Measuring UNUNs for Dual-Band EFHW and High-Impedance Loads
Measuring High-Ratio UNUNs for Dual-Band EFHW
When you measure an unloaded EFHW autotransformer (49:1, 64:1, 68:1, 70:1), you are really seeing the magnetizing inductance in parallel with leakage and stray capacitance. Below the winding’s self-resonant frequency (SRF) it looks inductive; above SRF it flips capacitive—common on mix-77 around ~1–2 MHz.
Unloaded sweeps only prove the magnetizing branch is high enough at the lowest band. They do not validate the impedance transformation—use a resistive load for that.
Dual-Band Design Philosophy
Rather than forcing one transformer to cover all of HF, we optimize for specific pairs:
- 68:1 → 160/80 m (optimized 160 m)
- 70:1 → 80/40 m (optimized 80 m)
- 49:1 → 40/20 m (optimized 40 m, usable 20 m)
This keeps loss low, heating minimal at QRO, and behavior predictable.
Measurement Expectations
Unloaded (primary only)
- Inductive below SRF; capacitive above SRF (expected on mix-77 into 80/160 m).
- The lowest band should see XL ≥ 10–20 × 50 Ω.
Loaded (secondary with test resistor)
Attach a non-inductive resistor equal to the intended high-Z load (~3.3–3.6 kΩ for EFHW). The input should be close to 50 Ω on both bands; a little reactance is fine.
Primary Inductance Targets & Bands
Targets below reflect real builds that run cool at power (your recent measurements: 170 µH 160-opt; 84 µH 80-opt; ~40 µH 40/20 on mix-43).
| Ratio | Best Dual-Band Pair | Target Lm | XL @ 1.8 MHz | XL @ 3.5 MHz | XL @ 7 MHz | XL @ 14 MHz |
|---|---|---|---|---|---|---|
| 49:1 | 40 m / 20 m | 30–45 µH (≈40 µH) | ≈340–510 Ω | ≈660–990 Ω | ≈1.32–1.98 kΩ | ≈2.64–3.96 kΩ |
| 68:1 | 160 m / 80 m | 140–200 µH (≈170 µH) | ≈1.58–2.26 kΩ | ≈3.08–4.40 kΩ | ≈6.16–8.80 kΩ | ≈12.3–17.6 kΩ |
| 70:1 | 80 m / 40 m | 70–100 µH (≈84 µH) | ≈0.79–1.13 kΩ | ≈1.54–2.20 kΩ | ≈3.08–4.40 kΩ | ≈6.16–8.80 kΩ |
XL = 2πfL. Choose Lm so the magnetizing branch is “invisible” on the lowest band, without adding excess stray C.
Ferrite Mix Guidance
- Mix-77 (stacked): excellent for 160/80 when band-optimized; SRF often ~1–2 MHz. Expect open-circuit capacitive behavior near/above 80 m, but a proper 3.3–3.6 kΩ load damps it. Runs cool at QRO if cross-section is adequate.
- Mix-43: lower loss into 40/20 and above; easier to keep SRF high. Use for 40/20 (and 15/10 variants with slightly lower Lm).
Workflow (quick)
- Unloaded: verify inductive region on the intended band(s); SRF location noted.
- Loaded: add 3.3–3.6 kΩ on the tap; confirm ~50 Ω input near each target band.
- Choking: place a 0.05 λ choke at the feedpoint and another before the shack entry.
- Thermal: brief QRO key-down; enclosure should stay cool/warm-to-touch on the optimized band.
Mini-FAQ
- Why no capacitors? — Proper winding (trifilar start → bifilar bulk) provides the needed L and coupling; shunt caps are unnecessary in these dual-band builds.
- Why 68:1 for 160/80? — Higher turns and cross-section raise Lm (~170 µH target), keeping the magnetizing branch out of the way at 1.8–3.5 MHz.
- Can one box do all bands? — Not efficiently. Dual-band, band-optimized units are predictable, cool at power, and easier to choke correctly.
Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.
Questions or experiences to share? Feel free to contact RF.Guru.