Posted by
Dave Van Wallaghen on
Feb 21, 2008; 12:07am
URL: http://elecraft.85.s1.nabble.com/Roofing-Filters-tp461501p461507.html
Hey Lyle,
Thanks for the reply. Man, if anyone knows about how the K? DSP works...
So if I read this right (bear with me, I'm learning), in keeping strong
signals out of its passband, it will keep the hardware AGC from activating
to eliminate "pumping". And that makes perfect sense. Keeping the high
energy out from interfering with the signal you want.
But, in the same scenario, if I had a 2.7khz filter and a 200hz filter with
no strong signals, just band noise (or maybe static crash type QRN), the DSP
filter would act the same and my desired signal would come through the same
way regardless of which roofing filter I used. In other words, if I had a
weak signal just above the noise floor in the above scenario, it would not
make a difference which roofing filter I chose, the signal would sound the
same.
Obviously, I would narrow the DSP filter to chop off some of the noise. And
that is where, in my above scenario, the elimination of extra energy from
noise would be eliminated. Reduction in bandwidth in the roofing filter
would have no redeeming effect.
And if you would permit me one more scenario. In the case of strong static
crash type QRN, in my case, either the noise blanker or noise reduction
would be able to handle the really heavy stuff. I know the hardware noise
blanker is set in front of the roofing filters and well ahead of the DSP,
therefore protecting it. I guess that statement is more of a question than a
statement.
Am I getting this clear?
73,
Dave W8FGU
>
> > But I wonder (always dangerous), if operating on some of the noisier
> bands,
> > would there be a benefit to using a narrower filter to cut down on the
> > amount of energy hitting the DSP in order to help with digging out weak
> > signals? Or is that not a consideration with this particular
> implementation
> > of DSP in the K3?
>
> In general, the DSP can handle signals and noise below about S9+40 or so
> without undesired effects. Signals between about S9+40 and S9+60 will
> start to activate the hardware AGC (to protect the analog to digital
> converter from overload), so if there are strong signals inside the
> roofing filter passband but outside the DSP IF filter passband, they can
> make their presence known by activating (or "pumping") the AGC.
>
> So, it isn't the weak signals or noise that are of concern, but the
> presence of adjacent signals that are (a) stronger than the one you are
> trying to copy and (b) above the hardware AGC threshold.
>
> Contests -- and therefore weekends -- come to mind :-)
>
> 73,
>
> Lyle KK7P
_______________________________________________
Elecraft mailing list
Post to:
[hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft
Help:
http://mailman.qth.net/subscribers.htmElecraft web page:
http://www.elecraft.com