Hiya Carl
Hmm, lots of factors here. OK, as regards the height, unless there were any obstructions near by I would say that the height at which you have your dipole mounted is probably OK. However, once a dipole is moved above about one half wave above the ground (assuming it is a perfect conductor), you tend to get multiple lobes in the radiation pattern. Also, if you get the height 'wrong' you can end up firing a lot of your precious RF energy straight up. Even at one wavelength height (the height you mention above), your antenna will be producing two main lobes, though depending upon the radiation angle they may help or hinder you. Here's an illustration below of dipole height versus radiation pattern:
You can read the full article here:
http://www.astrosurf.com/luxorion/qsl-antennas-basics3.htm
I would say most definitely use a balun, since there is nothing worse than feeding a balanced antenna with an unbalanced feeder. Guaranteed to mess up the radiation pattern for sure, and in so doing waste precious RF energy, and mess up the directivity of the antenna system!
Again, with the feeder, most certainly use 75 Ohm feeder, or perhaps use a low ratio balun at the feed-point and still keep your 50 Ohm feed. You would need about 1.2:1 turns ratio to give about 1.44:1 impedance ratio (close enough to the required 1.5:1 ratio to transform 50 Ohms into 75 Ohm, though some sources suggest 76 Ohms for halfwave dipoles).
Another way would be a 75 Ohm 1:1 balun at the feed-point, 75 Ohm low loss feeder, and then EITHER an antenna coupler to match your 50 Ohm transceiver to the 75 Ohm feeder, OR, use a 1.44:1 Un-Un (unbalanced to unbalanced) transformer.
The Un-Un could take the form of a 12 turn inductor, with the 'cold' end of the coil being common to both the input and output, the 'hot' end (full twelve turns) going to the 75 Ohm unbalanced coaxial feeder, and a tap two turns down from the 'hot' end (i.e. 10 turns) going to the 50 Ohm side of the system (i.e. to the transceiver). The tap two turns down from the 'hot' end will give a turns ratio of 1.2:1, and so an impedance ratio of 1.44:1 (close enough for what we want). A small toroid or maybe even a short length of ferrite rod could be used as a suitable core.
As mentioned elsewhere on this site, keeping the coaxial feeder at multiples of a halfwave length (allowing for the velocity factor of the cable, rule of thumb is 66% for most types of coax), should ensure that you obtain a good match across a wide area of the ten metre band. Worst case scenario is having a feeder that is an odd multiple of a quarter-wave length, so you end up with a phase shift on the coax that makes your 75 Ohm feedpoint impedance look more like 800-900 Ohms. However, odd multiples of a quarter-wave length are sometimes used in some antenna designs to actually help in the matching process (i.e. making a highish feed-point impedance 'appear' lower).
Long live QRP!
Best 73, Mark.
