Hello,
I would like to discuss about delay time in the modeling of CP. (Listing 10 in paper jitter+phase noise)
"
...
parameter real td= 200e-12;
parameter real jitter=20e-12;
...
dt=jitter*$dist_normal(seed,0,1);
...
I(out)<+transition(Iout*state, td+dt,tt)
...
"
The vaule of the delay time, i.e. td+dt, should avoid negative. There are two metods to aovid that.
Method 1 :
Set the "td"  enough large, e.g. 500e-12 or even larger.  I would like to confirm my question: Does the delay time not influence the jitter calculation for the PLL. Because I though the delay times are same for all transition edges. Is it right???
Method 2 :
Set the delay time td= 3*jitter, and cut other values, if the td+dt is negative. Actually, the distribution is not true gaussian distribution, but the influence for jitter calculation is very small.
"...
if td+dt<0
 td+dt=0;
end
I(out)<+transition(Iout*state, td+dt,tt)
"

Which method is better from your side?  Any anser would be appreciate!!!