When a transmission line is terminated with an impedance, ZL, that is not equal to the characteristic impedance of the transmission line, ZO, not all of the incident power is absorbed by the termination. Part of the power is reflected back so that phase addition and subtraction of the incident and reflected waves creates a voltage standing wave pattern on the transmission line.
The ratio of the maximum to minimum voltage is known as the Voltage Standing Wave Ratio (VSWR) and successive maxima and minima are spaced by 180° (l /2).
Also of considerable interest is the mismatch loss. This is a measure of how much the transmitted power is attenuated due to reflection. It is given by the following equation:
Mismatch Loss = -10 log ( 1 - r2 )
For example, an antenna with a VSWR of 2:1 would have a reflection coefficient of 0.333, a mismatch loss of 0.51 dB, and a return loss of 9.54 dB (11% of your transmitter power is reflected back).
In some systems this is not a trivial amount and points to the need for components with low VSWR.
only in situations where the length of the feedline is sufficient to introduce substantial amounts of loss is it absolutely necessary to measure swr at the feedpoint of the antenna itself as under these conditions the line (and the excessive length) will create an illusion of having a better antenna VSWR than is actually the case.
for the short lengths that are commonly used by most of us in mobile applications measurements at emax/emin points at the appropriate electrical 1/2 wave intervals are more than sufficient, providing of course that the feedline is 180° in length or longer to begin with.
for example, when working with 50 ohm feedline there is a range of real impedances available along the length of the line from roughly 33 - 75 ohms when the measured swr is 1.5:1. the correlation between the range and the swr is directly related. reactance aside for the moment, since these abnormal dynamic impedance values are available along the line then being able to determine the spots in the line where emax/emin gives values of impedance closer to 50 ohms can be used to restore full output when the feedline is terminated at these points at the source.
the question then becomes, are there also wider ranges of real impedance points along the line at even higher levels of swr? the answer to this question is yes. and there are non-resonant antenna designs in the cb market. one that comes to mind is the francis line of antennas. while in a given situation where the manufacturers recommendation in reference to the required feedline length is completely ignored, as in the case where a 12' foot line of RG8M is installed instead of the recommended 18' length of the same type of feedline, measured swr at the source can rise as high as 3:1 depending on the end of the 40 channel band that we're talking about. in retrospect, when the 12' foot line is replaced with 18' feet of the same feedline we see the swr across the band drop significantly with all channels showing below 1.3:1, the best match in the center of the band and rising slightly at the ends. here is one actual and specific instance where the "whatever it takes" approach simply just doesn't work. this isn't the only one.
it's clear here that even at an swr as high as 3:1 that there is a real range of available dynamic impedance values along the line from at least 16.66 - 150 ohms that when used with a non resonant antenna design can be length adjusted to improve the match but more importantly to allow the transmitter to see some value of impedance closer to the required 50 ohms that is required to facilitate maximum transfer of available developed transmitter power. excessive swr measured at the feedpoint in and of itself is not the problem, since ALL POWER developed by the transmitter is ultimately absorbed and radiated, regardless of the match. and here is the REAL PROBLEM. the problem IS the reduction of power developed into the load due to the reflection mismatch presented there.
the mismatch loss represented by a 3:1 swr is roughly 25%. however, if we can restore the impedance that the transmitter sees looking into the feedline at the termination point to something that more closely resembles 50 ohms we can eliminate a large portion of the mismatch loss. at that point the antenna has no choice but to absorb and radiate. the swr is still 3:1 as measured at the feedpoint but now there is negligible reduction in transmitter power.
now, coming back to reactance, either inductive or capacitive values can be cancelled by the use of stubs. an open stub under 1/4 wavelength can be used to cancel inductive reactance and a shorted stub can be used to cancel capacitive reactance.
these are but a few of the many principles that when applied allow what most of us would consider to be ridiculous levels of swr to be tolerated at frequencies in the ghz. band involving satellite and space communications without a second thought.
amazingly enough, it's the very existence of reflection from the load present on the line to begin with that allows us to make it all work, regardless of the particular solution applied. just another example attesting to the fact that swr is not the be all end all indicator to efficient antenna performance, something else that Maxwell alluded to in his writings/papers and lectures.
