R
Member
Calibration removes systematic errors up to the calibration reference plane; it does not remove what is between that plane and the DUT. Choosing the method is mostly about deciding where that plane can physically be placed.
SOLT: convenient with coaxial standards, broadband, but depends on how well the open/short/load standards are modelled.
TRL: needs a thru and a line with a known phase difference (commonly 20-160 degrees), works well for non-coaxial media such as microstrip or wafer probes, but is band-limited per line.
ECal / unknown-thru: fast and repeatable, good for production setups, still needs a valid kit definition.
Practical sequence:
1) Define the reference plane at the DUT terminals. If a fixture, adapter or launch is in between, plan a de-embedding or fixture-characterisation step.
2) Warm up the instrument, set IF bandwidth and averaging, then calibrate with the same cable routing used for measurement. Moving cables after calibration is a classic source of phase error.
3) Verify the calibration on a known device (a matched line, a known attenuator, or a check standard) before measuring the DUT.
4) Record calibration kit, method, frequency range, power level, IF bandwidth and date in the data file.
Common mistakes: trusting a two-port calibration after changing port power; forgetting that a 3.5 mm to SMA adapter has its own loss and mismatch; comparing S-parameters measured with different reference planes.
Takeaway: report the calibration method together with the reference plane. Without that, two labs can measure the same DUT and legitimately disagree.
SOLT: convenient with coaxial standards, broadband, but depends on how well the open/short/load standards are modelled.
TRL: needs a thru and a line with a known phase difference (commonly 20-160 degrees), works well for non-coaxial media such as microstrip or wafer probes, but is band-limited per line.
ECal / unknown-thru: fast and repeatable, good for production setups, still needs a valid kit definition.
Practical sequence:
1) Define the reference plane at the DUT terminals. If a fixture, adapter or launch is in between, plan a de-embedding or fixture-characterisation step.
2) Warm up the instrument, set IF bandwidth and averaging, then calibrate with the same cable routing used for measurement. Moving cables after calibration is a classic source of phase error.
3) Verify the calibration on a known device (a matched line, a known attenuator, or a check standard) before measuring the DUT.
4) Record calibration kit, method, frequency range, power level, IF bandwidth and date in the data file.
Common mistakes: trusting a two-port calibration after changing port power; forgetting that a 3.5 mm to SMA adapter has its own loss and mismatch; comparing S-parameters measured with different reference planes.
Takeaway: report the calibration method together with the reference plane. Without that, two labs can measure the same DUT and legitimately disagree.
