Question About the PM of a PLL

[LuncentWish]

I have designed a PLL. I use CP-PLL structure, loop-filter is 3rd-order loop-filter. Then I use Matlab to plot the Bode figure with the transfer function.

I am confused because at 0Hz the phase is -180°. Is PLL's Hopen(s) supposed to have such a phenomenon?

 

[BigBoss]

It's an unusual PLL Loop Filter characteristic. There must be somewhere errors.

 

[LuncentWish]

Thanks for your reply. I would like to know what the typical PLL's phase Bode figure looks like, referring to the Hopen(s).
At 0Hz, it should start from 0° and drop gradually towards -180°？

 

[KlausST]

Hi,

I´d say it depends on the circuit and how you measured it.

There are filters with "inverting characteristic" (at DC) .. then they usually go into the non inverting (instead of the inverting) input of the error amplifier.
And a 3rd order filter may even go towards 270° phase shift. (not only 180°)

So show your complete setup. The better your information the better our answers.

Klaus

 

[LuncentWish]

Hi,

I´d say it depends on the circuit and how you measured it.

There are filters with "inverting characteristic" (at DC) .. then they usually go into the non inverting (instead of the inverting) input of the error amplifier.
And a 3rd order filter may even go towards 270° phase shift. (not only 180°)

So show your complete setup. The better your information the better our answers.

Klaus

Hi Klaus,
Thanks for your reply, I have the following loop filter:

It's a pure passive loop-filter, the values are as indicated. The left node is terminated by a Charge-Pump(open-loop CP) circuit, where nominal Icp=17uA; the right node is terminated by a VCO with KVCO=0.8GHz/V.
This PLL's post-layout simulations at 5 different corners with large variations of temperature are good. But the Bode figure's phase characteristic looks strange to me, and I don't know how to evaluate that Bode figure.
Is the above information sufficient?

Lucent

 

[KlausST]

Hi,

I asked for the complete setup. You showed just the filter components.

So I have to assume you just used the given circuit - nothing more.
And you state that the analysis gives -180° for the DC phase between nodes "CP" and "VCO"

I´d say this is impossible.

Klaus

 

[LuncentWish]

Hi,

I asked for the complete setup. You showed just the filter components.

So I have to assume you just used the given circuit - nothing more.
And you state that the analysis gives -180° for the DC phase between nodes "CP" and "VCO"

I´d say this is impossible.

Klaus

Hi,
Sorry for my vague statement. The given circuit is LP(loop filter), and the complete setup is as follows:

This is a typical PFD&CP-based PLL(No digital trimming in this circuit), in order to calculate the Hopen(s) of the above system, the following parameters are all you need: Kcp=Icp/2/pi, with Icp=17uA; resistor and capacitor values off Loop Filter; KVCO=0.8GHz/V; N=4. The Z(s) of the loop filter will be a long equation.

Lucent