PLL phase noise question

[super]

Dear all:
in PLL , in band phase noise is degrade by VCO or Crystal ??

thanks.

 

[biff44]

If by "In band" you mean carrier sideband noise that is closer to the RF carrier than the open loop bandwidth of the PLL control loop, then it is mostly due to the crystal.

The PLL will try as hard as possible to match the RF oscillator's phase, after dividing it down, to the crystal oscillator's phase. So if the crystal oscillator phase moved 1 degree, the RF oscillator phase moves N x 1 degree, where N is the divisor in the PLL.

If the RF Oscillator moves in phase due to some random noise source, the PLL control loop quickly sends a correcting tuning voltage to the RF Oscillator to move it back to where it was. The only problem with this is that there is a limited amount of open loop gain available to correct the RF Oscillator noise, so you can not take a horribly noisey oscillator and expect the PLL to completely clean up the noise by simply locking it to a pure crystal oscillator.

 

[melc]

If by "In band" you mean carrier sideband noise that is closer to the RF carrier than the open loop bandwidth of the PLL control loop, then it is mostly due to the crystal.

The PLL will try as hard as possible to match the RF oscillator's phase, after dividing it down, to the crystal oscillator's phase. So if the crystal oscillator phase moved 1 degree, the RF oscillator phase moves N x 1 degree, where N is the divisor in the PLL.

So, what's happening if the PLL is a fractional one ?

 

[ashutosh_g]

if i understand correctly than by inband you mean the noise inside the loop bandwith
of loop filter ..right ?
since VCO and its phase noise is always there at the output this means that PN due to VCO will be there however if the loop is closed you can actually get bettter performace than specified by the VCO vendor.

now as far as the noise inside the band contributed by the crystal osc. is concerend
that wont be affected at all beacuse PLL is actually a narrow band multliplier
so any noise within the b/w of the loop filter will also be simply multiplied by factor N

HTH
ashu