[TowerTalk] Active phased arrays.
jimlux at earthlink.net
Mon Mar 7 17:14:09 EST 2005
At 09:14 AM 3/7/2005, Tom Rauch wrote:
> > then the DSP can buy you a whole lot. I think it's safe
>to say that on low
> > bands, the dominant noise source is not receiver noise,
>but comes from
> > "outside", and the interference and signal are hardly flat
> > so there's great potential for DSP.
>Not at my QTH (or at many others).
>Noise floor here is white noise that is the result of random
>combinations of countless distant noise sources that
>propagates in from sources as far as many thousands of miles
>away. There are so many random distant sources it is
>indistinguishable from smooth hissing "receiver noise".
But, if you had a highly directional antenna, and scanned it, you'd
probably see some variation.
>Desired signals also vary in phase, level, and direction
>over very short periods of time. I can't even combine two
>antennas 1000 feet apart on 160 or 500 ft apart on 80.
You can't, but a computer might. You can't keep up with the millisecond by
millisecond changes in propagation, but the computer can. This has been
demonstrated by more than one researcher.
>I'm not disagreeing that adaptive noise canceling receiver
>systems could be useful in an environment with dominant
>local noise sources, I just think the concept of adaptive
>phasing isn't the great advantage we might think.
>Heck, I'm still waiting for a DSP radio that outperforms a
Kind of depends on what "perform" means, doesn't it? And, it depends on
whether a manufacturer is willing to design and sell it into the ham
market. Right now, a manufacturer would be a fool to do so. As this
discussion shows, not many hams see the value in a phased array, not to
mention that the manufacturer would need to face the considerable
engineering challenges in such a design. On a commercial scale it would be
a several million dollar investment, just to do the prototyping and initial
development. Would YOU be willing to pay $10,000 for such a radio? Would
5000 other hams be willing to do so? Probably not, but that's what it
would take (figuring a 5-10% investment in NRE for a new radio) to make it
Why do that when hams are perfectly willing to buy what they sell now, and
for which most of the development cost has already been recouped. Such
features as are in current DSP radios are more in the nature of bells and
whistles (nifty ones, I grant you, like adaptive tone cancellers) and
probably didn't have a very high development cost. An LMS tone canceller is
probably a few tens of thousands of dollars to develop, at most.
The best bet for this to become anything other than a curiosity is for
someone with a big pile of bucks to invest in it for the thrill of it. I'm
an optimist. It could happen. Paul Allen is investing big bucks in a
multichannel DSP radio telescope doing just the sort of processing we're
talking about here. (google for Allen Telescope Array)
More information about the TowerTalk