[Elecraft] K3 Noise Reduction

[Joe Planisky]

Hi Dave,

Thanks for trying to understand.  I can't speak for others, but here's  
my issue with NR.  It's not how it works on weak signals, it's how it  
works on strong ones.

Sometimes I have AGC SLP set to a low value (i.e. this makes weak  
signals sound quieter and strong signals sound louder). I usually have  
RF gain set to max and set AF gain to a comfortable level.  I can tune  
around the band and hear weak and strong signals pass by, but I set my  
AF gain so the strongest signals are tolerably loud.  Now, if I turn  
on NR and set it to a relatively aggressive setting and then tune  
around the band, suddenly the strong signals that would have been  
merely loud with NR off are now painfully loud with NR on.  It's  
hearing-damage loud if I have headphones on (well, that's how it seems  
when you suddenly tune across a S9+20 signal.)

So OK, some say, don't tune around with NR on.  But the same effect  
makes it nearly impossible to use aggressive NR with low AGC SLP  
values on signals with deep QSB.  With NR off, a signal that fades in  
and out between, say, S2 and S9+10 is easily copyable.  With NR on, it  
goes from nearly inaudible to wake-the-neighbors loud.

Yes, there are all kinds of ways to get around it. Increase the AGC  
SLP, ride the AF or RF gain, reduce bandwidth.  All very valid work- 
arounds and I use them all at times.  But I really can't imagine that  
the extra 15 to 18 dB boost that aggressive NR gives already strong  
signals is normal, beneficial, or intended behavior.  (If it IS  
considered normal or intended, I wish Wayne or Lyle would speak up and  
I'll shut up about it. :-)  My only other experience with NR was on  
the K2, and I loved it there (although I didn't experiment with the  
settings on the K2 as much as I have on the K3.)

73
--
Joe KB8AP


On Apr 29, 2009, at 1:41 PM, David Gilbert wrote:

>
> Is most of this discussion on NR in the K3 from users who would like  
> to
> improve the SNR on louder signals for easier listening in noisy
> conditions?
>
> I get the impression we have people expecting different things from
> these controls.  The optimum AGC and NR design strategy to improve SNR
> for weak signals might be quite a different story than for stronger
> signals and noise, even if the SNR's are initially the same.  That  
> might
> show up fairly dramatically on a rig with as large a dynamic range as
> the K3.