[NLRS] Give away + Need advice on DirecTV dishes

Sat Jun 8 12:05:00 EDT 2013

It certainly is practical to enclose the Vivaldi taper in a horn, I saw 
one on line just with top and bottom triangular plates. I'm not sure 
which came first, the planar Vivaldi or the double ridged horn with 
exponential curved ridges. The technique certainly makes a horn with a 
very broad band match in broad band waveguide with an easy transition to 
coax. I wonder about the location of the phase center which ought to be 
at the focal point of the dish being fed and I'm pretty sure the gain is 
a function of the aperture in wavelengths so at low frequencies the 
spill would be great and at high frequencies the gain would be down 
because the dish is underfed. Somewhere the gain is optimum and that is 
selectable by the horn aperture. Usually a horn does not give equal 
E-plane and H-plane beamwidths for the same aperture so often feed horns 
have different E and H plane dimensions to equalize the patterns.

It is possible to make the Vivaldi without dielectric but at least one 
of the references says the pattern is cleaner with the dielectric than 
without it or with it very thin. One of the references also show that 
the outside shape of the metal affects sidelobes and the overall pattern 
and broadband performance, impedance and radiation. As does adding 
places or a horn. As the 54 references to the TechRpt135 show there is 
much room for variation and innovation on the topic.

Somehow in Paul Wade's dual band 10/24 feed, the corrugated horn that 
sets the aperture for 10 has little effect on 24 where the dual mode 
sections seem to set the beamwidth so the feed beamwidths are comparable 
on the two bands. Perhaps its the deep corrugations that do that, though 
its a variation on a design with a wide flare without corrugations on 10 
that gave practical results.

Long ago Jodrell Bank used concentric coaxial waveguide feeds to get the 
phase centers in the same spot, though that worked best for their short 
focal length dish where the small achievable aperture in wavelengths of 
the coaxial guide gave a broad feed pattern. The cutoff frequency of the 
lowest frequency mode in coax is where the average circumference (inner 
and outer) is 1/2 wavelength or one wavelength. So the smallest aperture 
is smaller than wavelength divided by pi.

73, Jerry, K0CQ

On 6/8/2013 8:55 AM, Marciniak, Ed wrote:
> How about a double ridged horn with a not necessarily square cross section (or equal tapers)?
>
>
>
> ----- Original Message -----
> From: Dr. Gerald N. Johnson [mailto:geraldj at netins.net]
> Sent: Friday, June 07, 2013 09:41 PM
> To: nlrs at mailman.qth.net<nlrs at mailman.qth.net>
> Subject: Re: [NLRS] Give away + Need advice on DirecTV dishes
>
>
>
> I've done some more reading, thinking one problem with the Vivaldi is
> that the phase center is going to change with frequency throwing it out
> of focus.
>
> Some say that is absolutely true, some say not. Some say the Vivaldi can
> be designed for a clean pattern with at least 10 dBi broadband gain,
> some average 6 dBi and peak at 8 dBi. From my reading I concluded that
> for every one who seriously designs a Vivaldi (and there are major
> variations with different aperture tapers, linear, circular, and others
> described only by their creator's name) there are at least a dozen
> different designs with different characteristic.
>
> www.cresis.ku.edu/sites/default/files/TechRpt135.pdf seems to put the
> design emphasis on compactness much smaller than most designs but
> includes a considerable survey of the literature and discussion of
> design tradeoffs.
>
> www.ieee.cz/mtt/soutez07/Nevrly.pdf put emphasis on broad band with a
> frequency insensitive phase center so the UWB transmitted and received
> pulse is not broadened unreasonably. So their antenna and feeds were
> experimentally (much with analytical software) optimized for a
> compromise on impedance match giving priority to pulse quality.
>
> It is possible to design for higher and lower gains depending primarily
> on the aperture height and the length of the taper which is exponential
> if its a true Vivaldi. But Neverly shows at least 8 different
> exponential curves and their effect on pulse quality and impedance match.
>
> It begins to boil down to simply assemble Vivaldi and dish and have Kent
> test it on multiple bands at CSVHF in 7 weeks or Donn in a week, perhaps
> build it with the capability of easily optimizing the feed to see how
> much that particular Vivaldi changes with frequency so a compromise
> position (if compromise is needed) to favor the band most in need of
> gain. I think it will show gain, but a bit off what the DirectTV dish
> would do with the optimum feed on all your bands of interest. Then the
> question is having one antenna with less gain better than hauling 4 or
> more, can enough of the gain imperfection be made up by site location,
> antenna elevation above ground, TX power, and receiver preamp?
>
> 73, Jerry, K0CQ
>
> On 6/7/2013 8:55 AM, David Palm wrote:
>>
>>
>> Jerry wrote:
>>
>> The Vivaldi that Kent sells is 10-25 GHz, not very useful on 3.4 GHz. His
>> PCboard log periodic that does cover 2.1 to 11 GHz has less gain and I
>> think the spill over the sides of the dish will be great, hurting the
>> transmit gain and the receive signal to noise ratio.
>>
>> The Vivaldi that I got from Kent is 5 - 18 GHz, but he says it works down
>> to 3.5 GHz.  It's quite a bit bigger than the 10 - 25 GHZ model.
>>
>> So Jerry, I wonder if you can clarify: are you saying that the Vivaldi will
>> do a better job at limiting overspill than the log periodic?  Or are they
>> both equally bad in that regard?
>>
>> Donn wrote:
>>
>> I look forward to seeing the multi-band dish on the antenna range.
>>    (Assuming the WX is such that we can do another antenna range.  We got
>> skunked at Aurora last month and next week's make up is iffy with the WX
>> pattern we seem to be in.  Maybe later in the summer....
>>
>> Thanks for your encouragement, Donn.  Yes, if the antenna range is offered
>> later in the summer I'm going to try hard to get up there for that.  The
>> earlier dates have not worked for me, but I'd love to have another crack at
>> it.
>>
>> 73,
>>
>> David  W9HQ
>>
>>
>>
>> On Fri, Jun 7, 2013 at 12:23 AM, Dr. Gerald N. Johnson
>> <geraldj at netins.net>wrote:
>>
>>> The Vivaldi that Kent sells is 10-25 GHz, not very useful on 3.4 GHz. His
>>> PCboard log periodic that does cover 2.1 to 11 GHz has less gain and I
>>> think the spill over the sides of the dish will be great, hurting the
>>> transmit gain and the receive signal to noise ratio. The dish will
>>> concentrate the received signal on the feed, but the feed will see noise
>>> sources like hot earth from the spillage.
>>>
>>> You could make a Vivaldi for a lower frequency range, just make it 3 times
>>> the size of his 10-25 GHz and it has better gain than the LP, but still may
>>> spray RF past the edges of the dish and pick up some noise sidelobes. It is
>>> possible to array Vivaldi's but it gets messy.
>>>
>>> 73, Jerry, K0CQ
>>>
>> ______________________________________________________________
>> NLRS mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/nlrs
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:NLRS at mailman.qth.net
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>>
>>
>
> ______________________________________________________________
> NLRS mailing list
> Home: http://mailman.qth.net/mailman/listinfo/nlrs
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:NLRS at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>