Phasing Harness / Power Divider for Stacked Arrays
Any array with more than one driven element — a stacked Yagi pair, a Lazy H, a 4-square vertical array, or a crossed-Yagi for satellite work — depends on a phasing harness to split power between elements and hold them at the correct phase relationship. Get the harness wrong and the array's carefully designed pattern simply doesn't form, no matter how well the elements themselves are built. This guide covers how transmission line length sets phase, a phasing-line-length calculator, equal-split power dividers, and the assembly details that matter most.
What a phasing harness actually does
A multi-element array's radiation pattern — its gain, directivity, and front-to-back ratio — depends not just on element spacing but on the phase relationship between the currents in each element. A phasing harness is the feed network that sets that relationship: it splits power from a single feedline between two or more elements while introducing a deliberate, controlled phase difference (or deliberately keeping them in phase) between the ports. Get the split even but the phase wrong, and the elements can partially cancel each other instead of reinforcing — the array can end up performing worse than a single element on its own.
How line length sets phase
A signal traveling down a transmission line accumulates phase delay proportional to the line's electrical length, which depends on both its physical length and the cable's velocity factor (how much slower the signal travels in that cable compared to free space). Two feedlines of different lengths reaching two elements from a common split point therefore deliver the signal to each element at a different phase — cut the length difference to correspond to your target phase angle, and you've built a phasing harness.
Equal-split power dividers
- Direct parallel connection with a matching section: two equal-length, equal-impedance feedlines connect in parallel at a common point, which halves the impedance seen at that point (two 50Ω lines in parallel present 25Ω) — a quarter-wave section of 75Ω-ish coax transforms that back up to 50Ω for the main feedline. This is the classic method behind most stacked-Yagi and Lazy H feed harnesses.
- Equal lengths keep the split in phase: when both branch lines are cut to the same electrical length, the split is a pure equal-power, in-phase divider with no deliberate phase offset — the phase-shaping in an array like this comes from physical element spacing and orientation, not the harness itself.
Where phasing harnesses show up across this site's arrays
The Lazy H Stacked Array and stacked-Yagi style arrays typically use an equal-split, in-phase harness (the physical bay spacing does the phase work). The Satellite Crossed Yagi uses a 90° phase delay harness between its two crossed elements to produce circular polarization. The 4-Square Phased Array uses a more involved feed network with 90° phase steps between successive elements around the square, typically combined with impedance-matching networks at each element rather than a simple delay line alone — that array's own guide covers its specific feed network in full; this page covers the general phasing-line-length principle those designs build on.
| Harness Type | Purpose | Typical Application | Key Design Parameter |
|---|---|---|---|
| Equal-split, in-phase divider | Split power evenly with no deliberate phase offset | Stacked Yagi pairs, Lazy H bays | Equal branch line lengths, plus an impedance-matching section at the split point |
| 90° phase delay harness | Introduce a quarter-cycle phase offset between two feeds | Crossed-Yagi circular polarization, some phased-array elements | One branch line cut a quarter-wavelength (electrical) longer than the other |
| 180° phase delay harness | Feed an element fully out of phase with another | Certain bidirectional and end-fire array configurations | One branch line cut a half-wavelength (electrical) longer than the other |
Phasing Line Length Calculator
Materials for a two-element phasing harness
A two-way phasing harness with a weatherproofed T-junction and clearly labeled branch lines.
Building a Phasing Harness
Measure your actual cable's velocity factor before cutting anything — the number printed on the jacket is nominal, not a guarantee for your specific reel.
Determine the required phase and split for your array
Check your specific array design (stacked Yagi, Lazy H, crossed-Yagi, or similar) for the phase relationship and power split it needs between elements. Most stacked/bay arrays want an equal, in-phase split; crossed-Yagi and some phased designs want a deliberate 90° or 180° offset.
Measure your actual cable's velocity factor
Use a NanoVNA to measure the true velocity factor of the specific coax reel you're using, rather than trusting the manufacturer's nominal published value. Even a small VF error compounds into a real phase error over a long phasing line, especially at VHF/UHF where the whole line is physically short to begin with.
Calculate and cut each branch line
Use the calculator above with your measured VF to find the length for each phased branch, then cut each line with a small amount of extra length and trim to final size after an initial measurement — it's much easier to trim a bit more off than to add length back.
Assemble the split point and matching section
Assemble the T-junction or split block, adding a quarter-wave matching section if you're using the direct-parallel equal-split method. Keep connections at the split point mechanically solid — this point carries the full array's power and takes ongoing weather exposure at most installations.
Test and weatherproof before final installation
Check each branch line's electrical length and the overall harness's SWR with a NanoVNA before mounting the array. Weatherproof every connector in the harness with self-amalgamating tape or heat-shrink, since a harness typically has more exposed connections than a simple single-feedline run.
| Symptom | Most likely cause | Diagnosis | Fix |
|---|---|---|---|
| SWR looks fine but the array's pattern doesn't match the expected gain/directivity | Phased branch lines swapped between elements, or a length error in one branch | Check each labeled line goes to the correct element, and re-measure each branch's electrical length against the calculated target | Correct the wiring to match the design, or re-cut a mislengthed branch line |
| SWR is elevated at the main feedline input despite good element resonance | Missing or incorrectly sized matching section at an equal-split T-junction | Check the impedance transform math for your specific split method (e.g. two 50Ω lines in parallel present 25Ω, needing a quarter-wave 75Ω-ish section to restore 50Ω) | Add or correct the matching section length/impedance at the split point |
| One element runs noticeably hotter or the harness shows uneven power distribution | Branch lines aren't actually equal length/impedance despite being labeled as an equal split, or a connector at the split point is faulty | Re-measure both branch lines' electrical length and check each connector for a clean, low-resistance connection | Re-cut the mismatched branch or repair/replace the faulty connector |
| Phasing works correctly on the design frequency but the array performs oddly off-frequency | Normal behavior — a phasing line's electrical length (and therefore its phase delay) is frequency-dependent, so it's only exactly correct at its design frequency | Compare the operating frequency against the harness's design frequency | Accept the narrower effective bandwidth as inherent to fixed-length phasing harnesses, or design specifically for the center of your intended operating range |
| Water intrusion or corrosion found at the split point after a season outdoors | Inadequate weatherproofing at the T-junction or one of the branch connectors | Inspect all connections at and near the split point for moisture or corrosion | Clean, repair, and re-weatherproof with self-amalgamating tape, paying particular attention to the split point since it usually has the most connectors in one place |
Does the phasing harness itself set the array's gain?
Not directly — the harness sets the phase and power-split relationship between elements, but the actual gain and pattern shape also depend heavily on physical element spacing and the individual elements' own design. The harness is necessary to realize the array's designed pattern, but it isn't the sole source of gain.
Does the coax's velocity factor really matter that much?
Yes, especially at VHF/UHF where the phasing line itself is physically short — a velocity factor error of even a few percent shifts the achieved phase delay by a real, measurable amount relative to a short line's total length. Measuring your actual cable's VF with a NanoVNA rather than trusting the nominal datasheet value is worth the extra few minutes.
Is a 90° phasing line the same as a quarter-wave matching section?
They're built on the same underlying phase-length relationship, but they do different jobs. A quarter-wave matching section is specifically sized to also perform an impedance transformation between two different impedances. A pure phasing line's job is only to introduce a phase delay — it doesn't necessarily do any impedance transforming on its own, though the two functions can be combined in some harness designs.
Can a phasing harness handle full legal-limit power?
Yes, as long as the coax, connectors, and any matching section components are rated for your actual power level — this is fundamentally a coax and connector sizing question, not a limitation of the phasing-harness concept itself. Size every component in the harness the same way you would for a single high-power feedline run.