Ah, my kind of cheapskate. Wanna build an antenna instead?
Omnidirectional antennas suck. What you want is a sector antenna. Lots of gain, cheap, fair vertical radiation pattern, cheap, easy to build, cheap, and fairly good bandwidth. Oh yeah, they're cheap and easy to build. 5 sections (110 degrees horizontal beamwidth) yields a theoretical 13.6dBi gain.
I'm practicing being subtle. I don't think it's working.
Omni antennas are incredibly difficult to build. That's because there's nothing to support the driven elements. More specifically, everything above the connector is part of the driven elements. There's also nothing worse than building an omni antenna, testing it perfectly on the bench, stuffing it into a plastic or fiberglass tube, and finding that everything has moved down in frequency.
For example, here's an allegedly "easy" to build omni:
That's not my idea of easy (or a good antenna as the coax cable parts have fairly high losses).
Here's a 5dBi vertical colinear kit, which is also not my idea of easy to build:
Antennas double in size for every 3db gain, so a 12dBi will be about
4-6 times larger (and more complex).
On the other foot, the AMOS/Franklin sector antennas have the back reflector to support the driven elements. Much easier to build, less critical, fewer parts, and cheaper. Using a plastic rain gutter as a radome has resulted in only a tiny frequency shift. Also, sector antennas are much better for most applications as they can be tilted downward, while omnis have an inherent uptilt.
My 80+ year old (at the time) piano teacher used to say "after a certain age, it's not worth the effort resisting temptation". I think I was about 12 years old at the time, and never forgot the line as a great excuse to do what I damn well please.