GH with Parabolic NARODs for flat VHF-HI response?

[RamKat]

Hi, I am a total newbie here. Don't have any experience with antennas other than bending simple reflectorless 4 bay bow tie and HG antennas.

My goal is to have a couple of small compact antennas that I can put in the attic to cover the local TV stations (Digital and Analog) from 2 transmitting towers here in Ottawa Canada .

I have followed some of the discussions on 300 Ohms excellent work on the SBGH with NARODs. Kudus to all you contributing folks and Stampeder running this forum.

As I have said I don't have any antenna experience but I do have worked with some optics in my career. What bothered me about the VHF hi response of the NAROD design was the peaky frequency response in the VHF hi band and the fact that the pattern "flips" at the lower end of the response i.e. being more sensitive from "behind" the antenna . An idea came up to "focus" more energy onto the NAROD. Being an optics guy my first thought was a parabolic reflector. So I downloaded 4 NEC 2, plugged 300 ohms NEC file in and played around with some additional NAROD reflectors in a parabolic pattern behind the NAROD.

I ended up with 4 additional elements that provide, what looks to me as a newbie, some promising results. The NAROD reflectors are placed so that the path length from the source to the NAROD Reflector and onto the NAROD is constant.

I have increased the length of the NARODs to 300 Ohms suggested 30.25" but left the spacing to the 1st reflector at 10" as his first design. Still have to play around with that parameter (Don't know how to use 4NEC2's optimization feature yet)

Since I am a newbie I cannot post figures but here are the first results in the VHF Hi band.
Freq[MHz] Gain SWR
175 8.75 8.65
180 9.6 2.88
185 8.95 3.89
190 9.26 5.12
195 8.84 4.69
200 8.49 4.26
205 8.2 4.12
210 7.97 4.48
215 7.77 5.24

(all in front of the antenna)

The UHF band's dip has shifted maybe a dB or so deeper.

By moving the NORAD reflectors around it is possible to change the slope of the response curve so that the 9dB gain is obtained on the hi end etc (saw 9.65 dB @ 213 Mhz on one config)

The SWR values above looks a bit on the high end.

Before carrying on on this path I would like to hear from the gurus if there is there something that this newbie is missing.

Below is the NEC file

Oh, the PR in the heading is for Parabolic Relector - for the lack of a better term

thanks



CM
CE
GW 30 15 0 1.73228346 0 0 6.73228346 5 0.05094856
GW 31 15 0 1.73228346 10 0 6.73228346 5 0.05094856
GW 32 15 0 1.73228346 10 0 6.73228346 15 0.05094856
GW 34 15 0 -1.7322835 0 0 -6.7322835 5 0.05094856
GW 35 15 0 -1.7322835 10 0 -6.7322835 5 0.05094856
GW 36 15 0 -1.7322835 10 0 -6.7322835 15 0.05094856
GW 38 15 0 1.73228346 0 0 6.73228346 -5 0.05094856
GW 39 15 0 1.73228346 -10 0 6.73228346 -5 0.05094856
GW 40 15 0 1.73228346 -10 0 6.73228346 -15 0.05094856
GW 42 15 0 -1.7322835 0 0 -6.7322835 -5 0.05094856
GW 43 15 0 -1.7322835 -10 0 -6.7322835 -5 0.05094856
GW 44 15 0 -1.7322835 -10 0 -6.7322835 -15 0.05094856
GW 79 25 -4.1338583 -0.3937008 15 -4.1338583 -12.440945 15 0.1875
GW 80 25 -4.1338583 0.39370079 15 -4.1338583 12.4409449 15 0.1875
GW 81 25 -4.1338583 0.39370079 -15 -4.1338583 12.4409449 -15 0.1875
GW 82 25 -4.1338583 -0.3937008 -15 -4.1338583 -12.440945 -15 0.1875
GW 84 25 -4.1338583 0.39370079 2.5 -4.1338583 12.007874 2.5 0.1875
GW 85 25 -4.1338583 -0.3937008 2.5 -4.1338583 -12.007874 2.5 0.1875
GW 86 25 -4.1338583 0.39370079 7.5 -4.1338583 12.4409449 7.5 0.1875
GW 87 25 -4.1338583 -0.3937008 7.5 -4.1338583 -12.440945 7.5 0.1875
GW 88 25 -4.1338583 0.39370079 -2.5 -4.1338583 12.007874 -2.5 0.1875
GW 89 25 -4.1338583 -0.3937008 -2.5 -4.1338583 -12.007874 -2.5 0.1875
GW 90 25 -4.1338583 0.39370079 -7.5 -4.1338583 12.4409449 -7.5 0.1875
GW 91 25 -4.1338583 -0.3937008 -7.5 -4.1338583 -12.440945 -7.5 0.1875
GW 103 13 0 6.73228346 15 0 12.3228346 15 0.05094856
GW 104 13 0 -6.7322835 15 0 -12.322835 15 0.05094856
GW 106 13 0 6.73228346 -15 0 12.3228346 -15 0.05094856
GW 107 13 0 -6.7322835 -15 0 -12.322835 -15 0.05094856
GW 138 29 0 -15.125 15.5 0 0 15.5 0.0511811
GW 142 29 0 0 15.5 0 15.125 15.5 0.0511811
GW 139 29 0 -15.125 -15.5 0 0 -15.5 0.0511811
GW 143 29 0 0 -15.5 0 15.125 -15.5 0.0511811
GW 144 23 -10 15.125 15.5 -10 -15.125 15.5 0.1875
GW 145 23 -10 15.125 -15.5 -10 -15.125 -15.5 0.1875
GW 144 23 -9.35 15.125 -20.5 -9.35 -15.125 -20.5 0.1875
GW 144 23 -9.35 15.125 20.5 -9.35 -15.125 20.5 0.1875
GW 144 23 -7.5 15.125 -25.5 -7.5 -15.125 -25.5 0.1875
GW 144 23 -7.5 15.125 25.5 -7.5 -15.125 25.5 0.1875
GW 146 3 0 -1.7322835 0 0 1.73228346 0 0.01
GS 0 0 0.0254
GE 0
GN -1
EK
EX 0 146 2 0 1 0
FR 0 38 0 0 473 6
RP 0 1 10 1510 90. 0. 0. 20. 0. 0.
EN

[300ohm]

Hi, I am a total newbie here. Don't have any experience with antennas other than bending simple reflectorless 4 bay bow tie and HG antennas.

Pretty good for a newbie, welcome to the forum.

As I have said I don't have any antenna experience but I do have worked with some optics in my career.

That may actually be more a help than a hinderance. A lot of times, the guys with the EE backgrounds get hung up re-studying the tree and missing the forest, heh. I too believe that a wave is a wave is a wave, be it water, RF or light (discounting the photon effect) and should respond similiarly in nature. I think once you learn the terminology youll do fine and it will be easier.

I ended up with 4 additional elements that provide, what looks to me as a newbie, some promising results. The NAROD reflectors are placed so that the path length from the source to the NAROD Reflector and onto the NAROD is constant.

Yeah, I also did that earlier, but didnt bother to post it because it makes the build harder. (but using Mec_Eng's build design and bending the pvc with heat, it can be done fairly easily) Ive also had no real success putting directors in front of the NAROD.

The UHF band's dip has shifted maybe a dB or so deeper.

Yep, theres always a compromise. Its like fixing an old garden hose in one spot, only to have it develop a new leak elsewhere, heh.

Before carrying on on this path I would like to hear from the gurus if there is there something that this newbie is missing.

Well, typical newbie mistake of not adjusting for AGT = 1.0 (0db). Directions on how to do this are in the modeling thread and also in the 4nec2 help (F1) under AGT in the index.

[300ohm]

Since I am a newbie I cannot post figures but here are the first results in the VHF Hi band.

You mean the charts or pictures ? Just capture the image with a program like SWBcapture. Then set up an account with a place like Imageshack and upload the image there, choosing the resizing option to keep the image 56K dial up friendly. Then click on the "Insert Image" icon here in the Reply to Thread page and copy and paste the Image Shack "Direct Link to Image" address in the pop up box here.

Like so:




It really is that easy !

You can also post your picture here on digitalhome.ca, but there is no user selectable resizing option, so you have to go thru another step.

[RamKat]

Thanks 300ohm! . I will look into setting AGT=1.0. There was this little voice..........

As far a construction goes the thermal forming of the PVC pipe also came to mind. I was planning on cutting a profile from a corriboard which can be picked up at the Dollarstore as "For Sale" Signs. Since it is going into the attic (in my case at least) wind resistance will not be a problem.

Thanks for posting the 3D

[300ohm]

As far a construction goes the thermal forming of the PVC pipe also came to mind.

Sch 40 pvc pipe, both the white plumbing stuff and the much cheaper, better UV resistant electrical gray stuff, bend very easy with a heat gun or a common cheap electric room heater. All you need is a form in the pattern you need, like made from some nails in a board, and a cold wet rag to help cool it down after you bend it.

[flcs3]

Originally Posted by RamKat

Being an optics guy my first thought was a parabolic reflector. ...

A while back I was thinking of the same thing, though in terms of gain and
apeture rather than bandwidth. (My market has only one VHF-hi station,
and though my mother's market will have 7, 10, 12 and 13, her antenna will
be mounted behind a large picture frame, such that it cannot have any
reflectors at all. For my own attic installation, a large planar reflector is, for
the moment anyway, under the most consideration.) So, I eagerly anticipate
the findings of your studies.

One thought did occur to me back then: distorting the parabolic shape a bit
might help broaden the bandwidth. If the standard circular arc is used as an
approximation to the parabolic arc in order to simplify construction, the error
so introduced actually may help with the bandwidth. The "error" caused by
larger diameter conductors should help, as it usually does. Additional
geometric distortion might help. Even a simple small shift off of the focal
point (well, axis) might help with bandwidth, at the cost of some gain.

As for your optics background, you probably have seen the use of corner
reflectors with TV antennas. Somewhere I came across the use of corner
cube reflectors (or should I say retroreflectors?), and Kraus's 1950 book
"Antennas" contains a section on LENSES! Alas, RF lenses present quite
a few practical problems, construction being just the first. (For the curious,
think of a large number of small directors arranged in a bounded space
shaped similarly to an optical lens. Then recall that "refraction" basically
is "diffraction" around a large number of atoms in a regular lattice.)

Since you are concerned with bandwidth, you might consider using a corner
reflector, though at VHF frequencies it will be smaller than "ideal". For an
attic antenna, it could be constructed with a pair of folded thin panels faced
with aluminum foil and placed behind each of the NARODs. A slight variation
is the "trough" reflector, where the "V" cross-section is truncated at the apex
and replaced with a flat secion. This moderates the depth of the reflector
assembly a bit, which might be of some concern. You might refer to the
ARRL Antenna Handbook (2005) ch 18 "VHF and UHF Antenna Systems".

Enjoy your new hobby! 8-)

Oh, the PR in the heading is for Parabolic Relector - for the lack of a better term

That works for me in context, though it momentarily may confuse some
newcomers, who may expect a large dish illuminating the whole GH. 8-)

[300ohm]

You might refer to the ARRL Antenna Handbook (2005) ch 18 "VHF and UHF Antenna Systems".

I picked up the 21st Edition 2007, which you can get cheap now, which also has the coupled resonator dipole section in chapter 7. (Of course, the NAROD or GHs arent dipoles).

who may expect a large dish illuminating the whole GH. 8-)

Heh, yeah like a couple hundred feet in diameter to illuminate the "whole" GH.

[RamKat]

Originally Posted by flcs3

So, I eagerly anticipate
the findings of your studies.

Will post it as soon as I get a something better. I am expecting to have a busy week so maybe have a go at it again next weekend

One thought did occur to me back then: distorting the parabolic shape a bit
might help broaden the bandwidth. If the standard circular arc is used as an
approximation to the parabolic arc in order to simplify construction, the error
so introduced actually may help with the bandwidth. The "error" caused by
larger diameter conductors should help, as it usually does. Additional
geometric distortion might help. Even a simple small shift off of the focal
point (well, axis) might help with bandwidth, at the cost of some gain.

That is exactly what I tried this PM - with some success - got two peaks with a nice gain at the bottom and middle but I could not move the high end higher than about 7.7 db. Initialy started with a circular shape - that is the one where the high end ended up with the highest gain > 9.5db and the bottom end was at about 7.5 db - so it should be possible. Still have to learn and see what affects what. In most cases I think is like to filling a balloon with water and then to press down on it with your hands.

You might refer to the
ARRL Antenna Handbook (2005) ch 18 "VHF and UHF Antenna Systems".

Certainly going to see if I can find the 2007 copy - as per 300ohm's post.

Enjoy your new hobby! 8-)

Was already in trouble this PM for spending too much time with NEC

[300ohm]

Was already in trouble this PM for spending too much time with NEC

Get used to it, heh.

I will look into setting AGT=1.0. There was this little voice..........

Some notes and tips on doing that : If AGT is greater than 1.0, decrease the radius of the wire. If AGT is less than 1.0, increase the radius of the wire. If that doesnt do it, resegment the whole model and start again.
Print out and keep an AWG wire size chart handy, with conversion to radius in both millimeters and decimal inches. (a print out of TV channel frequencies also comes in very handy) In other words, do what you have to do to make life easy for yourself and post the tip here for others.

Will post it as soon as I get a something better.

Yeah, like I noted to JED a while back, we have to be very careful about posting too many models here. (Ive only been at this modeling for about a year, and Ive got pretty close to about 1000 or so different models per se. (including infamous 16 bay GHs that no sane person can build) 4nec2 and a fast 'puter makes it all too easy, heh) Keep in mind, regular casual builders read this site and they already find what is posted to be too overwhelming for them and thus become discouraged. This is not my goal, and Im sure its not yours. I would suggest, going on Autofils suggestion, that we not post models unless there is at least a .5 dbi increase in gain somewhere somehow, or at least its a unique design. (Face it, .2 dbi gain is nothing, heh)

[300ohm]

Certainly going to see if I can find the 2007 copy - as per 300ohm's post.

Another must very good read is Adam Smiths (my Professor of Economic Thought pronounced him Adamas Mace. It took 3/4 of a semester to figure out who the heck he was talking about, he-he) Wealth of Nations, published by Harvard Classics. It will help you to understand the importance of being cheap.

[j3d]

Keep in mind, regular casual builders read this site and they already
find what is posted to be too overwhelming for them and thus become
discouraged.

If this is the case, then perhaps the scope of this forum is too
broad. In fact, the forum's description is: "Discuss OTA TV Antenna
Design, Theory, Computer Modeling, Building". I feel that the forum
should be split into a "building" component and a "modeling" one.
--John

[stampeder]

I've been concerned about the broad mix of the practical and the theoretical in the Antenna Research & Development forum but actually I'm seeing that they feed each other symbiotically, so my preference is to keep fine tuning this introductory thread that I recently renamed the "Which Gray-Hoverman TV Antenna Should I Build?" thread:

http://www.digitalhome.ca/forum/showthread.php?t=81982

I'll also work on writing an advisory post to that thread to let casual builders know that they should be aware of the theoretical nature of new models and wait for the thumbs up before attempting them.

Sort of like the sword swallower saying "Do not attempt this at home, I am a professional!"

[RamKat]

Hi 300ohm, thanks for your response - I started to look at adjusting the AGT, still have alot to learn - but that is the another fun part it.

Being cheap - I have dutch ancestors, that is our motto

[300ohm]

Arie Voors is Dutch. (I think. If not, thats one odd Eskimo name, heh.)

[rbhaug]

I am a newbie, very creative and useful forum(s). Since virtually all users are very likely to have VHF ch in their view, the NAROD version is very desirable.
While many can live with the narrow VHF peak, any help in widening the acceptance (frequency) is desirable. Has anyone tried (is it valid ?) tuning the top and bottom to different frequencies/element lengths, one would expect a drop in response, but it may spread the response.
The SBGH is easy to modify for the VHF with the length adjustment in the additional elements of the NARODs.
I am reticent to begin modelling, as my experience has been it takes a very sophisticated RF type to have the right insight to 'guide' the approach and random 'freshman rule of enthusiasm' mods may tend toward more 'noise' than 'real' progress -- I am VERY sensitive to 'buildability' and simplicity at that stage, with great sophistication in the modelling and 'tooling' phase.
With early ATSC tuners a YAGI was often useful to reduce multipath, I do not know if the current gen 5,6 etc are completely past that.