I’ve heard this for decades, and I first heard it when I got into CB radio when I was a kid.
You’ve gotta to cut your coax to the right length if you want low SWR’s
More accurately, if you are driving a mismatched load (e.g: a poorly tuned antenna) it depends where along the feeder you measure your SWR. In the following animation, there is a coaxial cable connected to the transmitter which is a multiple wavelength long, (in this case, four wavelengths) and nothing connected to the other end of the cable where the antenna usually goes. It is an open circuit. This is a most extreme case of mismatch, where (for sake of argument) 100% of the transmitter power is reflected.
When you key your transmitter, the transmitted (incident) wave travels down the coaxial cable and meets an impedance mismatch (in the above case, an open circuit) and no current is leaving the cable since there is no load. If no load is present for the transmitted energy to be dissipated into and/or radiated from, the energy has nowhere to go except back where it came from. Current cannot flow in an open circuit, and if there is zero current flowing at the end of the transmission line, there is maximum voltage appearing at the end of the transmission line. As the reflected wave travels back up the cable, it adds to the magnitude of the incident wave where it arrives in-phase with it, and subtracts from the magnitude of the incident wave where it arrives out of phase with it.
Where these nodes (points of minimal energy) and anti-nodes (points of maximum energy) appear in the cable depend on the frequency. Cutting the coaxial able to such a length where the end which plugs into your SWR meter or radio is at a point where the incident and reflected waves arrive out of phase with each other will fool your meter or radio (and you) into thinking you have a well tuned antenna system. However, energy is still being reflected and it radiates back and forth along your coaxial cable until it is either radiated by the antenna, the coaxial cable, or dissipated as heat by your meter, tuner, or coaxial cable, etc. It all finds a way out of the system somehow. Damage to final output transistors can be due to large VSWR induced voltage swings arriving at the device and exceeding the operational parameters. Keep a quarter-wavelength jumper cable handy for an experiment someday, and if you think your SWR reading is artificially low, just add this into the chain. Failure can also be due to the excessive reactance presented to the output stage causing excessive current flow in the output device.
For the sake of example, the ideal standard 50 ohm transmission scenario occurs when the characteristic impedance of the RF source, the transmission line, and the load are all 50 ohms resistive. In this scenario, there is no variation along the length of the transmission line, current and voltage are in-phase, and all power delivered from the RF source to the load is transmitted. However, reactance in an RF system makes this practically impossible. Impedance is made up of a resistive element, and a reactive element, and there will certainly be a reactive element in your system which causes some energy you deliver to your antenna to be reflected, and so you will be able to see a small variation in measured VSWR in different places along the coaxial cable even in the most well matched of antenna systems.
Always measure the antenna VSWR at the antenna feedpoint. Changing the length of coaxial cable does not change the characteristics of your antenna.
Don’t get overly obsessed with SWR if your antenna is fairly well matched. A 1:1 or 1.1:1 VSWR seems to be the “holy grail” of SWR, and I really have no idea why there is such an obsession with it. Really, anything less than 2:1 is usually more than adequate. Here’s a table of VSWR and reflected power for a 100W transmitter.
|VSWR||Mismatch Loss (dB)||Reflected Power (W)||Radiated Power (W)*|
Even this table isn’t entirely accurate in the real world. This is just for one reflection. In reality, reflected power is re-reflected back down the feeder towards the antenna, and will be dissipated in the load eventually, so the real world loss is much less.
Additionally, in the event of no load (as above) or a load mismatch of extraordinary magnitude, and especially with long transmission lines, the Ferranti effect may contribute to increased voltage on the line, which may cause damage to the connected equipment, and/or the transmission line itself.
Edit 26th July 2017
I came across an article from a 1956 issue of QST which describes the misconception I outline here.