[time-nuts] timing properties of "spectrum spreading" clocks

[Bob kb8tq]

Hi

I think the gotcha is that the PLL (where one is involved) may be in a dynamic
state a good percentage of the time. Eventually it gets close to setting and then
it’s time to hop yet again. That would make the “back out” process pretty involved.

Bob

> On May 15, 2019, at 1:15 PM, jimlux <jimlux at earthlink.net> wrote:
> 
> Has anyone measured the details of the spread spectrum clocks used to help meet emission limits (like FCC Part 15)?
> 
> It seems, superficially, that they do this with some sort of deterministic process (like a linear feedback shift register), and so, one might be able to "back out" the (Pseudo)randomization to get better timing performance.
> 
> 
> https://www.maximintegrated.com/en/app-notes/index.mvp/id/3503
> seems to indicate that they use a up/down sweep approach.
> 
> https://www.maximintegrated.com/en/products/all-products/archive/MAX9492.html
> 
> https://www.idt.com/products/clocks-timing/application-specific-clocks/spread-spectrum-clocks/180-01-low-emi-clock-generator
> "IDT's proprietary blend of analog and digital PLL technology"
> 
> https://www.idt.com/document/dst/idt5p50911-2-3-4-datasheet
> 
> seems to show a triangle wave frequency modulation
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.