Antenna Polarization: The 3 dB You Cannot See

Two antennas with identical gain, identical power and identical readers produce completely different read zones. The difference is polarization, and it is the parameter most often chosen by habit rather than by measurement.

What polarization actually is

Polarization describes the plane in which the electric field oscillates.

• A linear antenna radiates in one plane, horizontal or vertical, and a tag aligned to that plane receives it best.
• A circular antenna radiates a field that rotates, so it can be received by a tag in any rotation.
• Circular comes in right-hand and left-hand variants, and a tag matched to one is rejected by the other.
• A tag rotated 90 degrees against a linear antenna loses most of its coupling, which is the classic "it works until it is turned sideways".
• Real antennas are not ideal; an "elliptical" pattern is a linear antenna with some circular tolerance and less gain.

The 3 dB question, honestly

The 3 dB loss of circular polarization is real, and it is not always a loss.

• A circular antenna splits its energy across two orthogonal components, so each component carries half the power: that is the 3 dB.
• A tag aligned with the field recovers that loss only partly; the number is a design trade, not a defect.
• In exchange you gain orientation tolerance, which is worth far more than 3 dB when tags are randomly oriented.
• If every tag in the process is guaranteed to be aligned, linear gives you the extra decibels for free.
• If any tag might be rotated, circular will beat linear on average, by a wide margin.

Where each one wins

The correct choice follows the material flow, not the datasheet comparison.

• Apparel on a rail: tags hang at arbitrary rotations, so circular is almost always correct.
• Pallet portals: cartons present mixed orientations, and circular tolerates that.
• Conveyor reading of flat cartons with a fixed orientation: linear can double the range.
• Document and file tracking, where labels lie flat: linear is efficient.
• Metal surfaces and liquids: polarization interacts with the reflection, so test both rather than assume.

Read zone shape is not a circle

The "read zone" drawing in a brochure is a simplified top view, and the real one is a three-dimensional lobe.

• The main lobe narrows as gain rises, so a higher-gain antenna reads further but not wider.
• Side lobes exist, and they are where the neighbouring aisle picks up your stock.
• Tilting an antenna downward moves the zone, and can be cheaper than adding power.
• Concrete and steel change the lobe through reflection; the zone you measured in a corridor is not the zone you get in a rack.
• Zone boundaries are gradual, not a clean line, which is why portal design needs a debounce rule.

Controlling the zone in practice

Read-zone control is a system of small decisions, not a single setting.

• Set power to the lowest level that satisfies the process, then verify with the stock present.
• Aim the antenna so the main lobe follows the flow of goods rather than crossing a boundary.
• Use racking and steel as shielding, which costs nothing and removes side-lobe reads.
• Separate readers in time as well as in space; a scheduled window is the cheapest shield.
• Measure with the product that will actually be there, because an empty shelf flatters every configuration.

Tags are half the equation

Antenna selection cannot compensate for a tag whose orientation changes inside the box.

• A tag lying flat in a carton may be vertical after the carton is stacked, so design for both.
• Some tags are designed to be orientation-insensitive, at the cost of range or size.
• Two tags placed orthogonally can stabilise reads when orientation is truly random.
• A tag on a curved surface bends its own antenna, shifting its resonant frequency.
• Test the tag and antenna as a pair; neither has a meaningful specification alone.

Measuring instead of arguing

Polarization debates are settled in ten minutes with a tag and a walk.

• Fix the reader and antenna, then rotate a test tag through 360 degrees and log the read rate at each angle.
• Repeat with the product in place; the product changes the answer more than the antenna does.
• Walk the boundary of the intended zone and mark where reads stop, on both axes.
• Test at the highest and lowest stock levels, because the vertical lobe matters as much as the horizontal.
• Compare two antennas on the same reader in the same session; comparing datasheets is not comparing hardware.

Stacking, nesting and the vertical lobe

A zone measured on an empty rack says nothing about a full one.

• Stacked cartons put tags at different heights, and the vertical beam profile decides which layer reads.
• Nesting turns a flat label into a vertical one, which is exactly where linear polarization fails.
• Dense stock absorbs energy, so a range measured in free air does not survive contact with goods.
• Tags close together detune each other; keep spacing above the manufacturer's minimum when the zone is tight.
• If one layer is systematically missed, move the antenna rather than raise the power.

Documenting the choice so it survives you

An antenna decision that is not written down will be reversed by accident.

• Record polarization, gain, tilt, power and mounting height per portal, with photographs taken from the operator's viewpoint.
• State which product orientations the installation was validated against, and which were not.
• Note the rejected alternative and the reason, or the next engineer will re-open the argument from scratch.
• Re-validate when the carton, the tag or the racking changes, because those three drive most zone regressions.
• Hand over one measured read-rate figure per lane, so a later drop shows up as a number rather than a complaint.

Common mistakes

These four recur in almost every review of a struggling site.

• Choosing circular everywhere "for safety", then complaining about range on flat cartons.
• Choosing linear everywhere "for range", then finding a third of the stock unreadable.
• Raising power to fix a polarization mismatch, which widens the zone and creates cross-reads.
• Judging the installation by a single successful pass with a test tag held in the hand.

The fix in each case is the same: define how the tag will actually be attached and oriented, then choose the antenna to match that reality, and confirm it with the stock on the floor.