[Elecraft] Starting values for AGC with MCU 4.51 ???

[Guy Olinger K2AV]

I moved my AGC values around some to check it out, and then finally put
them back where they were, because I STILL like them there.  My intercept
value was worked out when I first got my K3, which was not at the extreme.
 4.51 has not changed the AGC behavior if your previously picked AGC was
within range to start with.  For some they were at the extreme of the old
value limits and wanted more numbers.  Those folks now have something to
play with.

What I got out of 4.51 is a remarkable change in clarity and a coincident
improvement in DSP NB (no change in IF NB).  This has been explained as
using numbers with more places after the decimal in logarithm functions,
which would have removed a kind of "distortion" at AGC threshold.  This
"distortion" had already been reported as being eliminated by moving the
threshold higher, or by adding attenuation to the RF, which effectively
raises the threshold.

Many of us for other reasons had made these changes ourselves some time
ago, and simply were not affected by the complaints from others, making
this problem perversely one of those A hears it, B does not, kind of things

I note that for a long time,  it was being blamed on the hardware AGC.  If
there WAS something still left there, I don't know what it would be.  Maybe
all of us need to apologize to the hardware AGC genie in our K3.  False
accusations may have cause hard feelings and reduced mojo.

At some point after I got the K3, I quit using fast AGC at all, and set the
slow AGC to its fastest value, used that all the time  just because I
"liked it better" (lack of technical specificity noted :>)).  Also I "like"
the fast AGC again (lack of technical specificity still noted).  I'm sure
it's obvious to all, the slow AGC punches high, and takes its fat time
coming back down, which puts the majority of signals under what is a
rolling threshold where only a yet stronger signal would push it higher.

So the combination of slow AGC and managing the RF input had the effect of
minimizing transaction in the area fixed by the log function improvements,
and the statement "Guy just is not having this problem" was both true AND
covering up the problem.   It also explains why I could never come up with
anything to ID problems reported on the reflector using steady state
signals  :>)   If ever there was one, this was one of those
"self-disguising" problems that confounds solution.

The clarity improvements in 4.51 occur for me because, even with my coping
adjustments to AGC and RF attenuation, the **slice of time where the AGC
was moving from not invoked to set-at-value-by-signal** produced a
distortion on the signal that happens only on the edge of signals.  This
also explains why the NB DSP t1-7 and t2-7 are better than before on key
clicks (turn off IF blanking).

Way to go, guys.   This one's really a winner.

73, Guy.

On Tue, May 22, 2012 at 5:48 PM, Peter Chamalian <w1rm at arrl.net> wrote:

> I have not had a sufficiently large or loud pile up to test the AGC mods
> under "battle conditions".  Perhaps this weekend...
>
> I have my AGC values all set at the default levels except AGC THR which is
> 13.
>
> I'll be interested to hear what others are doing.
>
> Pete, W1RM
>
>
> -----Original Message-----
> From: Jack Colson [mailto:jcolson7 at tampabay.rr.com]
> Sent: Tuesday, May 22, 2012 5:19 PM
> To: elecraft at mailman.qth.net
> Subject: [Elecraft] Starting values for AGC with MCU 4.51 ???
>
> With all the comments about the upgraded or improved AGC is their a
> consensus as to starting values?  I realize that these can be tailored for
> specific situations but a default starting set would be helpful.
>
> Thank you in advance,
> 73,
> Jack, W3TMZ
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>