On Sat, 2006-03-25 at 09:58 -0500, Bill Tippett wrote:
W9OY wrote:
>
> >I have been playing with a SDR-1k recently. The
> filter scheme in that radio allows down to a 10hz
> bandwidth. What I have discovered is there is no
> ringing in a filter this narrow generated by the
> algorithm used in the SDR, even if the signal being
> received has a bandwidth considerably wider than 10
> hz. (up to 200hz for example). Precise tuning of the
> received signal is a requirement.
>
> Lee I have no doubt you are reporting
> exactly what you observed, but doesn't this seem to
> violate the laws of physics? (i.e. receiving a 200
> Hz signal through a 10 Hz filter). Have you ever
> measured the actual BW of the "10 Hz" filter? If
> not, I suspect there may be some optimism in the
> setting. Many manufacturers including Ten-Tec
> do this. For example Orion's "100 Hz" 6 dB BW
> is actually around 150 Hz. Another issue is
> the shape factor of the filter. If a "10 Hz"
> filter had an extremely wide shape factor, it
> would actually pass energy much wider than the
> stated 6 dB BW. If I recall, Orion's "100 Hz"
> setting has a shape factor around 3:1. This
> coupled with the actual 150 Hz measured BW
> accounts for the fact that Orion never rings
> even at its lowest DSP BW settings.
>
> If the SDR-1000 folks have discovered
> some way around Nyquist, Shannon and Hartley,
> I would love to know how they did it.
>
> 73, Bill W4ZV
>
>
> A filter need not have square corners, nor a ringing response. A Butterworth
> filter
will take at least 5 times the speed of a square cornered filter and a Bessel
will do
better than that. The DSP filters in common use do far better than similar
analog
filters because the DSP filters have far higher orders allowing a better
approximation
of the amplitude response desired along with a better time response. This can be
accomplished in analog filters too, but hardly ever is because of the
significantly
greater filter complexity.
--
73, Jerry, K0CQ,
All content copyright Dr. Gerald N. Johnson, electrical engineer
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|