Devices inserted in transmission lines often characterised by one or more of:
- Insertion VSWR (the input VSWR when terminated with a matched load);
- Return Loss (RL) in dB (20 times the log of the magnitude of the complex reflection coefficient); and
- Insertion Loss.
Practitioners often find Insertion VSWR (1) of most use as it indicates whether the device is worse than other system devices, the weak link in the chain if you like. You might see a coax antenna switch specified to have InsertionVSWR<1.2 to 60MHz.
Return Loss (2) is a function of VSWR and vice versa, so it appeals when the designer thinks in terms of Return Loss rather than VSWR (and it is a better metric for VSWR<1.2). You might see a coaxial relay specified to have ReturnLoss>30dB to 500MHz.
Insertion Loss (3) is not so readily compared to the other two which are measures of input reflection with a matched termination. It often yields some numbers that appear very acceptable, but might be deceptively so. You might see a coaxial relay specified to have ReturnLoss>30dB to 500MHz. You might see a coax antenna switch specified to have InsertionLoss<0.2dB to 100MHz.
Insertion Loss has two components, Transmission Loss and (input) Mismatch Loss. InsertionLoss=TransmissionLoss+MismatchLoss.
In many applications, the device has relatively low Transmission Loss, so InsertionLoss≈MismatchLoss. This is often the case for adapters, coax switches, coax relays below say 500MHz. It is not the case for attenuators, or most filters.
In turn, MismatchLoss is related to VSWR, so for the case where TransmissionLoss is negligible we can convert InsertionLoss to VSWR and vice-versa.
Above is a plot of InsertionVSWR vs InsertionLoss for a lossless device. It can be used to convert one to the other for comparison.
Taking the third example above, a coax antenna switch specified to have InsertionLoss<0.2dB to 100MHz is equivalent to InsertionVSWR<1.6.
Whilst some practitioners might regard InsertionLoss of 0.2dB to be quite acceptable, the InsertionVSWR≈1.55 might be more concerning in terms of the way in which money is allocated to different devices for an overall performance level.
Above is a plot from RFSim99 of a 40mm length (vf=0.66) of 35Ω lossless transmission line (as might be expected from a couple of UHF series adapters). The red curve is S21, InsertionLoss=-S21 in this case; and the blue line is S11, RL=-S11. It can be seen that although RL climbs steadily, InsertionLoss climbs very slowly at first. Whereas at 500MHz InsertionLoss=0.2dB, RL=13.6dB and from the graph above VSWR=1.55, at one tenth the frequency (50MHz), InsertionLoss=0.00235dB, RL=32.7dB, and from the graph above VSWR=1.05.
A real world example
The Zetagi V3 coax switch is advertised to have InsertionLoss=0.2dB at 500MHz (to imply perhaps no worse than 0.2dB below 500MHz).
The switch probably has very low transmission loss, and so from the graph above we might expect that InsertionVSWR≈1.55 (RL=13.3dB), perhaps a little better depending on transmission loss (which is not good either). The device appears to use UHF series connectors, so it would be quite naive to expect very low InsertionVSWR to 500MHz (Exploiting your antenna analyser #12). InsertionVSWR to say 60MHz would be a more relevant, more interesting parameter.
When sellers use a less common metric to characterise their product, it might just be that they are trying to frustrate apples for apples comparisons with competitor’s products.
The relatively small value of InsertionLoss appeals to the cry for mediocrity in modern ham radio,
if it doesn’t cost an S-point, it doesn’t matter.
Footnote: Mismatch Loss though not a particularly complex concept, is not a well understood concept in the ham community. Whilst often misused, it is correctly used above.