Dear all,
I am designing a transmitter with active mixer. The supply voltage is 1.8V and the current is about 100 mA.

I use a parallel capacitor and series inductor to match the output impedance of balun to 50 Ohm at 2.35 GHz.

After impedance matching, the simulation result shows that the output power is -8 dBm at most.

However, that output power is not the maximum output power. The output power can hit 3 dBm which is pretty close to my goal. I put  Port1 at the output of matching network. In this case, the magnitude of Z11 is growing in the range of 2.2GHz to 2.4 GHz. And the value at 2.35 GHz is 90 Ohm, while it drops to 50 Ohm in the case of impedance matching.

I wonder why this happen? Does the output power maximize when the output network is matched, doesn't it?


Yutao Liu

Yutao,

Conjugate matching does insure maximum power transfer from a source to the load, but it doesn't say anything about maximizing the available power that the source could generate given the power supply voltage and signal current swing. So, it may be the case that conjugate matching provides maximum power for the source impedance, but by mismatching the the source you are able to better utilize the available voltage/current swing from the supply to generate more power.

In your case you said you have 1.8V supply and 100mA current. How much of the current is signal swing at RF (after upconversion how much RF current is swinging at the output of the mixer devices)?

From this information you can determine what load impedance on the mixer will produce the maximum signal voltage swing (taking into account the voltage headrooom needed on the active devices) for the peak upconverted RF signal swing. This maximizes the total RF power the circuit can produce (without increasing the signal and bias current). Now conjugate match the output and you should get maximum power transfer of the maximum available power as dictated by the maximum supply swing and signal current.

Hope this helps.