rfmems
|
In a sigma delta fractional N pll, the fractional spurs are shaped by the SD modulator which controls the multi-modulus divider. In the literatures I found, the first order approximation of quantization noise always neglect the effect of the number of bits of MMD and assume the quantization noise white.
My questions are: (1) How does the bit-number of MMD effect the quantization noise and/or frational spur? (2) How can the fractional spurs be treated as noise or even white noise as in numerous literatures? (3) When the PLL has a fixed fractional division factor N+x, the quantization error should be x and 1-x with different probability density. How can this be treated same as SD ADC and get a rms value of 1/12?
Thanks in advance. chenyan
|