discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] RFC: A generic 'Loop Filter' for PLL type blocks


From: Colby Boyer
Subject: Re: [Discuss-gnuradio] RFC: A generic 'Loop Filter' for PLL type blocks (Work with Freq and Time synch)
Date: Tue, 31 May 2011 08:17:50 -0700

I've taken a look at Gardner's text earlier and he does mention 3rd order PLLs. Also, I found the text with the chapter by Harris. Good reference.

Thanks.

On Tue, May 31, 2011 at 6:16 AM, Tom Rondeau <address@hidden> wrote:
On Sat, May 28, 2011 at 2:29 AM, Colby Boyer <address@hidden> wrote:
Hi All,

Recently, I've gone down the road of timing synchronization and ended
up learning basics for PLLs and such. So far in GNURadio, it seems
that the loop filters used in the existing GNURadio synch blocks are
only first order. From what I understand, increasing the filter order
of the PLL can help it to lock in the presence of noise, for both
timing and frequency. Any else care to comment on the performance
gains for higher orders?

I propose to come up with a block that can cascade a series of
integrators and gains to form a loop filter. The total order and gains
of the filter can perhaps be defined in the block constructor. The
idea is make the block generic enough so that you can drop it in with
an outside NCO and phase error detector.

Thoughts, questions or concerns?

Colby


Colby,
You're general thought about having an implementation for these loops could work well, since we repeat them frequently in various blocks.

I will point out, though, that the PLLs are (I'm pretty sure) second order loops since they have the two accumulator stages in them. These of course look different than the types of circuits you normally see for PLLs, like in Gardner's book. But if you can get it, fred harris has a paper called "Let's assume the system is synchronized" that gives a great overview of the various synchronization problems. It also contains many block diagrams for these different systems, including a block diagram for a digital PLL. You can pretty much trace the code in these PLL blocks to his diagrams.

Tom
 
 


reply via email to

[Prev in Thread] Current Thread [Next in Thread]