Gary,
Your use of the term "DSP" almost sounds like you're limiting its
definition to noise reduction only. There are many more applications of DSP
than just noise reduction.
73, Duane
Duane Calvin, AC5AA
Austin, Texas
www.ac5aa.com
-----Original Message-----
From: tentec-bounces@contesting.com [mailto:tentec-bounces@contesting.com]
On Behalf Of Gary Hoffman
Sent: Thursday, December 07, 2006 11:36 PM
To: Discussion of Ten-Tec Equipment
Subject: Re: [TenTec] Noise Reduction Setting
Ok Grant....
I'm not trying to split hairs with you or be blindly
argumentative....however, a DSP processor does nothing else, by the
definition of the term. If it is a general purpose processor, capable of
doing "something else" then by definition, it is not a DSP processor, which
is not capable (in general) of doing something else, and in practice, is not
used to do something else.
Now we come down to the crux of the matter. If the "DSP" processor(s) in
the Orion are used to do general purpose processing, in addition to any
"real" DSP processing, then I'd not call them actual DSP processors at all.
And then I would have to agree that there are likely nowhere near enough
processing cycles available to do the type of processing I'm talking about,
in near real time.
What I don't know is if this is in fact a correct statement of the facts.
I was told, at the Hamfest at Ten Tec where the radio was introduced and
which I attended, in person, by Doug, that there was true DSP noise
reduction. I was very impressed, and pressed the point as to whether this
was true. I was told yes. I was shown the "lab rat" radio....and the DSP
processor was pointed out to me, physically.
Now, I will admit that (at least) 3 things were possible:
1. They did not understand my question in the way I meant it.
2. I did not understand their answer in the way they meant it.
3. Something has changed since then.
Would it not be nice if Ten Tec could comment ??????????? Please ???
As far as "some kind of LMS algorithm".....I never heard of that in the
sense of true DSP. I would guess that LMS means "Least Mean Square" which
is a computation found in endless kinds of apparatus and data reduction, but
has, in general, nothing specific to do with SAW processors nor with
subtraction of uncorrelated noise data from a larger data set. I don't even
know why it would be used...unless it just refers to a general "smoothing"
of the signal, which might tend to blur out noise spikes ? I'm guessing
there. I would imagine such blurring would also smear out the desired
signal to some extent, which may be why people tend not to like it. Yet
others do !
Alas, what is the poor DSP signal analyst to do when faced with such
questions :)
Bottom line....I know the kind of signal processing I've been talking about
can be done in a very small space with small chips (but at a relatively high
price). I've seen this. I was led to believe that a low frequency version
of this was done in the Orion. The low frequency made it affordable. Now
I'm told, more or less, the opposite.
I don't know the facts well enough, and I'd sure like Ten Tec to offer an
informed view.
73 de Gary, AA2IZ
----- Original Message -----
From: "Grant Youngman" <nq5t@tx.rr.com>
To: "'Discussion of Ten-Tec Equipment'" <tentec@contesting.com>
Sent: Friday, December 08, 2006 12:01 AM
Subject: Re: [TenTec] Noise Reduction Setting
>
> > Now, on the other hand, the kind of processing I've been
> > talking about (and was led to believe is being done by Ten
> > Tec...but who knows, maybe not) does not take huge parallel
> > processors nor does it involve latency of any serious amount.
> > In the military applications for example, SAW processors
> > (Surface Acoustic Wave) handle this nicely, as do various
> > hard wired ASICS. The latter are comparatively cheap.
> > Neither do much actual processing, as the algorithm is pretty
> > much hard wired in.
>
> It's done in the DSP processors that handle all other receiver tasks. Not
> sure who might have led you to believe that there were dedicated or
> specialized NR processors. As far as I know, NR is implemented as some
form
> of LMS alogrithm, as it is in most other radios in the ham market.
>
> Grant/NQ5T
>
> _______________________________________________
> TenTec mailing list
> TenTec@contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
>
>
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|