The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Question about power supply sensitivity of LC VCO!
https://designers-guide.org/forum/YaBB.pl?num=1252141021

Message started by Yutao Liu on Sep 5^th, 2009, 1:57am

Title: Question about power supply sensitivity of LC VCO!
Post by Yutao Liu on Sep 5^th, 2009, 1:57am

Hello everyone,

I want to compare the power supply sensitivity between a LC VCO (I call it VCO1 here for convenience.) and LC VCO powered by a LDO (VCO2). One VCO1 is powered directly by 1.6V while the other is powered by LDO whose output is 1.6V as well. And I take PXF analysis to both of them, frequency range is from 1Hz to 1MHz.

The sensitivity of VCO1 is lower than that of VCO2 between the range from 1Hz to 1MHz, when I choose "absin" as the "freqaxis",. However, sensitivity of VCO1 is higher than that of VCO2 in the range from 1Hz to 1MHz offset from the oscillation frequency, when I choose "in" as the freqaxis,. Why the result change in different frequency change? Is there any problem with my simulation?

I am concerned about the low frequency supply noise rejection and I suppose the LDO would decrease the power supply sensitivity at the beginning.

Could anybody help explaining? Thanks in advance.

Regards,
Yutao

Title: Re: Question about power supply sensitivity of LC VCO!
Post by loose-electron on Sep 5^th, 2009, 11:41am

What is the model you are using for the noise on the "directly powered" version?

The LDO version has the noise model associated with it introducing noise.

You need to include a reasonable approximation of your power supply noise model in **both** simulations. One will reduce the power supply noise effects due to the LDO being there and the other will let the power supply noise be present on the oscillator directly.

At the same time, you will be introducing noise from the LDO circuit in one case and have no LDO noise in the other case.

It now becomes a quantitative analysis of what works better.

Title: Re: Question about power supply sensitivity of LC VCO!
Post by Yutao Liu on Sep 7^th, 2009, 1:24am

Thanks for the reply.

As mentioned by loose-electron, I think the cause to the problem is that I directly powered VCO1 with "vdc" in "analogLib".

Since I should include a reasonable approximation of power supply noise model in both circuits, which kind of sources should I choose to model the power supply noise when I am not able to measure it?

From the topics in this forum, I see someone uses sinusoid wave as the ripple in the supply path, while some recommend impulse. Which should I choose? Is there any criteria for choosing this model? Or is there any typical method to model power supply noise?

Thanks a lot!
Yutao

Title: Re: Question about power supply sensitivity of LC VCO!
Post by loose-electron on Sep 8^th, 2009, 8:55pm

simple: what else is connected to the power and what kind of noise does it cause to a non ideal power supply with realistic impedeces??

Its going to be unique to your design. Generally its not nice clean one frequency sine waves.

Title: Re: Question about power supply sensitivity of LC VCO!
Post by Yutao Liu on Sep 9^th, 2009, 6:21pm

loose-electron wrote on Sep 8^th, 2009, 8:55pm:

Thanks for your reply, "loose-electron".
But I don't quite understand "what kind of noise does it cause to a non ideal power supply with realistic impedance??" Could you explain it with a simple example?

And I wonder the power supply noise is present only when else circuit is connected to it. If only the VCO connects to power in my circuit, it is unnecessary to build a LDO to block the noise?

Thanks in advance

Best regards
Yutao

Title: Re: Question about power supply sensitivity of LC VCO!
Post by loose-electron on Sep 11^th, 2009, 4:53pm

There is no such thing in the real world as an ideal voltage source. They all have some impedance.

With that "real world" voltage source, you are connecting loads to it. The currents in those loads (generally transient in nature) cause variation in the impedance in the power supply and consequent noise