[Elecraft] K3 AGC Mush

David Gilbert xdavid at cis-broadband.com
Sun Dec 4 20:50:42 EST 2011


It isn't an individual rig problem, and if everyone who has noticed the 
problem sent their rig back Elecraft wouldn't get anything else done.  I 
think recreating the conditions is the relevant issue ... not variation 
from rig to rig.

It isn't difficult to figure out how close the signals need to be.  They 
need to be close enough in frequency that their sum and difference 
frequencies when mixed overlap onto the great majority of the 
originating signals.  I can fairly easily discern two keyed signals 10 
Hz apart, but put five of them within 30 Hz and the K3 turns them all 
into pudding.

The mechanics of the problem seem pretty simple to me.  The closer the 
signals are the fewer are needed to cause an overlap problem.  The 
further apart they are, the more of them are required for the mixed 
products to land on originating signals.  I'm sure someone could 
describe it mathematically, and probably already has.

Here's a point that I think a lot of people are missing, though ... the 
signals need to be pretty weak.  As someone else recently mentioned, the 
problem doesn't occur on strong signals.  If you have a high ambient 
noise level you may never even have the opportunity to notice the problem.

Dave   AB7E


On 12/4/2011 1:34 PM, Brian Alsop wrote:
> Interesting.
>
> Just a thought:
> To rule out or identify hardware problems, I wonder if someone with
> "mush" problem would be willing to loan Elecraft the offending rig?
>
> I wonder just how "exactly" on frequency guys calling from spots really
> are?  Spots are quantized in 0.1 KHz steps.  Many radios are off by 100
> Hz.  So it is easy to imagine the responders calling within a 150-200Hz
> window.  That's still a mess.
>
> 73 de Brian/K3KO
>


More information about the Elecraft mailing list