This article reports a simple but robust measurement of the ReturnLoss of a pair of UHF adapters from 1 to 501MHz.
Load selection
To measure ReturnLoss of the adapter set, it must be terminated with a quality load, one that has an ReturnLoss better than the expected measurement.
Two loads are considered:
- Port 2; and
- a reference load.
Port 2 load
Above is a measurement of Port 2 ReturnLoss. This load cannot be used to measure ReturnLoss of higher than 14dB at 500MHz to uncertainty less than 3dB, so it is not a suitable load for measurement of ReturnLoss of the adapter set.
Reference load
The best available reference load is a SMA load that measures DC resistance of 49.805Ω and so we might infer that it has a low frequency ReturnLoss of 54dB.
This load was used to calibrate the VNA.
Let’s say we will accept uncertainty in ReturnLoss measurement of ±3dB.
Above is calculation of the ReturnLoss that can be measured using the 54dB reference to uncertainty of ±3dB, the result is 43dB.
The load will be a little poorer at higher frequencies, but the VNA firmware does not provide for a more detailed model of the load (or other calibration parts).
Above is measurement of the adapter set terminated in the reference load detailed above.
At 500MHz, the adapter pair could not be used in a test fixture to measure ReturnLoss of more than 10dB to uncertainty of ±3dB. ReturnLoss=10dB with uncertainty of ±3dB is equivalent to a VSWR confidence interval of 1.59-2.36.
At 100MHz, the adapter pair could not be used in a test fixture to measure ReturnLoss of more than 24dB to uncertainty of ±3dB. ReturnLoss=24dB with uncertainty of ±3dB is equivalent to a VSWR confidence interval of 1.09-1.18.
What is going on?
Above is a Smith chart view of s11 which gives insight. Note that it is an arc from about chart centre at lowest frequency, clockwise as frequency increases.
Let’s zoom in on it for more detail and analysis.
Above is a zoomed in view. The trace is not quite a circular arc, but there is a strong underlying circular trend with centre of the arc somewhere around 40+j0Ω which hints that the imperfection can be approximated as a short section of 40Ω transmission line in this case.
A 50mm bulkhead adapter example
Above is a longer adapter, a 50mm bulkhead adapter commonly used in panels for wall penetrations etc.
Above is a ReturnLoss plot to 500MHz. Note the different scale to the earlier chart.
Above is a Smith chart view of s11 which gives insight.
See Study of suitability of UHF bulkhead adapter to a Diamond x50-A antenna system for a deeper analysis of this adapter.