Doublet Antennas and Balun's
My friend John (GI0HWO) and I carried out a lot of research and testing on feeding a doublet antenna and in the process learnt a lot about balun's, mainly due to a great article by Tom W8JI.
The document below was written by W8JI and both John and I can confirm that what Tom suggests is 100% correct and as a result our doublets' are now performing better than ever, so with credit to Tom we re-produce an extract from his article below:
I would just add that this article blows out of the water all the common thoughts on baluns.
Like most people I thought that a 4:1 was the way to do it.
A lot of people think that what Terry & I are doing is not far short of herecy but the proof of the pudding as they say. Try it, but you need a good balun (or a big pile of inferior ones HI HI)
John - GI0HWO
"Multi-band Dipoles (Doublet's)
A Multi-band Dipole antenna is cut for a specific frequency, but plans call for it to be used on any of the frequencies above that..
The correct construction technique is based on making the antenna element a half-wavelength on the lowest frequency on which the antenna will be used based on the 468/F formula. The parallel conductor, air-dielectric feedline is led away from the antenna at right angles.
Lower impedance feedlines result in less extreme impedance changes from band-to-band. For instance, 600-ohm feedlines tend to present wider load impedance excursions to the tuner in multi-band applications than 300-ohm feedlines.
Coax has a still lower impedance, but the non-air dielectric causes great loss within the feedline with high SWR. For instance, at 10:1 SWR you lose half of your power for each 100 feet of coax, parallel feedlines with air dielectric have much less loss at the same SWR.
Why you don't want to use coax when the SWR is high?
For each 100 feet of coax, you lose half your power at an SWR of 10:1 and at frequencies higher than 14MHz, it's worse. For higher loss coax like RG-58 or RG-8X it is worse, too.
Feedline length is critical to antenna performance. Always choose a feedline (connects the antenna to the balun, in this instance) that is 1/8th wavelength or some odd-multiple of 1/8th wavelength long on the lowest band.
Often, a 4:1 balun is suggested for Multi-band Dipoles; however, the best balun to use for this application is a 1:1 ratio. The impedance at the end of the feedline is going to vary greatly from very high to very low. Tuners have an easier time with high impedance than a low one. A balun with a ratio of 4:1 or more will transform already low impedance to an even lower one that will make the antenna hard to tune. The 1:1 ratio balun will just pass the low impedance through.
The parallel conductor feedline should be led to the balun located at a convenient location. The balun should be located such that the coax between it and the tuner is as short as possible BUT do not route the parallel line so close to the tuner or the rest of the station equipment that RF feedback occurs. This will manifest itself by making the tuning of the antenna very difficult and the controls will be very “touchy”. There may also be RF present on the microphone, key, etc.
Even when properly done, this arrangement will subject the coaxial line between the tuner and balun to very high standing waves and high voltage and/or current. You should use good low-loss coaxial line and keep the coaxial line length as short as possible, RG8X and smaller will not do a proper job. Belden RG-213 or equivalent is the minimum."
As Tom states there will be 'high voltages and /or currents' on the feeder so please make sure that you select a high power balun, John and I both use the excellent DX engineering 10 KW baluns, at full UK legal power limit these baluns dont even get warm!
A Rig with Maximum Soul: The 5 Band Transceiver of Glenn KU4NO - That's the front panel. Glenn's kids and grand kids liked sitting on his lap while he played radio and putting stickers on it This is one of the best HB2H...
2 days ago