FR1 conducted vs OTA: when a clean conducted EVM still fails the chamber

[Technical article, rfforum Editorial Team]

A UE or CPE that passes conducted EVM and then fails OTA is a common programme delay. The conducted path removes antenna mismatch, body/fixture loading, and chamber multipath — exactly the terms that show up in the field.

1. What conducted hides
Antenna VSWR and efficiency, front-end load-pull into a non-50 Ω antenna, and housing detuning never appear on a conducted jig that presents 50 Ω.

2. Bridging the gap
- Keep a conducted baseline, then add a radiated smoke test early (even a simple chamber or quiet room)
- Load-pull or at least VSWR-stress the PA/FEM into expected antenna Γ before locking EVM gates
- Treat OTA failures as antenna + FEM interaction first, not only "modem settings"

3. Habit
Do not freeze marketing EVM numbers from conducted-only data for an antenna-integrated SKU.

Discussion: at what programme milestone do you require first OTA EVM, and what antenna Γ window do you load-pull before that gate?

— rfforum 编辑部 / Editorial Team

Conducted EVM is necessary and insufficient for antenna-integrated SKUs. Early OTA smoke + Γ stress closes the surprise gap.

When is first OTA mandatory in your schedule?

— rfforum 编辑部 / Editorial Team

We scoped the waveform on this once and could see the anomaly clearly. Worth a look.