Coax cable phase stability: the spec that ruins your phased array

Conclusion first, reasoning after.

1. Phase stability with temperature is usually specified in ppm per degree C, and it matters more than insertion loss.
2. Phase stability with flexure matters if the cable moves — even a fixed install can shift during maintenance.
3. Foam dielectrics are lighter and lower loss but generally worse for phase stability than solid PTFE.
4. Match cable lengths from a single batch, and keep spares from the same batch.

Discussion welcome, especially counter-examples.

Agreed on the ranking: for phased arrays and interferometers, phase vs temperature and phase vs flex usually dominate insertion-loss shopping.

Practical addenda:

1. Spec reading. "ppm/°C" is often electrical length change normalized to physical length. Convert to degrees at your frequency: Δφ(deg) ≈ 360 · f · L · (Δτ/τ) with Δτ/τ from the ppm figure (and confirm whether the vendor quote is group delay, phase, or mechanical length). A cable that looks fine at 1 GHz can eat your beamforming budget at 10–20 GHz.

2. Dielectric and construction. Solid PTFE is usually more predictable with temperature than foam/air-spaced low-density dielectrics; foam wins loss and weight, loses phase stability and crush resistance. Phase-stable "armor" or low-density PTFE constructions exist specifically for this trade - read the temperature *and* flexure curves, not only IL @ 25 °C.

3. Matching discipline. Same batch, same length cut, same bend radius history, and spares from that batch - yes. Also keep a golden reference cable in the cal kit drawer and re-check relative phase after any rework. For arrays, document which physical cable sits on which element; swapping "identical" assemblies without a phase note is a classic late-night failure mode.

Open question: at what frequency and array size did phase-stable coax stop being enough for you, and you moved to semi-rigid, delay trim, or active calibration?

— rfforum 编辑部 / Editorial Team