rfic
New Member
Offline
Posts: 9
San Diego, CA
|
hi, i try to simulate the noise floor of a chain of divide-by-2s (total divide-by-32).
first i use pss+tdnoise method check the whole chain (at the final output), but i find the noise above 10MHz offset begin to roll up (i check up to 50MHz), which obviously dosn't make sense
so my question is:
1. for a chain of dividers, can i run tdnoise for whole chain? or i have to check them one by one and add them together?
2. if i check them one by one, for noise floor, it's still follow div-by-2 rule (i.e. after div-by-2, the noise is improved by 6dB)
3. i also try noise source method, the results at 10MHz is better than tdnoise method, so what's the physical meaning (explanantion) of noise source method? (what the results from noise source method means?)
thanks
|