On Thomas Lee's RF Book, 2nd Ed, Sec 12.3 page 375, he says:

".. both vacuum tubes and MOSFETs are devices with nominally capacitive input impedances. The key word, however is " nominally". for if the input impedances actully were purely capacitive then the input could never consume any power, and the power gain would necessarily be infinite even at infinitely high freq, sunch result defires common sense..."

Why the pure capactive input will have infinite gain? in such case S11=0dB all input power will be reflected back to source and output = 0, gain= - infinite dB. I like Lee's book and I believe he is right, but anyone can point out why I am wrong or there is something I missed?

Thanks

The amplifier need not draw power at the input port. The output signal power is the primary parameter, power gain is an artificial metric. So there is no need for the 'nominally' qualifier - a transistor (transfer resistor) can use battery energy to perform an impedance transformation from a purely imaginary input impedance to a purely real output impedance.

Also note that a source will always dissipate power in its Thevenin impedance, so even if an amplifier were purely capacitive it does not mean there is no power dissipation within the source. Most antenna impedance matching arguments are based on illumination from the far field. In the near field even purely reactive source impedances are possible. Check out RFID antennas, for example.

M.G.Rajan
www.eecalc.com