Where to Measure a Multiband Antenna: Feedpoint or Shack?
Where to Measure a Multiband Antenna: Feedpoint or Shack?
Both locations can give valid answers. The right connection point is the reference plane that matches the question you need to answer.
“Should I put the analyzer at the antenna or connect it in the shack?” has no one-location answer. A feedpoint measurement describes the load at the antenna terminals. A shack measurement describes the complete line-and-antenna load presented to the tuner or transmitter. Start by naming the desired plane; then arrange the calibration, fixture and installation so the reading belongs to that plane.
The practical rule: measure at the feedpoint when you need feedpoint impedance for antenna adjustment, modelling or component stress. Measure at the shack when you need to know what the station equipment sees. When both results matter, take both and relate them with a characterized feedline rather than treating either as the universal truth.
Every Reading Belongs to a Reference Plane
An antenna analyzer reports complex impedance or reflection at its calibrated measurement plane. Everything beyond that plane is part of the device under test. Everything between the instrument and that plane must either be included intentionally, calibrated out at the remote end, or removed mathematically using a valid model.
| Connection and plane | What the reading includes | Question it can answer |
|---|---|---|
| At the antenna terminals | Radiator, return structure and any components intentionally left on the antenna side | What load does the antenna present here, and how should its geometry or matching component be adjusted? |
| Coax side of a balun, transformer, choke or matching unit | The antenna plus every device on the antenna side of the named plane | What does this assembled antenna subsystem present to its feedline? |
| At a remote tuner's input | Antenna, output-side line or wire and the tuner's current relay state if it remains in circuit | What raw or tuned load is present at the tuner's input under this explicit state? |
| At the shack end of the feedline | Antenna, matching devices, feedline transformation, attenuation, connectors and any unintended mode conversion | What impedance reaches the station-side tuner or transmitter? |
Record the plane with the sweep. “SWR 1.7” is incomplete; “SWR 1.7 at the shack end of 24 m of this coax, tuner bypassed” is an interpretable result. Record resistance and reactance as well as SWR, because different complex impedances can produce the same SWR.
What the Feedline Changes
A uniform transmission line transforms the load impedance according to its characteristic impedance, propagation constant and electrical length:
Zin = Z0 [ZL + Z0 tanh(γl)] / [Z0 + ZL tanh(γl)]
γ = α + jβ
Here, ZL is the load at the antenna end, Z0 is the line's characteristic impedance, l is physical length, α represents attenuation and β represents phase change per unit length. Electrical length therefore depends on frequency and velocity factor. Unless the load equals Z0, resistance and reactance measured in the shack can differ greatly from their feedpoint values even when the line is behaving perfectly.
Loss attenuates the reflected wave on its round trip. A long or lossy line can therefore display a lower shack-end SWR than exists at the antenna end while also dissipating more power. That does not make the shack reading false: it is the real load presented at that plane. It simply cannot be interpreted as the feedpoint impedance, antenna efficiency or radiator quality.
Keep two effects separate. Differential-mode transformation inside a characterized line is predictable. Exterior-shield or other common-mode current changes the installed current paths, so moving the cable or analyzer can change the system itself. A line calculation can correct the first effect; it cannot remove an unknown radiating return path.
Measure at the Feedpoint When the Antenna Is the Question
A feedpoint plane is appropriate when trimming a dipole, comparing a measured load with an antenna model, characterizing a matching network's load, or estimating terminal voltage and current. Calibrate at the connector or terminals whenever practical. If a short adapter or jumper remains beyond the calibration plane, include its electrical length, loss and mismatch in the result.
Preserve the antenna's normal installed geometry. Do not lower the radiator, coil the feedline differently or replace the intended return structure merely to reach it. Those changes can matter more than the instrument error. On an unbalanced or asymmetric antenna, the analyzer case, operator, USB lead and temporary jumper can provide capacitance or another return path. That is a fixture and current-path problem—not a universal rule that the feedpoint lies too deep in a single fixed “near-field distance” to be measured.
A short repeatability test is useful: hold the intended antenna geometry fixed, then make small, documented changes to analyzer position, jumper routing or operator proximity. A shifting trace shows that the setup is sensitive to those changes. It does not by itself identify the coupling mechanism or prove that a particular choke location is correct.
Measure in the Shack When Station Equipment Is the Question
The shack plane is the useful place to check whether the installed feed system stays within a transmitter's foldback range, a tuner's matching range or a station-side component's impedance boundary. The result should include the feedline because that is what the equipment actually sees.
For a multiband antenna, save R, X, |Z|, reflection or SWR across every operating band. A single minimum-SWR marker can hide severe impedance at another frequency. Compare the results with the exact tuner's frequency, impedance, voltage, current, power and duty-cycle limits; those limits are manufacturer- and configuration-specific.
For fault finding, a shack sweep is also valuable because it tests the accessible installed system. A new ripple, discontinuity or band shift can point to a changed connector, cable, matching unit or antenna. Distance-to-fault modes have finite bandwidth and spatial resolution, so confirm suspected locations with a physical inspection or another measurement.
Calibration, Port Extension and De-Embedding Have Limits
Open-short-load calibration corrects systematic errors to the plane where the standards were connected. Calibrating at the analyzer connector and then adding a cable does not put the reference plane at the far end automatically. If the instrument permits, perform the calibration at the far end of the measurement cable using standards suitable for that connector and frequency range.
Port extension usually compensates electrical delay. Some instruments can also estimate line loss. Keysight's current FieldFox documentation warns that a basic port extension does not correct the added line's mismatch, and its fixture guidance uses characterized network data when fuller de-embedding is required. The exact capability depends on the analyzer and software, so “cable compensation” is not a sufficient method description.
- Port extension can rotate phase to a new plane for an adequately uniform line; it cannot generally repair an unknown adapter, transformer, severe mismatch or radiating fixture.
- De-embedding can remove a characterized linear fixture over the characterized frequency range. The fixture model and its orientation, reference impedance, ports and calibration planes must match the measurement.
- Renormalization changes the reference impedance used to express measured S-parameters; it does not move the physical calibration plane.
- Unknown common mode cannot be de-embedded as though it were a stable two-port line. First control or characterize the installed current path.
Common Mode Can Move the Boundary
In the intended coaxial mode, centre-conductor current returns on the inner surface of the shield. Current on the shield's exterior follows a different circuit through the antenna structure, mast, ground, equipment, other cables and displacement capacitance. That exterior current can change tuning, pattern, received noise, equipment voltage and the analyzer reading.
A choke is not automatically a clean reference plane. Its impedance is complex and frequency-dependent, and its installed effect depends on the common-mode driving and return-path impedances. Winding capacitance and the surrounding conductors can create resonances. Use measured R+jX data for the finished choke and, where common mode matters, measure exterior current along the line before and after changes.
Choose a choke location to establish an intended system boundary, then verify that boundary on every required band. A position that works at 3.5 MHz may not provide the same suppression at 28 MHz. Avoid fixed prescriptions such as “always use two chokes” or “measure a set distance after the transformer”: the correct number, impedance and placement follow from the installed current paths and the required result.
Tuners Need an Explicit State and Plane
A tuner can make its input look close to the chosen system impedance while its output still sees a demanding complex load. A sweep through a tuned tuner therefore verifies the tuner's input result; it does not reveal the raw antenna or feedline load.
- To measure the raw system at the tuner input, use a documented bypass state or remove the tuner from the path.
- To measure what the transmitter sees after tuning, leave the tuner in its declared tuned state and connect at the transmitter-side plane.
- Do not assume “power off” means bypass. Latching relays can retain their last switched state, and switching details vary by model.
- Do not infer full-power suitability from the analyzer's low-level sweep. Tuner loss, arcing, component current, thermal rise and control behaviour require separate validation under the permitted operating conditions.
Balanced Line Needs a Balanced Measurement Boundary
A single-ended analyzer connected directly across ladder line introduces its own unbalance and common-mode path. For practical field work, use a characterized wideband fixture or balun and include or de-embed its response within its measured limits. For laboratory work, a multiport VNA can measure single-ended ports and convert them to differential and common-mode quantities; Rohde & Schwarz documents this mixed-mode method for balanced devices.
Whatever method is used, state the differential reference impedance, fixture topology and calibration plane. A convenient balun with unknown amplitude balance, phase balance and common-mode impedance is part of the measurement, not an invisible adapter.
A Safe, Repeatable Field Workflow
- Define the question. Decide whether you need antenna-terminal impedance, an assembled antenna subsystem, the remote-tuner input or the load seen in the shack.
- Freeze the installed geometry. Record radiator position, return structure, feedline route, chokes, bonds, tuner state and nearby conductors.
- Make the system safe. Disconnect and positively inhibit transmitters and amplifiers. Do not measure during thunderstorms or with nearby transmitters active. Discharge the line using the instrument manufacturer's approved method; never use the analyzer as the discharge path.
- Check the analyzer limits. Confirm connector, frequency, maximum external RF and DC limits. Remove bias voltage and any active device that can drive the port unless the analyzer documentation explicitly supports the setup.
- Calibrate at the intended plane. Use suitable standards and avoid disturbing the cable after calibration. If that plane is inaccessible, document the cable or fixture model and the exact port-extension or de-embedding method.
- Measure the feedpoint plane when needed. Save R, X, |Z| and SWR across the complete frequency span; repeat after a controlled fixture-position change to test sensitivity.
- Measure the shack plane when needed. Keep the normal feedline and declared tuner state in circuit so the sweep represents the station boundary.
- Relate the planes. Use measured cable length, velocity factor, attenuation and characteristic impedance, or a characterized two-port file. Investigate differences that the line model cannot explain.
- Check exterior current. Map the installed line with a calibrated or transfer-factor-documented current probe when common-mode behaviour could affect the conclusion.
- Archive the evidence. Keep raw sweeps, calibration details, fixtures, reference plane, routing, weather and uncertainty so a future change can be compared honestly.
Analyzer safety: manufacturer instructions are controlling. RigExpert's AA-230 ZOOM manual, for example, says not to connect during thunderstorms, not to inject RF or DC into the antenna port, not to connect with active nearby transmitters, and to control static before connection. Apply the limits and discharge procedure for your exact instrument and installation.
How Common Multiband Systems Change the Choice
| System | Useful feedpoint-side question | Useful shack-side question | Special boundary to verify |
|---|---|---|---|
| Centre-fed dipole or OCF dipole | Impedance on the feedline side of the actual balun or choke | Load delivered through the installed coax | Current balance and exterior-coax current, especially with asymmetry |
| Vertical with radials | Terminal load with the intended radial system connected | Complete load reaching the tuner or transmitter | Analyzer and temporary cable must not replace or alter the intended return structure |
| End-fed wire | Load presented by the radiator and declared return conductor to the matching device | Installed transformer, feedline and common-mode result at the station | Which exterior-coax section, counterpoise or other conductor is intentionally part of the antenna |
| Remote-tuned wire | Raw wire-and-return load at the tuner's output plane | Tuned input presented through the station feedline | Exact bypass or retained relay state and the tuner's control/bias arrangement |
| Ladder-line doublet | Balanced load at a defined differential plane | Load at the balanced tuner's or transition's declared input | Fixture or mixed-mode method, line routing and common-mode conversion |
Engineering References
- Keysight FieldFox: Network Analyzer Mode, calibration and port-extension limits
- Keysight Fixture Simulator: embedding, de-embedding and reference planes
- Keysight Network Analysis: transmission-line theory, reflection and calibration
- Rohde & Schwarz 1EZ53: balanced and mixed-mode VNA measurement
- Fair-Rite: Notes on Impedance Measurement
- RigExpert AA-230 ZOOM Antenna and Cable Analyzer User Manual
Mini-FAQ
- Should I measure a multiband antenna at the feedpoint or in the shack? Measure at the feedpoint for antenna-terminal impedance and in the shack for the load presented to station equipment. Take both when both questions matter, and record each reference plane.
- Does a different shack reading mean the feedline is faulty? Not necessarily. A mismatched line normally transforms complex impedance with electrical length, and loss attenuates reflection. Compare the result with a characterized line model before diagnosing a fault.
- Can port extension remove any measurement cable? No. Basic port extension corrects electrical delay and may optionally estimate loss; it does not generally remove fixture mismatch, unknown mode conversion or radiation. Full de-embedding requires a valid fixture characterization.
- Is the first choke always the best measurement plane? No. Choke impedance and its installed effect vary with frequency, geometry and the complete common-mode path. Define the intended boundary, then verify exterior current and repeatability on every required band.
- Can I measure through an antenna tuner? Yes, but the result belongs to the tuner's input in its declared state. It does not reveal the raw output-side load. Use documented bypass or remove the tuner when that raw load is the target.
- Can a low-power analyzer sweep prove the antenna is safe at full power? No. It does not establish tuner loss, arcing margin, conductor or component current, ferrite heating, common-mode voltage, RF exposure or behaviour at the intended waveform and duty cycle.