When I simulate a Gilbert double balanced mixer,it seems the higher the IF,the lower the gain.But I think during the mixing process,the mixer quad(in Gilbert Cell )just operates as switch and the mixing product will be independent on the value of the IF.
So does the value of the IF affect the mixing process?Or the lower gain is just due to the output match??
If you are using an LC circuit at the IF output, the maximum conversion gain of the mixer you will get close to the resonate frequency of this circuit.
Check the resonant frequency of the LC circuit, and maybe is higher frequency that you believe.
If there are no resonant circuits in the mixer, the gain should decrease with frequency, generally with 6dB/octave.
If you are using an LC circuit at the IF output, the maximum conversion gain of the mixer you will get close to the resonate frequency of this circuit.
Check the resonant frequency of the LC circuit, and maybe is higher frequency that you believe.
If there are no resonant circuits in the mixer, the gain should decrease with frequency, generally with 6dB/octave.