It depends on what 'RF' is with respect to your transistor's f_max. In most RF amplifiers the transistor reactances dominate the impedance matching, plus the matching network is usually narrowband with a Q < 10. So the load line won't be a pure resistance - it will swing around on the Smith chart.

Is this amplifier inside an IC, or is it external? If external you will be better off with a 25 cent part from one of the major vendors.

M.G.Rajan

I got this equation out of a book by Schubert and Kim.

Q = [(mdc/mac)*Vcc+C]/[1+mdc/mac]

mdc = slope of DC load line
mac =  slope of AC load line
C =  saturation voltage of the transistor

mdc = 1/[Rc +Re(beta+1)/beta]
mac = 1/Rc

Q then will give the 'best' Vce for low frequency approximations

If you don't want to do the math, a good rule of thumb says you should drop about a third in voltage between Re, Vce, and Rc each.

A couple of pointers on this application. You will probably need a high DC current in the transistor. This will help with linearity at higher powers (800mW = 29 dBm) which is a lot of power. I personally have not designed any power amps, which is what you are trying to do. Also, this amplifier configuration has a high output impedance compared to your low antenna impedance. You will need some sort of tuning network to match the amplifier to the antenna, otherwise you will have a horrible VSWR problem.