documentationOriginal technical references
- Coilcraft — Testing Inductors at Application Frequencies
- Coilcraft — Measuring Self Resonant Frequency
- Coilcraft — 0402HP-10N original product and characteristic curves
- Coilcraft — 0402HP-12N original specifications and curves
Check the latest documentation for the full ordering code and validate the intended circuit before ordering or substituting a part.
