[Elecraft] NB/NR Noise Blanker/ Noise reduction

[GW0ETF]

If I look at the audio response in Spectrogram while feeding in a noise
source at the antenna the trace shows definite nulls when NR is switched in.
In SSB at max width (2.7kHz) the number of nulls vary with the settings as
follows:
F1-# 9
F2-# 7
F3-# 4
F4-# 3

As # is increased from 1 to 4 at each F setting the depth of the nulls
decreases and it's almost flat at #4. However RF gain and AGC slope settings
drastically affect the depth of these nulls but at maximum the nulls are
about -30db which would presumably affect the 'loudness' of a signal if it
happened to fall in one of the holes? I don't have time at the moment to
investigate further how the AGC settings affect the nulls but backing off
the RF gain definitely flattens the response.

According to Lyle back in 2007 the F parameter indicated the number of taps
(?) in the adaptive filter and the second number the amount of signal 'bleed
through' to reduce distortion effects. I'm not sure my limited knowledge of
dsp and filters ties this in with what I see in Spectrogram.....

Having said this I must admit I tend to use filtering in preference to NR
even though I seem to be able to get NR to work pretty well.

73,

Stewart Rolfe, GW0ETF


Steve Ellington wrote:
> 
> This is amazing....We have some who complain about NR increasing the audio 
> and some complaining about it reducing audio!........
> 
> 

-- 
View this message in context: http://n2.nabble.com/NB-NR-Noise-Blanker--Noise-reduction-tp2728863p2732318.html
Sent from the Elecraft mailing list archive at Nabble.com.