Passive ARDF Probes: Preserve the Null Before Adding Gain
Passive ARDF Probes: Preserve the Null Before Adding Gain
In amateur radio direction finding, more output is not automatically more information. A passive, balanced probe is often the cleanest starting point because it protects symmetry, receiver headroom and a repeatable bearing. Gain belongs only where the complete noise, balance and overload budget shows that it helps.
The practical ARDF question is not “how many decibels can I add at the probe?” It is “does the complete antenna and receiver still give me the same trustworthy direction when I rotate, move, attenuate and approach the transmitter?” My default is to earn the bearing with passive symmetry first. If the system then proves receiver-noise limited, gain can be added deliberately instead of by reflex.
Joeri's design order: protect the directional pattern, control common mode, remove unwanted spectrum, preserve overload margin and only then add the minimum gain the measured system needs.
The Null Carries the Directional Information
An electrically small loop has a bidirectional figure-eight response in its plane of rotation. In an ideal far field, a simple angular model is:
Vloop(φ) = A sin φThe two zeros are 180 degrees apart, so the loop gives a line of bearing but not by itself the sense of that line. A non-directional sense response can be combined with the loop response to produce a single-null cardioid:
V(φ) = A sin φ + B ejψThe useful single null appears only when amplitude and phase have the required relationship at the operating frequency. If B/A or ψ changes, the null depth and direction change. A passive combiner, active circuit or digital processor can all perform this job; none is excused from amplitude, phase and common-mode error.
A cardioid is not created by the words “E plus H”. The directional and sense channels must be combined with controlled magnitude and phase. Changing distance, polarisation, ground coupling or nearby conductors can change the two channel responses and fill or move the null.
Balance Is Part of the Antenna Pattern
A differential loop can reject fields that couple equally to both terminals only while the complete path preserves that symmetry. An unequal load, one-sided capacitance, hand coupling, an asymmetric cable, enclosure current or an unbalanced amplifier input can convert common-mode pickup into differential output.
That extra response does not rotate with the intended loop pattern in the same way. It fills the null, makes the bearing depend on how the receiver is held and can make the coax or operator part of the antenna. A high signal level from such a system is not evidence of a better direction finder.
Balance must therefore be followed through the probe terminals, transformer or amplifier, interconnect, receiver enclosure and operator coupling. Differential gain and low noise are useful specifications, but so are common-mode rejection, input symmetry, parasitic capacitance and cable-current control.
An LNA Is Not the Villain—Unexamined Gain Is
A probe-mounted LNA can be technically sound. A genuinely differential, sufficiently linear input can buffer a high-impedance or low-level sensor, overcome cable loss and improve system noise figure when the receiver is the limiting noise source. Active aircraft direction finders have used loop and sense amplification successfully for decades.
But adding gain ahead of selectivity also raises every signal admitted by the probe. A single-ended input can disturb balance. An input that is poorly matched to the sensor can change amplitude and phase. Strong broadcast, amateur and local signals can drive the amplifier or following receiver into compression or intermodulation. The cable then carries a larger mixture of wanted signal, unwanted signal and distortion.
Analog Devices' published receiver example demonstrates the trade: an external LNA improved sensitivity while substantially degrading third-order intercept in that particular system. The numbers are not transferable to ARDF, but the engineering lesson is. Noise figure, gain, balance, filtering and linearity must be evaluated together.
Why Passive Is Often the Better Starting Point on 3.5 MHz
IARU ARDF includes competition in the 3.5 MHz amateur band. At these frequencies, atmospheric and man-made radio noise can dominate the thermal noise of a competent receiver, depending on site and bandwidth. ITU-R P.372 treats those external noise sources as central inputs to radio-system design below 100 MHz.
When external noise already arrives well above the receiver noise floor, more front-end gain raises wanted signal and noise together without improving input SNR. A passive probe can then offer several practical advantages:
- no active device at the antenna to compress or generate intermodulation;
- no local power lead or bias network adding another coupling path;
- easier preservation of a symmetric differential input;
- predictable attenuation close to the transmitter; and
- a cleaner baseline for deciding whether gain is actually required.
This is a design default, not a universal superiority claim. A very small lossy sensor, long lossy interconnect or noisy receiver may justify a balanced active stage. The proof is system SNR and bearing repeatability under the real signal environment, not the presence or absence of a transistor.
Filtering Usually Buys More Headroom Than Broadband Gain
A preselector reduces energy that cannot help the bearing before it reaches nonlinear stages. That can improve usable dynamic range even when it introduces some insertion loss. In a strong-signal environment, removing an out-of-band blocker can matter more than lowering noise figure by a small amount.
The placement depends on the system. A filter before an amplifier protects the amplifier but its loss contributes directly to noise figure. A filter after an amplifier benefits from preceding gain but cannot prevent that amplifier from overloading. A switched bypass, selectable attenuation and more than one gain state often give a handheld receiver a wider useful operating range than one fixed high-gain path.
Near the transmitter, attenuation is not a concession. It is part of direction finding. The objective is to keep the receiver within its linear indication range while preserving enough directional contrast to find the null.
Resonance Trades Bandwidth for Selectivity and Sensitivity
Resonating a loop or sense channel is not fundamentally wrong. It can raise voltage, improve selectivity and reduce the noise contribution of a following stage. It also makes amplitude and phase change more rapidly around resonance. Component tolerance, hand capacitance, temperature, nearby objects and frequency error can then move the combined null.
A broadband or deliberately damped probe gives up some resonant voltage in exchange for a response that is easier to keep stable while moving through terrain and approaching the transmitter. The appropriate loaded Q follows the event bandwidth, required settling, available tuning, environmental variation and receiver sensitivity. “Highest Q” and “lowest Q” are both incomplete design targets.
Use the Complete ARDF Receiver as the Test Object
| Question | Evidence to collect | Warning sign |
|---|---|---|
| Does the probe preserve balance? | Differential/common-mode response, cable current and repeatability with receiver orientation | The null moves when the cable, enclosure or operator position changes |
| Is gain needed? | Wanted signal, noise and receiver indication with restored-baseline gain/bypass comparison | Signal and noise rise together or intermodulation appears |
| Is filtering helping? | In-band SNR and blocker levels before/after the filter at equal receiver settings | A quieter output is mistaken for better SNR |
| Is the null trustworthy? | Full rotation, repeated bearings, reverse approach and known-source checks | Only one convenient heading or distance was tested |
| Is the close-in range controlled? | Step attenuation and linear indication over decreasing distance | The receiver remains pinned or the minimum disappears near the source |
| Is the sense channel correct? | Amplitude/phase sweep and single-null direction across the event band | The cardioid reverses, fills or turns with frequency |
Compare one change at a time and restore the baseline. Record receiver bandwidth, gain, attenuation, AGC, frequency, cable route and operator geometry. A handheld ARDF system is an antenna, receiver, enclosure and human body operating as one electromagnetic installation.
My Practical Bottom Line
The lesson from field direction finding is not that every LNA is bad. It is that gain cannot repair a damaged null. Start with a passive, balanced probe because it makes the current paths and failure modes easier to control. Keep the sense combination, filter, attenuation and receiver inside one declared design.
If the passive system already places external noise comfortably above the receiver floor, spend the available margin on filtering and close-in attenuation. If it is receiver-noise limited, add a balanced, linear and bypassable active stage whose amplitude, phase, common-mode behaviour and overload margin have been measured. The winner is the configuration that gives the most repeatable bearing across the course—not the highest S-meter reading on the starting line.
Primary and Authoritative References
- IARU Region 1 — Rules for Championships in Amateur Radio Direction Finding, Part A
- ITU-R P.372-17 — Radio Noise
- US Patent 2,308,521 — Automatic Radio Direction Indicator
- US Patent 3,967,280 — Integrated Loop and Sense Antenna Direction Finder
- Analog Devices — Improving Receiver Sensitivity with an External LNA
- Analog Devices — Use Selectivity to Improve Receiver Intercept Point
- Keysight — Differential, Common and Mixed-Mode Measurements
Mini-FAQ
- Is a passive ARDF probe always better than an active one? No. It is often the cleanest starting point for balance and headroom. A balanced, linear active stage can help when the receiver is genuinely noise limited or interconnect loss matters.
- Does an LNA improve bearing accuracy? Not by itself. Gain can improve sensitivity, but bearing accuracy depends on pattern, amplitude and phase balance, common-mode control, receiver linearity and calibration.
- Why can a cable or hand fill the loop null? Asymmetric coupling or common-mode current creates an additional receiving path that does not share the intended differential loop pattern.
- Does combining a loop and sense antenna automatically make a cardioid? No. The directional and sense responses need the correct relative amplitude and phase across the operating range.
- Should an ARDF probe be resonant? Resonance can add selectivity and voltage, but it also increases sensitivity to detuning. Choose loaded Q from the required bandwidth, stability and receiver-noise budget.
- What should come before more gain? Verify balance and the null, remove unwanted spectrum, preserve linear headroom and provide enough attenuation for close-in bearings.