some questions about "Gilbert cell mixer design tutoria

[skysky]

hello everybody!
i am reading "Gilber cell mixer design tutorial" from www.rfic.co.uk. there are some questions befor me. i wish that somebody here could help me. thanks!
1.about the equation 1,2 in page 3.
where are they from? i have read Thomas H Lee's book(second edition), but i could find them.
at page 3, there is u1=u0+2*theta*Vsat*L, but in page 5,theta=0.5,u1=0.433+2*0.1*1.72e5*0.6e-6, why? i think it should be u1=0.433*2*0.5*1.72e5*0.6e-6.

2.page 4
there is "for correct operation these devices should not be driven into saturation", but the calculation formula of gm in the same page corresponds to the saturation condition. why?

3.page11
what is the turn ratio of CMP3 CMP1? how to calculat the input resistance at the rf and LO port?

4.page 14
"to achieve the required lineraity the bias current was increased from 6mA to 9mA. This of course will degrade the DSB noise figure slightly"
why the change of bias will degrade the NF?

5.page20 alternative de-generation scheme
in Lee's book, there is cross-quad mosfets to ensure that the potential at the terminals of the R equal each other. but in the figure 19 at page 20, there is no such cross-quad mosfets,i am wondering whether circuits like figure 19 could achieve fine performance? in the paper ,need for well-matched current sources is mentioned, how can i do for it?

thanks!

 

[safwatonline]

Re: some questions about "Gilbert cell mixer design tut

hello everybody!
i am reading "Gilber cell mixer design tutorial" from www.rfic.co.uk. there are some questions befor me. i wish that somebody here could help me. thanks!
1.about the equation 1,2 in page 3.
where are they from? i have read Thomas H Lee's book(second edition), but i could find them.
at page 3, there is u1=u0+2*theta*Vsat*L, but in page 5,theta=0.5,u1=0.433+2*0.1*1.72e5*0.6e-6, why? i think it should be u1=0.433*2*0.5*1.72e5*0.6e-6.

2.page 4
there is "for correct operation these devices should not be driven into saturation", but the calculation formula of gm in the same page corresponds to the saturation condition. why?

3.page11
what is the turn ratio of CMP3 CMP1? how to calculat the input resistance at the rf and LO port?

4.page 14
"to achieve the required lineraity the bias current was increased from 6mA to 9mA. This of course will degrade the DSB noise figure slightly"
why the change of bias will degrade the NF?

5.page20 alternative de-generation scheme
in Lee's book, there is cross-quad mosfets to ensure that the potential at the terminals of the R equal each other. but in the figure 19 at page 20, there is no such cross-quad mosfets,i am wondering whether circuits like figure 19 could achieve fine performance? in the paper ,need for well-matched current sources is mentioned, how can i do for it?

thanks!

2- they shoyuld be in saturation , i think what is meant in saturated output voltage i.e. linear region.
4-this is not a positive feedback but it is still negative feedback"degeneration", whee he just used one resistor instead of the two "this will remove the drop on the two resistors but need good matching for the two current sources , this is done in layout by Common centroid "