V
v_naren
Guest
Dear members
I have designed a LNA(inductive source degeneration) at 900MHz with a Q of 36 using negative resistance circuit. I am trying to make a variable gain LNA from here on. I have cascaded the high Q cascode LNA with a source follower buffer to provide good S22 to match 50ohms. unfortunately since the Q is directly related to the load resistance across the LC tank...I have a gain of about 40dB at 900MHz with Q of 36. I need to make a gain variable from -3dB to 22 dB...any ideas on how to proceed?...I need to cascade a variable attenuator I suppose!!
but problem is that the S22 and NF degrade quite a lot and also making variable attenuator with MOSFETs seems a lot tougher due to poor linearity and related bias problems...I tried to make a gilbert cell too....but still am unable to achieve the range of attenuation need which will be from -43dB to -18dB...problem is that I am using 0.18um full CMOS tech...please help me
I have designed a LNA(inductive source degeneration) at 900MHz with a Q of 36 using negative resistance circuit. I am trying to make a variable gain LNA from here on. I have cascaded the high Q cascode LNA with a source follower buffer to provide good S22 to match 50ohms. unfortunately since the Q is directly related to the load resistance across the LC tank...I have a gain of about 40dB at 900MHz with Q of 36. I need to make a gain variable from -3dB to 22 dB...any ideas on how to proceed?...I need to cascade a variable attenuator I suppose!!
but problem is that the S22 and NF degrade quite a lot and also making variable attenuator with MOSFETs seems a lot tougher due to poor linearity and related bias problems...I tried to make a gilbert cell too....but still am unable to achieve the range of attenuation need which will be from -43dB to -18dB...problem is that I am using 0.18um full CMOS tech...please help me
