Strange simulation result of NF

Dear all,

Now, I simulate NF of a receiver which is consisted single-ended input of Tx/Rx + LNA (which give out the differential output) + down-conversion mixer (IF is 1G-1.6GHz, differential output). It gives a strange NF value to me. As I simulated the NF of the Tx/Rx switch + LNA, it gives NF = 4.5dB. If I simulated included the down-conversion mixer together, the NF value would give me 14dB. It is extremely high in value. Moreover, it is un-resonable. If I plotted the NF(double-side band), it would give a 4.9dB around. Actually, what is my simulation setting problem? Or, what is my problem in simulation. Could anyone give me suggestion?

Thanks

wccheng

when you simulated your design with mixers, did you add IF port at mixer output?
if so, cancel the port and use two output pin instead.
try it.

Hi tangqin55,

Yes, I use IF port at the mixer output before. I have follow your suggestion and re-do the simulation. However, the same simulation result gives out. What is it problem?

Thanks

wccheng

u need to add some sort of filter

khouly

Hi wccheng,

At first, you could verify if your ifoutput matches the ifport because your if frequency is high. Secondly, try to see your oscillator, because your mixer depends on its amplitude (have you swept the power of the oscilator to increase the conversion gain? is it matched to the oscillator port?). Thirdly, look at your transconductance stage, if it isn't well designed it won't give you the gain you want and your NF will be very high. Try to simulate your mixer first separately from the LNA (don't forget to match input of tranconductance stage).

Good luck