Comment on W6VAT article on VSWR

A correspondent has challenged some of the discussion of VSWR on VK1OD.net based on James G Lee's (W6VAT) Antennex article entitled The Effects of VSWR on Transmitted Power. This article is a response.

W6VAT's article gives an approximate model for understanding standing waves on a practical transmission line, but the approximation gives rise to some differences with the more detailed and more accurate model used in RF Transmission Line Loss Calculator and RF Two Wire Transmission Line Loss Calculator. My own VSWR calculator does use a simpler model similar to that used by W6VAT, and is less accurate as discussed on the calculator page.

W6VAT's Fig 3 sets out some loss figures for a range of VSWR figures. The suggestion is that the reduction of power delivered to a load with a certain VSWR is simply a function of the line VSWR at the load end. The quantity he refers to as Transmission Loss is more commonly known as Mismatch Loss (and Transmission Loss usually has a different meaning).

MisMatchLoss(dB)=10*log(1-ρ²) where ρ=|Γ| and Γ is the complex voltage reflection coefficient. Since ρ=(VSWR-1)/(VSWR+1), MisMatchLoss(dB)=10*log(1-((VSWR-1)/(VSWR+1))²).

So MisMatchLoss for VSWR=2 is 0.512dB, easy!

But what does it mean?

MisMatchLoss is the ratio of power available from an ideal source in a load that is the conjugate of the source impedance, to the power in the mismatched load. It means here that if a load with VSWR=2 was connected to the source, that the power in the load would be 0.512dB less than in a matched load.

So what is wrong with Fig 3 and the related discussion?

It is the assumption that ham transmitters are well represented as an ideal linear source with a Thevenin equivalent source impedance of 50+j0Ω.

It is a fairly simple experiment to measure 'Forward Power' and 'Reflected Power' for an ideal load and mismatched load using a quality directional wattmeter, then to calculate VSWR and net power, and compare the results with Fig 3. Fig 3 will not give a good estimate in general, it may sometimes agree with measurement, but not in general. In fact, in cases where 'Forward Power' varies with load VSWR for whatever reason, Fig 3 will be wrong. a write-up of such an experiment is at A test of Zs on an IC7000.

For more detailed discussion, see MisMatch Loss given VSWR.