The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> spreading spectrum pll
https://designers-guide.org/forum/YaBB.pl?num=1074267312

Message started by Dze on Jan 16th, 2004, 7:35am

Title: spreading spectrum pll
Post by Dze on Jan 16th, 2004, 7:35am

Hello,

I would like to use the spreading spectrum technique on a pll. I have access to its VCO input and of course to the reference clock (crystal).
What should I put between the reference clock and the VCO input in order to spread the spectrum?
A simple capacitor did not manage it.

Thanks a lot,
Dze

Title: Re: spreading spectrum pll
Post by Eugene on Jan 20th, 2004, 11:27am

Are you asking about a fractional-n synthesizer? Here, the divider ratio dithers about the desired ratio, which no longer must be an integer. The dither pattern is designed to push the phase noise out to high frequencies where the loop filter can remove it. There are lots of papers and books on the subject. Just search for "Fractional N Synthesizers".


Title: Re: spreading spectrum pll
Post by Dze on Jan 21st, 2004, 1:21am

It's not about dividing frequencies.
I have a clock (HF) in my microcontroller that is obtained from a pll that multiplies a reference crystal clock frequency.
I would like to use the so called "spectrum spreading" in order to modulate the clock (HF) with the low frequency, reference clock (from the crystal). The result should be that when I look at the spectrum of the HF clock (output of the pll), there would not be narrow spike at the HF, but a larger band with a smaller dB value.
I tried a low cost solution, by just inserting a capacitor but it creates only 2 additional smaller spikes in the spectrum around the HF one, while the HF is unchaged. And it creates instability in locking the pll. So I need another solution, low cost, if it exists. ???
Dze

Title: Re: spreading spectrum pll
Post by Jitter Man on Jan 21st, 2004, 9:52am

I do not understand why you want to spread the spectrum of a microcontroller clock (EMI perhaps), but you can probably accomplish it by injecting noise either before or after the loop filter. The noise could be in the form of a digitally generated dither sequence, or perhaps it could be amplified device noise, perhaps avalanche noise from a zener diode. You could filter it to get the desired spreading spectrum.

[glb]Jitter Man[/glb]

Title: Re: spreading spectrum pll
Post by Dze on Jan 22nd, 2004, 6:57am

Thanks for your answer. You're right. It's indeed for EMI purpose.
It makes sense to try with some filtered noise. If you have already tried it, did it create any pll instability? I guess it will be a trade off on the coupling capacitor value.
I'll let you know if it worked.
Dze

The Designer's Guide Community Forum » Powered by YaBB 2.2.2!
YaBB © 2000-2008. All Rights Reserved.