Your Active Antenna Isn’t the Upgrade ... This Is

Most of us know the sequence. Install a new active antenna, see more level on the display and call it an upgrade. Then the noise rises, strong stations generate signals that are not on the air, or touching a cable changes the band. The antenna may be working perfectly. The receive system is telling you where the real limitation lives.

A receiving station is successful when it improves the wanted signal relative to noise and interference at a declared detector, bandwidth and operating condition. More gain and more bars do not prove that. My practical rule is to follow the current and level paths before replacing the antenna.

The Upgrade Is the Complete Receive Chain

The chain begins in the field around the antenna and ends at the detector or decoder. Between those points sit the antenna pattern and polarization, antenna factor or effective length, active element, feed line, bias network, protection, filters, receiver front end, gain controls and digital conversion. The installation also contains unintended paths through the coax exterior, power and control wiring, equipment enclosure and nearby conductors.

Each part can change a different quantity. Antenna placement changes the mixture of wanted field and local interference. Gain changes level and the contribution of later receiver noise. Filtering changes the spectrum presented to a nonlinear stage. Attenuation trades sensitivity for headroom. A choke changes one common-mode loop. None of those actions is a universal substitute for the others.

Define the Wanted Mode Before Chasing Noise

In coax, the wanted transmission-line mode carries current on the centre conductor and the inner surface of the shield, with equal-and-opposite terminal current at an isolated two-terminal port. The shield exterior can carry an additional current relative to the surrounding installation. That exterior current is commonly called common-mode current, although the exact modal description depends on the complete multi-conductor structure.

A balanced two-wire port likewise has equal-and-opposite desired terminal currents. Physical symmetry can help maintain balance, but geometry alone does not prove it. Capacitance to the mast, cable routing, power leads and nearby metal can create a third path. Once that happens, the feed line and station can receive local fields in addition to transporting the antenna’s wanted differential signal.

A change when you touch a connector or move the cable is therefore a useful clue, not a diagnosis. It may reveal exterior current, a changing capacitance, a poor connection, a level change or several effects together. Reproduce the observation at safe receive-only levels, then measure the relevant current, impedance and SNR before choosing the fix.

Measure SNR, Not Display Height

Record the wanted signal and noise with the same receiver mode, bandwidth or resolution bandwidth, detector, AGC state, preamplifier, attenuation, reference level and averaging. For digital modes, use a repeatable decoder metric as well as spectrum measurements. A wider bandwidth contains more noise power, so two noise-floor numbers are not comparable unless their bandwidths and detectors are known.

Make the reference plane explicit. “At the antenna,” “at the shack entry” and “at the receiver input” describe different systems once cable loss, filters and gain are present. For an active antenna, record whether quoted gain, noise figure, compression or intermodulation performance belongs to the element electronics alone or to the installed chain.

Recommendation ITU-R P.372 distinguishes atmospheric, galactic and man-made external noise. A receiver adds its own noise. Strong wanted or unwanted signals can also create blocking, intermodulation or converter clipping. Those mechanisms can look similar on a casual waterfall but demand different remedies.

Gain Helps Only Within Its Boundary

Friis’ cascade relation explains why gain early in a receive chain can make the noise contribution of later stages less important. It does not say that arbitrary gain improves SNR. Noise added ahead of that gain, noise generated by the active stage and interference already captured by the antenna are amplified too.

The active stage must also remain linear in the complete spectrum at its input—not merely in the narrow channel you want to hear. A wideband active antenna can overload before a shack filter ever sees the signal. A filter after that overloaded stage cannot undo products already created there.

Use enough active gain to overcome subsequent cable and receiver noise for the declared band and external-noise environment, with margin below compression and intermodulation limits. If a known input attenuation improves the spectrum or decoding nonlinearly rather than lowering wanted signal and external noise by the same amount, excess level or receiver overload is a strong suspect.

Placement and Antenna Architecture Are Conditional

Moving an antenna can change wanted signal, radiated interference, near-field coupling to wiring and current on attached conductors. Distance from a particular noise source may help; greater height may help or hurt; rotation may place a pattern null on an interferer while also changing the wanted signal. The result depends on frequency, polarization, geometry and the actual field distribution.

Small loops, electric-field probes, arrays and passive wires have different patterns, antenna factors, bandwidths, balance and linearity limits. A shielded loop is not inherently immune to electric-field noise, and an electric-field probe is not inherently noisy. Construction, feed symmetry, common-mode control, placement and the local source determine the installed result.

Test candidate positions with the same receiver settings, over representative frequencies and times. When possible, use a stable wanted signal or calibrated injected signal and repeat A/B/A changes so changing propagation is not mistaken for an installation improvement.

Find the Coupling Path

Start with the receiver input terminated in its specified impedance. That exposes receiver noise and internally generated spurs. Then connect the installed feed line with its far end correctly terminated. A new line or spur suggests coupling through the cable, shield exterior, power/control wiring or receiver environment. Finally connect the antenna and compare again.

A clamp-on current probe can map relative current on the coax exterior, bias lead and control cables. Use the same probe orientation and position, and scan more than one place because a standing-wave minimum at one point does not prove that the conductor is quiet. A portable receiver can help locate emissions, but final comparisons need stable settings, documented bandwidth and repeated positions; Recommendation ITU-R SM.2093 provides a disciplined framework for characterising indoor radio noise.

When practical, temporarily power the receive system from a suitable battery and disconnect nonessential data/control links without defeating protective bonding. A change identifies a path worth investigating; it does not by itself prove that the removed supply or cable is defective.

Choking, Isolation and Bonding Are Different Tools

A common-mode choke inserts complex impedance into a selected external-current loop. Its result depends on frequency, the installed loop impedance, cable route, ferrite material, winding capacitance and location. One choke may be sufficient, several may be useful, or none may address the dominant path. Measure exterior current and receiver SNR before and after each placement instead of prescribing chokes at three fixed points.

An RF isolation transformer separates conductive port references over a finite passband, but parasitic capacitance still provides a common-mode path. It also has differential insertion loss, amplitude/phase response, impedance limits and possible core nonlinearity. Verify both wanted-mode transmission and common-mode conversion or isolation across the frequencies and levels of interest. “Galvanic isolation” is not the same specification as infinite RF isolation.

Equipment bonding, protective earthing and lightning protection are safety functions governed by the installation and local rules. Do not disconnect a protective conductor or improvise a mains isolation arrangement to hunt receive noise. An RF current-path problem must be solved without weakening fault or lightning protection.

Filter Before the Stage That Overloads

A preselector, band-pass, high-pass, low-pass, notch or broadcast-band-reject filter is useful when its measured passband preserves the wanted signal and its stopband reduces a known interferer before the first stage that cannot handle it. The filter’s insertion loss, impedance, return loss, power and linearity belong in the calculation.

Placement is decisive. A shack filter can protect the receiver from out-of-band energy delivered by the feed line. It cannot protect an active antenna amplifier that has already overloaded at the mast. In that case, the options include reducing field level, using a more selective or more linear front end, adding filtering ahead of the vulnerable stage where technically possible, or changing location or architecture.

Receiver blocking, two-tone intermodulation and ADC clipping are not interchangeable. Keep the test tones, levels, spacing, reference bandwidth and gain settings with every result. Receiver standards such as ETSI EN 300 113 define these behaviours through separate tests for good reason.

Attenuation Buys Headroom at a Noise Cost

Attenuation before a receiver reduces both wanted signal and external noise while increasing the relative importance of the receiver’s own noise. That is a poor trade when the receiver is sensitivity-limited. It can be an excellent trade when excess gain or strong signals are causing compression, intermodulation or clipping and the remaining system noise is still safely above the receiver contribution.

Use a calibrated step attenuator and hold every other setting fixed. If 3 dB of input attenuation produces about 3 dB less wanted signal and 3 dB less external noise, SNR is broadly unchanged and the chain was probably not limited by that level. If false signals disappear, the displayed noise falls disproportionately or decoding improves, investigate overload. If wanted SNR falls immediately, restore the level and look elsewhere.

Power and Bias-T Paths Need Their Own Test

A switch-mode supply is not automatically noisy and a linear supply is not automatically quiet. Either can couple differential noise through its output or common-mode current through capacitance, protective earth, data links and enclosure. Characterise the supply in the installed frequency range and load state.

A bias-T is a three-port network, not an invisible DC wire. Its RF insertion loss, DC-current rating, isolation and the impedance of its inductive and capacitive branches vary with frequency. A DC filter may help only when it attenuates the measured noise in the real source/load network without destabilising the active antenna or exceeding component ratings. Compare the spectrum and SNR with a known-clean temporary supply before redesigning the filter.

Protection Is Not a Noise Cure

Limiters, surge protectors and static-discharge paths protect against specified events; they can also add capacitance, insertion loss or nonlinearity. Select them from declared surge, voltage, current, frequency and coordination requirements, then verify receive-chain performance at normal levels. A protector does not make a connected outdoor antenna safe from lightning.

A Measurement Sequence That Finds the Limiter

1. Freeze the receiver state. Record frequency, mode, bandwidth, detector, AGC, preamp, attenuation, gain, reference level and averaging.
2. Terminate the receiver input. Record internal noise and spurs with a suitable termination and safe procedure.
3. Test the installed line. Connect the feed line with its remote end correctly terminated; scan exterior current and note any new spectrum.
4. Connect the antenna. Record wanted signal, noise and strong out-of-band levels at the same reference plane and settings.
5. Apply known attenuation. Separate a receiver-headroom problem from an external-noise- or sensitivity-limited system.
6. Change one path. Move a cable, compare a choke position, substitute a supply or insert a suitable filter; return to the baseline for an A/B/A test.
7. Test position and orientation. Repeat at representative frequencies and times so a local spatial effect is not mistaken for a universal antenna ranking.
8. Keep the useful change. Document the measured SNR, current or overload improvement and its operating boundary.

This order is diagnostic, not mandatory hardware. It deliberately avoids buying a line of cures before the limiting mechanism is known.

When an Antenna Swap Is the Upgrade

Change the antenna when measurements show that its pattern, polarization, effective height or antenna factor, bandwidth, linearity, environmental tolerance or available placement cannot meet the objective. That is a technical decision, not a defeat of system thinking.

If common-mode current is controlled, receiver settings are repeatable, overload is excluded and the candidate antenna produces better wanted-signal SNR across the required conditions, then the antenna really is the upgrade. Until then, improve the chain that lets it be heard.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.