Re: [Discuss-gnuradio] AGC loop and digital mods

[Bob McGwier]

The feedforward AGC system in DttSP is more complex in that it is two 
track (fast and "user set").   I  believe I already mentioned this so I 
won't go over it again.  But let me reiterate one important feature of 
our work.  One of the things that was important to its success as a 
feedforward agc was to respond to a level change 3 times constants early 
in a simple predictive system.   The 3 time constants varied depending 
on the changing speed of the agc as set by the user.

Bob


'Eric Blossom' wrote:

> On Mon, Sep 25, 2006 at 11:35:33PM -0400, Tom Rondeau wrote:
>  
>
> > Eric,
> >
> > Here's a brief write-up on the tests I ran for the AGC loops. It looks like
> > the feedforward has a slight advantage over most of the test cases. 
> >
> > I thought I'd email this to the listserv in case anyone is interested in the
> > digital modulations. Basically, we're working on getting stable DBPSK and
> > DQPSK modulators into the code base. The AGC loops seems to be producing
> > problems, much of which is caused by driving the Costas loop crazy at the
> > next stage of the receiver. Future work will include creating a non-coherent
> > receiver for the differential modulations to take out the Costas loop.
> >
> > Tom
> >    
>
> Hi Tom,  thanks for making the measurements and writing them up.
>
> I'll be merging the results of all this work into the trunk tomorrow.
> Thanks to Tom for all the effort he's put into sorting out the DBPSK
> and DQPSK.  And thanks to Matt for suggesting the feed forward AGC.
>
> Eric
>  


--
Robert W. McGwier, Ph.D.
Center for Communications Research
805 Bunn Drive
Princeton, NJ 08540
(609)-924-4600
(sig required by employer)