I'm doing a small board with a 915 MHz transceiver. A small helical antenna is used, which is located on board with a very short path (10 mm) from the IC, including balun and a matching network. There are no RF connectors or real state available to include any. Since I'm constrained with the board space and the board needs a plastic case, I couldn't follow the manufacturer's guidelines for them so I need to at least do some antenna matching to ensure maximum transfer. The problem is that I don't have access to a network analyzer, so I'd like to know if there is some manual method I could use, besides randomly trying L and C values and measure the output. I don't need serious antenna performance, but some reasonably "good" one (a relative term, I know).
The board has footprints for two 0402 components, one series and one parallel. IC's crude power output is around 20 dBm. From the IC's development kit I've got boards that have a good antenna and good matching for it (other kind of antenna). I was thinking I could do some continous transmission from my board and get a relative power measurement from the dev kit receiver's perceived RSSI. I could compare the results with measures obtained from another dev kit board in countinuous transmission. Then I would perform some brute force by trying out L and C values. I could also choose from 10 different channels in the 915 MHz band (902~928) and see what diference I get. I know that's terrible, but hey, that's why it's called brute force.
Is there a better method? Is there any cheap (<$100) equipment I could use to get some measurements? Please mind that I live in an underdeveloped country so "buy some xxx on eBay" is not within my possibilities.
Any suggestions would be greatly appreciated.