The effect of the amplitude differences on RF power combining

[tptambusai]

I need to combine signal power in RF from each element of the 8 element array antenna using power combiner, the 8 signals are in phase but difference in amplitude, the variation of the amplitude difference is up to 60%, would you like to explain what is the effect of the amplitude differences to the combining process and to the output power? is the rule is still pure addition like p1+p2+..Pn? or there will be a loss?
Iam actually a new comer in the RF field, so sorry if my question is rather dummy.. thank you so much if you can explain as detail as possible

 

[FvM]

RF Power combiner

Usually, the power distribution/combining network of an array antenna has the purpose to achieve specific weighting factors for each antenna. Only with these factors, the intended directivity can be realized. Simply connecting the antennas by symmetric power combiners would destroy the designed directional characteristic.

To answer the specific question, for losless combination of signals of (known) different amplitude, you have to design a special matching network.

 

[tptambusai]

Re: RF Power combiner

thanks FvM,
actually the amplitude difference is random, sometime the p1>p2, some time p2>p1, some time p3>p1, or sometime p2>p3, and so on.., my next question is: if we use the 1 to 8 power splitter/combiner (i want to combine 8 ramdomly amplitude input , but they are exacly in phase), what is the effect of this ramdom amplitude variation? how much the loss will be?

 

[FvM]

Re: RF Power combiner

If I understand right, (all ?) symmetrical power dividers have this transmission factor:

1/√2*(A1+A2)

You can calculate the loss for a specific amplitude ratio. It varies between 0 (equal amplitude) and -3 dB (only one port stimulated). Unless you use a Wilkinson power divider, unsymmetrical components will be back-scattered to the antennas.

 

[tptambusai]

Re: RF Power combiner

thank you FvM,
it is to calculate 2 to 1, using the ratio of p1 and p2..
another question is if the input is 8 or 16, how to calculate the loss?

thankyou