160nH 500mA Broadband Inductors Robust DC Broadband Conical Inductors 100MHz - 22GHz

Why Bias Tee Design Starts with the Inductor

Ask any microwave systems engineer to name the component that most often limits their bias tee bandwidth, and the answer is nearly always the same: the DC choke inductor. The rest of the circuit—a coupling capacitor, a DC blocking cap, perhaps a resistor for low-frequency termination—rarely presents bandwidth constraints. The choke is the bottleneck.

Here is the fundamental problem: a bias tee inductor must appear as an open circuit to RF from the lowest operating frequency to the highest, while simultaneously conducting DC current to the active device. In narrowband systems, a high-Q resonator-based choke works. In broadband systems spanning multiple octaves, the approach collapses—the same reactive element that provides 1 kΩ of impedance at 500 MHz will inevitably self-resonate somewhere in-band, producing a catastrophic transmission null.

The HALT25008 bypasses this physics barrier. Rather than attempting to suppress parasitic capacitance in a uniform winding—a losing battle above several GHz—this component exploits the distributed nature of a conical structure. Parasitic capacitance exists, but it is spread continuously along the taper rather than concentrated at a discrete resonant frequency. The outcome is a bias tee choke that maintains RF-to-DC isolation exceeding 1.2 kΩ across the full 100 MHz to 22.0 GHz band, with no resonant dip disrupting the S21 transmission characteristic.

Five Hundred Milliamps Through an 80-Micron Wire: The Thermal Engineering Story

Pushing half an amp through a conductor thinner than a human hair would seem to invite thermal failure. Yet the HALT25008 handles this routinely. The explanation involves four interacting design choices:

1. Conductor purity matters at this scale. The 0.08 mm winding wire is drawn from oxygen-free electronic-grade copper (OFHC, 99.99% Cu). At 25°C, the DC resistance measures approximately 1.8 Ω. I²R losses at 500 mA amount to roughly 450 mW—manageable in a 3 mm air-core structure with natural convection.
2. Positive TCR feedback is controlled through thermal margin. Copper resistivity increases at +0.39% per degree Celsius. If the winding temperature rises, DCR increases, and dissipation grows—a positive feedback loop. The HALT25008 avoids this spiral by limiting the rated ΔT to 15°C above ambient. Even at +125°C environment, the junction stays within safe bounds for the polyimide insulation system (rated 200°C continuous). No thermal runaway path exists within the specified envelope.
3. No magnetic core means no hidden heat source. Ferrite chokes lose energy through hysteresis and eddy currents, converting RF power into core heat that compounds the DC I²R loss. The air-core HALT25008 has zero core loss. Every milliwatt of dissipation comes from a predictable, calculable copper resistance alone.
4. Open-air geometry provides passive cooling. Unlike potted or encapsulated inductors that trap heat internally, the exposed conical winding radiates and convects freely. A 0.5 mm air gap around the coil body is all that is needed to sustain the rated 500 mA at elevated ambient.

Environmental Ruggedness: From Cold-Soak to Desert Deployment

RF hardware deployed in the field sees temperature extremes that lab benches never encounter. A bias tee on an airborne radar array might cold-start at -55°C at altitude and, minutes later, reach +85°C from self-heating and solar load once the avionics bay warms up. Commercial-grade inductors fail under these swings—ferrite cores crack from CTE mismatch, solder joints fatigue, and enamel insulation embrittles.

The HALT25008 is hardened against all of these mechanisms:

• No CTE mismatch: There is only one structural material in the winding—copper. There is no ferrite-to-copper interface, no ceramic substrate, no epoxy encapsulant filling the coil. Thermal expansion is uniform and stress-free.
• Polyimide insulation system: Unlike conventional polyester or polyurethane wire enamels that soften above 130°C, the polyimide coating on the HALT25008 winding maintains full dielectric strength from -55°C through +125°C continuously, with short-excursion capability to +200°C.
• Vibration survival: A single dot of low-outgassing RF epoxy affixes the winding body to the substrate, eliminating the microphonic modulation that would otherwise introduce AM/PM noise. The low mass of the air-core structure (< 5 mg) means that even under MIL-STD-202 Method 204 vibration profiles, mechanical resonance occurs far above the excitation spectrum.
• Lead finish integrity: Gold-over-tin plating on the copper leads prevents oxidation during storage and ensures consistent wetting during soldering, even after 12 months in a cleanroom waffle pack.

Measured Performance Across the Bias Tee Operating Envelope

Each HALT25008 is characterized in a representative bias tee topology: apex lead soldered to a 50-Ω microstrip through-line, base lead connected to a DC feed pad with broadband bypass capacitance, and the full two-port network measured on a calibrated VNA with TRL de-embedding referencing the solder joints:

Specification Reference

Choosing Between HALT25008 and HALT20015 for Bias Tee Duty

Both are 160 nH conical inductors from Hoan, but they solve fundamentally different bias tee requirements:

Assembly for Reliable Bias Tee Operation

A conical inductor is not a drop-in replacement for a chip inductor. Assembly technique directly determines the realized bandwidth:

1. Apex-end connection: The narrow tip of the cone is the high-frequency port. Solder this lead directly to the 50-Ω microstrip trace with the coil standing perpendicular to the board surface. The perpendicular orientation minimizes parasitic magnetic coupling to adjacent traces and ground-plane eddy currents.
2. Lead trimming discipline: The apex lead between the solder fillet and the first winding turn must not exceed 0.5 mm. This is not an arbitrary guideline—every 0.5 mm of excess lead contributes approximately 0.3 nH of series inductance, which alone can pull the apparent SRF below 20 GHz. Use precision cutters under a stereo microscope for consistent results.
3. Base-end to DC: The wide end lead connects to the DC bias input pad. A broadband bypass capacitor network (typically 100 pF || 10 nF || 1 µF) should be placed within 1 mm of this pad to prevent RF from traveling up the DC supply line.
4. Epoxy application: After both solder joints are inspected, apply a single micro-dot (approximately 0.3 mm diameter) of Epotek H70E or H65 to the side of the conical winding where it contacts the substrate. The epoxy is not structural for DC current—its role is to arrest micron-scale mechanical vibration of the winding turns that generates phase noise through inductance modulation.
5. Soldering thermal profile: 280–320°C tip temperature. Keep dwell time under 3 seconds per joint. If reworking, allow the coil to cool completely before re-applying heat. Successive heating cycles can anneal the 0.08 mm copper, softening it and potentially altering the winding geometry.

Field-Proven Applications

• DC bias injection networks in 100G/400G/800G optical transceiver modules (TOSA laser driver bias, ROSA TIA supply)
• GaN-on-SiC HEMT power amplifier drain bias feed in X-band through K-band radar transmitters
• PIN diode switch bias routing in wideband electronic warfare receiver front-ends
• Active electronically scanned array (AESA) antenna element bias distribution
• Millimeter-wave VNA frequency extender bias tees for on-wafer device characterization
• Cryogenic low-noise amplifier bias injection (qualified for operation from -55°C cold start)

Frequently Asked Questions by Bias Tee Designers

Q: If I simulate my bias tee in ADS using the .s2p file and it looks perfect, should I still prototype before committing to production?
A: The .s2p data captures the HALT25008 behavior on a characterized microstrip fixture. It does not model your specific PCB stack-up, pad geometry, ground-plane spacing, or adjacent trace coupling. Below 18 GHz, the correlation between .s2p-based simulation and measured hardware is typically excellent (< 1 dB S21 deviation). Above 20 GHz, your PCB layout parasitics dominate, and we strongly recommend building a short test coupon with your exact dielectric material and stack-up. Hoan can review your layout and provide application-specific feedback.

Q: Our system requires operation at 85°C ambient with the HALT25008 passing 450 mA. Is this within the safe zone?
A: Yes. At 450 mA, I²R dissipation is approximately 365 mW. With ΔT of roughly 12°C at this current level, the winding temperature would be approximately 97°C at 85°C ambient—well within the polyimide insulation rating and within the rated ΔT ≤ 15°C margin. No derating is required for this operating point.

Q: Can the HALT25008 survive lead-free reflow if we want to use it in a hybrid SMT process?
A: The HALT25008 is a flying-lead component, not a surface-mount device, and is not rated for full reflow oven exposure. The 0.08 mm polyimide-insulated wire can survive brief excursions to reflow temperatures, but sustained exposure above 320°C risks insulation degradation. The recommended assembly method is post-reflow hand or automated micro-soldering of the flying leads. For fully automated SMT lines, contact Hoan about custom lead-frame packaging options.

Q: Our bias tee must pass 10W of RF through the transmission path. Does the HALT25008 need power derating?
A: The HALT25008 is on the DC bias side of the bias tee, not in the main RF through-path. The RF power traveling through the 50-Ω microstrip does not pass through the inductor. The choke only sees the RF voltage present at the tee junction, which induces a small AC current through the inductor’s reactive impedance. At 1.2 kΩ isolation, the RF voltage divider action keeps choke dissipation negligible. The 500 mA rating refers to DC current only.

Q: What is the lead time for custom variants if we need a different inductance value?
A: Hoan maintains a quick-turn conical inductor prototyping capability. Custom inductance values (typically 50 nH to 500 nH in the 0.08 mm wire series) can be sampled within 3–4 weeks. Volume production lead times are 8–12 weeks depending on quantity. Contact Hoan with your target inductance, frequency band, and DC current requirement for a feasibility assessment and quotation.