[Elecraft] APF vs 10 Hz DSP, why they don't sound the same

[David Woolley (E.L)]

Merv Schweigert wrote:

> APF does not distort the audio at all,  its extremely simple,  all it 
> does is peak
> an audio freq by X amount.  In older radios its done with a analog 
> circuit of

So how does that differ from the existing filters in the K3.  What 
people seem to be saying is that it actually boosts the audio, but that 
is just equivalent to turning up the AF gain, so I was speculating that 
the gain is actually turned up to the point where later stages overload, 
and that is why people think it differs from a simple filter.

> one or two chips and resistor capacitor sets to determine the freq of 
> the peak.
> I assume there is some feedback loop to increase the audio gain at the 
> peaked
> freq.

I.E. is a band bass FILTER!

>>   
> It does nothing as far as filtering,  years ago it was used in cheap 

If it boost one frequency with respect to another, it is, by definition, 
  filter.  This is what I mean by people claiming magical properties. 
It is just a linear filter (although possibly followed by non-linear 
output stages).

> receivers as a CW filter

> Since APF is totally an audio function the gain and AGC is not effected 
> by its use

The gain is influenced.  The audio gain is part of the overall gain of 
the receiver.  People seem to be claiming that the "peak" part of the 
name means that there is excess audio gain at the filter centre, over 
that without the filter.

The reason people introduced AGC was to point out that, being post AGC, 
the subjective effect of tuning through a signal was a much greater 
change in amplitude than pre-AGC, where the AGC would keep the amplitude 
more or less constant, although the signal to noise ratio would peak as 
you tuned through the signal.

> at all,  unless you have a radio that has audio derived AGC. 
> As I stated before APF is for very weak signal detection,  one would not 
> usually
> use it on signals that are copyable with normal filters etc.   It is 
> used for copying
> signals you normally tune across because they are too weak to copy or you
> perhaps dont even hear them.  

It's just a normal filter and if it is better than the existing filters, 
that is a matter of working out what it is about the filter shape 
(considering the overall effect of all filters in the receiver), in 
phase and frequency, that makes it subjectively better, and adjust the 
DSP filter to reproduce that characteristic.  If one understands exactly 
why it seems better, the flexibility of a DSP filter may mean that by 
designing a first principles solution, one can do even better.

> It will take a signal that is at the noise level,  and peak it to the 
> point of being
> able to copy.  

As does any narrow band pass filter.

The only difference between filtering at audio and at IF, is that 
distortion products will remain in band, and it is more likely that it 
will be done post-AGC.



-- 
David Woolley
"we do not overly restrict the subject matter on the list, and we
encourage postings on a wide range of amateur radio related topics"
List Guidelines <http://www.elecraft.com/elecraft_list_guidelines.htm>