Hi,


in the first place, you need to know what your load impedance is. In your case its not obvious because the only real resistance is coming from the transistor. So you will have to simulate it.

T-networks are just back-to-back L-match networks. So if you know how to design an L-match, its just a simple matter of deciding the intermediate impedance. There's tons of material online which you can reference and here's one I found at random:

http://rf-circuits.info/radio/impedance-matching/pi-matching/


regards,
Aaron

Hi,


if you're talking about designing a matching network for maximum power transfer, then you need to have some real part to the load impedance. Its a simple matter of only real loads dissipate power, and therefore, you can't talk about transferring power without a real load.

The MOSFET's input impedance is a capacitor with a small series resistance. The ratio of the capacitor's impedance to the small series resistance is equal to a Q value. At frequencies well below the fT of the device, this Q is very large, and its not practical to match this impedance since the matching network components are likely to have even lower Q than the load. If you look at most LNA designs, the matching works by deliberately introducing some resistance to the MOSFET's input impedance, usually by negative feedback.

For the PA, if S11 is important, then you need to introduce some real resistance to match to. A simple method is just to use a resistor. Alternatively, you can rely on the finite Q of the matching network components.

For the non-ideal components of the T network, their Q is probably high enough that you can ignore the resistive part. A little bit of iteration may be necessary, but using the smith chart will make it easy. In PA design, its important to get high Q components especially in the later stages of the PA, as they will dissipate power, and degrade your efficiency.

If you try simulating the circuits, I think you will understand this a lot easier. Also, when it comes to matching, the smith chart is your friend...


regards,
Aaron