Our original set-up used a Tepco product--http://www.rapidnet.com/~tepco/ I don't know if they still make boosters. This was one of the receiver modules out of their translator line, with its own power supply, and an output of the 10.7 IF, that fed the separate transmitter unit in another box. Believe the power came from the receiver through the center conductor of the coax. We put the receiver on one side of a hill--towards the main station--and buried about 1000' of cable over to the transmitter on the other side of the hill--towards the town.
(1200 ft amsl)
<to station .............................
....... .............
<<<<< ..... .............
[] ... .......... >>>>> to town in valley below
[] ... .......... [] at 600~640' amsl
receiver... (receiver and transmitter 1000 ft apart []
93.9 93.9 transmitter
As you can see by this set-up (if all the dots translate right) the problem is keeping transmit RF out of the receiver. Becomes a limitation on how much power you can put into the booster.
Our present set-up is much simpler. We built the booster in 1988, but abandoned it about ten years later when the long cable died. Then we built a Class A on the same farm above the town. Now we have two STL paths from our main studio, one carrying audio for the Class A, the other, the composite audio for the booster for our other station. All we use for the booster, equipment-wise, is a 30 watt BEXT (RVR) exciter, and a one-bay SWR antenna on 20 feet of 12" tower out behind a barn. The exciter was $1k used, the antenna $500.
Of course you will need STL, antennas, etc., to feed the booster; we got a 606 used for this purpose & our Omnia has two composite outputs--making it easy to feed both STL and main transmitter.
Before embarking on this path I would look at two other options:
1. Ditch the Phelps Dodge two-bay for an ERI antenna. Yes, you can do all kinds of modeling, etc., but I don't think you can get the $$$ from ownership to do that. Just swap out antennas. I'll bet you notice some improvement.
2. If you still want to look at a booster have a consulting engineer run a study. He will be able to give you the idea of how much power you can run on the booster, where the interference zones will be, etc. I've been quoted about 2500 for such a study