L
Admin
[Technical article, rfforum Editorial Team]
Phase and delay stability of the test jumper often dominate VNA uncertainty after the instrument itself is warmed up. Teams chase calibration kits and connector torque while a cable that was coiled differently between sweeps explains the whole delta.
1. What "phase stable" means
Vendors quote typical phase change vs flex and vs temperature. A cable that is excellent when fixed in a fixture can still move several degrees when you open a fixture door or drape it over a heat source. Absolute phase for embedding/de-embedding and for multi-port fixture removal cares about this.
2. Flex and memory
Repeated bend cycles change the dielectric geometry. If you calibrate, then re-route the cable for the DUT, you invalidated the cal in phase even if |S11| still looks pretty. Fixture the cable path; do not hand-form a new arc for every board.
3. Temperature
Thermal gradients along a cable look like electrical length change. Keep cal and measure at similar temperature; avoid running a hot PA exhaust across the measurement jumper.
4. Habits that help
- Dedicated, labelled phase-stable jumpers for cal vs rough RF work
- Torque and connector savers; never use the precision jumper as a daily torque wrench victim
- A quick "cable wiggle" test after cal: if phase jumps, fix the mechanical path before the DUT
Discussion: do you keep a written flex/temperature budget for phase-critical VNA work, or only swap to phase-stable cables when a program complains?
— rfforum 编辑部 / Editorial Team
Phase and delay stability of the test jumper often dominate VNA uncertainty after the instrument itself is warmed up. Teams chase calibration kits and connector torque while a cable that was coiled differently between sweeps explains the whole delta.
1. What "phase stable" means
Vendors quote typical phase change vs flex and vs temperature. A cable that is excellent when fixed in a fixture can still move several degrees when you open a fixture door or drape it over a heat source. Absolute phase for embedding/de-embedding and for multi-port fixture removal cares about this.
2. Flex and memory
Repeated bend cycles change the dielectric geometry. If you calibrate, then re-route the cable for the DUT, you invalidated the cal in phase even if |S11| still looks pretty. Fixture the cable path; do not hand-form a new arc for every board.
3. Temperature
Thermal gradients along a cable look like electrical length change. Keep cal and measure at similar temperature; avoid running a hot PA exhaust across the measurement jumper.
4. Habits that help
- Dedicated, labelled phase-stable jumpers for cal vs rough RF work
- Torque and connector savers; never use the precision jumper as a daily torque wrench victim
- A quick "cable wiggle" test after cal: if phase jumps, fix the mechanical path before the DUT
Discussion: do you keep a written flex/temperature budget for phase-critical VNA work, or only swap to phase-stable cables when a program complains?
— rfforum 编辑部 / Editorial Team
