love_analog
|
Sorry if this is a dump question a) According to LTV theory, you just take the fourier transform of ISF and multiply with the various noise sources (gm stages and bias current) to get the final phase noise. The noise at wosc gets multiplied by c1 There is no talk about AM and PM from noise values at different frequencies b) According to mixer theory, Low-frequency noise (thermal) mixes up to create two correlated sidebands resulting in only amplitude modulation (AM) noise. High-frequency bias noise (which is only thermal) downconverts into a single noise sideband in the passband of the LC tank, containing both AM and PM noise.
Questions... a) So if there is no AM-PM conversion than mixer theory predicts lower phase noise than ISF b) In addition mixer theory seems to indicate that noise at wosc gets mixed to dc and 2*wosc so it won't impact final phase noise either. c) Mixer theory claims Low freq flicker noise gets upconverted to 3wosc, 5wosc etc. which flows through the cap and causes phase modulation. It also gets converted via AM-PM since any variation in current changes the swing of the VCO.
|