Dear all,

Hello and happy new year 2013,

I'd like to ask a question based on pll/clock generation.

Based on this FAQ: **h****ttp://tinyurl.com/ah7r5ju** there is a benefit from running a VCO at very high frequency, higher than the Fpdf or Fref. But this leads to a high N value right?

Based on this pdf http://www.analog.com/library/analogDialogue/cd/vol33n1.pdf page 20, right column, it is stated that:

"Integer N synthesizers require a reference frequency that

is equal to the channel spacing. This can be quite low and thus

necessitates a high N. This high N produces a phase noise that

is proportionally high."

So, what i'm thinking of wrong?

Thank you.

Dear Gouzou,

> Based on this FAQ:

http://tinyurl.com/ah7r5juthere is a benefit from running a> VCO at very high frequency, higher than the Fpdf or Fref. But this leads to a high N value right?

> Based on this pdf http://www.analog.com/library/analogDialogue/cd/vol33n1.pdf page 20,

> right column, it is stated that: "Integer N synthesizers require a reference frequency that is

> equal to the channel spacing. This can be quite low and thus necessitates a high N.

> This high N produces a phase noise that is proportionally high."

What you're seeing in these articles are statements that are fundamentally correct, but they are making a lot of assumptions because they don't have the room for detailed explantions in the article.

For example, a higher VCO frequency output (with the same amplitude) will have a higher slew rate, but that only helps if your PLL noise is due to the slew rate of the VCO output. This is not always the case.

Or, this statement: "The low reference frequency means a long settling time, and the high value of N means larger phase noise." This is not always true. The settling time depends on the loop bandwidth and stability. A higher input frequency MAY allow you to have a faster phase detector frequency, and a higher phase detector frequency will allow you to have a higher loop bandwidth if you configure the loop that way. Anyhow, you can see that the statement is correct if you make the right assumptions. It's also true that a higher N can result in higher noise, but it's also assuming that you're not dominated by other sources of noise.

As with everything else, it's prudent to go back to PLL fundamentals.

I hope this helps.

-Paul Kern