Gain and Noise Figure discrepancy during characterization of RF front end.
In my front end package operating at 14.5 GHz, we observe that the gain and noise figure is not at our design specs.( measured NF~5.6dB and gain ~20dB). Here three LNAs are cascaded together, in order to bring NF and gain of the cascaded system back to the specs, we have tuned the second and the third amplifier stages using ribbon, the result is that gain has increased to the required level (~30db) but the noise figure remains same like that of 5.6dB but our required spec is 3.5dB respectively, my following doubts in my approach are as:
1. Is it that since the gain has increased to our required level, NF will be decreased simultaneously! , kindly guide me.
2. By placing the ribbons in third and second amplifier stages , is we are on the way of conjugate matching? then how NF will be improved?
waiting for the reply.
The noise figure is primarily determined by the first amplifier in your cascade. Consider Friis formula as first hint.
In cascade chained systems, first stage defines NF dominantly(look at Friis formulae about cascaded NF calculation) that's why your NF appears 5.6dB that is the NF of first stage.
All you should do, optimize your first stage for wanted ( or less ) NF by using "Minimum Noise Figure Matching" principal ( Gain may be less than desired, NF is more important ) and then make the other stages with higher gain by conjugate matching.It wil not increase NF because NF has already been defined by 1st stage..
What is the technical difference between minimum noise matching and conjugate matching? is it gamma optimum and s11* are different?
A conjugate match does not mean you will achieve a minimum noise figure. There is always a trade-off between minimum noise figure and conjugate match!