The problem with using software and/or DSP Chips is that no one, except for
the military, can afford DSP chips that can operate on signals up in the
megahertz range. Thus, all DSP work in amateur rigs to date has been done
at much lower frequencies. We say that the Orion does its DSP in an IF
stage, not an audio stage, but the Orion IF stage in question operates at
about 14 kilohertz, which of course is an audio frequency (no matter that we
call it an IF stage).
What does this mean practically ? It means that there has to be at least
one stage (call it RF or IF, whichever suits your fancy) that operates ahead
of the software DSP stage. This RF/IF stage, if not well designed, may be
subject to all kinds if selectivity and overload problems. Or, there may be
a radio designed somewhere which has no filters or other stuff, just direct
conversion, but has so robust an RF/IF stage that it is not subject to these
problems. I am not personally aware of one, but that certainly does not
mean that there is not one in fact.
The Orion gets around this problem by using multiple IF conversions, with
filters in the RF/IF stages and they also use many many other excellent
design techniques in the front end of the radio, and thus they present to
the last stage an excellent signal for the final, low frequency DSP
processor to work on. Many other radios - even the Pegasus for example -
don't have the awesome front end of the Orion, thus they present a less high
quality signal to the last, DSP stage.
Radios that have no DSP stage built in can have an audio DSP system (like
the timewave models for instance) added on. But, if the radio has a lousy
front end, the timewave can't "fix" that. If the radio in question has an
excellent front end ( Corsair comes to mind), then the timewave can add
value. Mine sure does.
Bottom line - until we get "affordable" DSP chips that can operate on
signals that are themselves at 30 megahertz or so, we are always going to
have some kind of RF/IF front end that can make or break the radio, no
matter what you may do afterwards. If it is a great front end, all is well.
If it is a lousy front end, no amount of post-processing is going to fix it.
----- Original Message -----
From: "Dr. Gerald N. Johnson" <firstname.lastname@example.org>
Sent: Saturday, July 15, 2006 10:52 PM
Subject: Re: [TenTec] TenTec] SDR-1000 MDS
> On Sat, 2006-07-15 at 19:38 -0700, Lee A Crocker wrote:
> > However 11 Hz is an attainable bandwidth with the
> > SDR-1000, so -160 is a valid real figure and not just
> > some imaginary figure. I use 10, 25 and 50 Hz all the
> > time, and I can clearly tell the difference between 10
> > Hz, 25 Hz, 50 Hz, and 100 Hz. The failure of other
> > rigs to be able to attain that level of selectivity is
> > notable.
> > Bill, thanks for the analysis I totally agree with the
> > data.
> > 73 W9OY
> With the audio DSP in my FT-857D, I don't find the DSP CW narrow filter
> enhances a weak signals signal to noise. It remove QRM but doesn't bring
> the signal out of the noise. The mechanical filter at the 455 KHz IF
> does enhance the signal to noise ratio of a really weak signal (ten or
> 15 dB below the noise).
> Attaining that narrow selectivity is one thing. Having it with out
> ringing is another, and having it with out too much time delay to
> interfere with manual tuning is yet another.
> 73, Jerry, K0CQ,
> All content copyright Dr. Gerald N. Johnson, electrical engineer
> TenTec mailing list
TenTec mailing list