Canadian TV, Computing and Home Theatre Forums banner

Introduction to the Gray-Hoverman TV Antenna (the original thread)

52K views 66 replies 15 participants last post by  Yurii Pylypenko 
#1 ·
Last summer I experimented with the Hoverman array antenna at a cottage near Lombardy and found that it had very wide bandwidth and could receive VHF and UHF channels quite well...

Perhaps you could consider the Hoverman array as an alternate antenna front-end to the Yagi for your preamp ultimate antenna.

I made it from #8 Al Grd wire with two lengths of 42". Each side of 42" in length, is bent into six 7" segments at right angles. The two bent 42" lengths are then mirrored in a vertical plane and feed from the middle using a 300-75 ohm bauln.

If you are interested, I have the array built and you are welcome to try it out. It might give you a single antenna front-end for the entire UHF band ...maybe even Ch7, since WWNY-Watertown Digital will revert from Ch35 back to 7 in 2009.

In consultation with Autofils and the main contributors to his ongoing Gray-Hoverman project it has been decided to make the designs, schematics, and diagrams of the Gray-Hoverman into public property, Copyright 2008, and distributed under the GPLv3 license:

http://www.digitalhome.ca/ota/superantenna
 
#3 ·
Stampeder,

Here is a photo of my modified Hoverman array: http://www.flickr.com/photos/21971251@N04/

This is only the active array, no reflector was used.

Each segment is 42" of #8 Al Grd Wire, bent in a zig-zag at right angles with six equal sub-sections of 7".
The spacing between the segments is 1.25" (not critical) and the array is feed from the middle of the array at nominal 300ohm.
For Coax, use a 300-75 ohm balun.

You can view the original patent at http://www.pat2pdf.org/404.html.
Just enter the following patent number 2918672
 
#4 ·
More on Hoverman

Stampeder,

Perhaps this link is the article that you were referring to re Hoverman.
VUD "The Hoverman" 2 page pdf article -Sept 1982.

http://www.wtfda.org/images/stories/pdf/hoverman.pdf

This is the Hoverman array that I built, plus it has an interesting twist for reflectors. Instead of using a screen type reflector, it uses 4 pairs of co-linear arrays. The top and bottom pairs appear to be larger than the the middle pairs.

The gain plot shown in the article states a 10-12 dB gain from Ch 14 to 58, with a sharp falloff above Ch 60. There are two peak gains at Ch 25 and Ch55. I believe these peaks are due to the pair of co-linear array reflectors. The top and bottom provide the Ch25 peak, and the middle pairs provide the Ch55 peak.

The overall dimensions of the antenna are 30" x 30" x 6"; so I'd guess the top and bottom reflector co-linear elements are 14-14.5" and the middle co-linear elements are probably around 10-10.5". I would guess the reflector spacing to be 4.5".

Anyone have the skills to run a gain emulation for this antenna?
Wondering if it would be worth while to give this a try, maybe even a double vertically stacked arrangement.
 
#5 · (Edited)
Autofils said:
maybe even a double vertically stacked arrangement.
This is where I think that a stacked yagi would be preferrable. I think the gain would be better because of the director elements. Nevertheless it does look like a good antenna at least from the photo. Would be more difficult to construct than a yagi.

If you want to make one I will make a low noise preamp for you to attach to it.
 
#6 ·
When I get a bit more time I may try to build this hoverman. I had read something on it a while back at another "do it yourself" antenna site. I think it would be interesting to see what the results at ch58 would be as it peaks there, its also high on ch33. These are my two local DT channels here in Vancouver.

It is raining out here in Vancouver so before I can test Gerry's super antenna again I need to waterproof it I suppose. I am amazed at the reception over 200 miles away, If only I could get it to go at 100 miles.
 
#8 ·
The original Hoverman patent did not have any reflector and used 56" segments with eight zig-zag 7" sub-elements. The two added horizontal 7" sub-sections per segment, provide VHF low channel coverage down to Ch 2.

I've only tried the Hoverman with 42" segments and without the reflector, and I was absolutely surprised with the wide bandwidth. This turned out to be very good for reception of all local channels here in Ottawa, ranging from Ch4 (CBC) up to Ch65 (CityTV)

Come the spring, I plan to experiment with the 4 pair of co-linear reflectors, as I had discussed in my earlier post.
I'm in Bells Corners and I'm hoping to get WNPI digital on Ch23, and maybe with a preamp get WWNY digital on Ch35.

Can anyone point me to any info on how to model this antenna? That would be very helpful for experimentation.
 
#9 ·
Autofils said:
Can anyone point me to any info on how to model this antenna? That would be very helpful for experimentation
I think Old Sparks has a good antenna modelling program:
Old Sparks said:
Let me know if you want me to model your antenna using NEC II before you build it. NEC II will give you a very accurate representation of performance. It doesn't 'do' boom correction, though...
You out there, Russ? :) I'd love to see how the Hoverman models.
 
#10 ·
I'm warming up to the Hoverman. Especially after seeing this:
The auction, which begins on Jan. 24, is for radio spectrum situated around the 700 MHz band, an asset that will be made available in 2009 when analogue television broadcasting is shut down.
I forgot there will be nothing above channel 51 (700 MHz) in a year from now. Which makes the Hoverman a good antenna for the entire band. BUT the modified yagi I built will also work really well too. It all depends on the shape of the antenna you want I guess.
 
#13 ·
Old Sparks,

Thanks so much for modelling the Hoverman. The results clearly show that it would be worth while to experiment with the original 56" (8 sub-sections of 7") using a screen reflector.

When I tried the Hoverman (array only) at a cottage near Lombary this past summer I was quite surprised that it got a clear signals on analog channels 4-CBOT Ottawa; 7-WWNY Watertown; and 11-CKWS Kingston; as well as the UHF channels.
So this makes an excellent antenna for all the local Ottawa stations.

This leads me to respectfully ask for a couple of additional variations on your modelling of the Hoverman.

1. Examine gain and SWR of Hoverman (56" per segment with 8 sub-sections of 7") array only; over the VHF range (50 to 220Mhz),

2. Examine gain and SWR of Hoverman with Screen reflector with a reflector spacing of 38cm, in place of 10cm. This is a compromise VHF-UHF relector spacing; which is about 0.22y at Ch7; and 0.65-0.81y for UHF range ch20-50; (where y = wavelength).
Run this for 50-220Mhz and also 400-800Mhz.

If this shows some gain for Ch7-13, as well as UHF, then real experimentation with a Hoverman stack (horizontal) with a preamp might give us Ottawa folks a chance to get WWNY digital when they revert back to Ch7 from 35 in Feb 2009 at their full power.

Your link to results on other commercial antenna, had an interesting Ch7-13 VHF for the highly recommended CM-4228 (8-Bay), which shows high gain ch10-12, but much lower on Ch7. Would be interesting to compare a Hoverman Hor Stack with screen reflector to the CM-4228.

Again, my thanks for your modelling work. It is certainly a great Christmas present for me and likely all who follow this forum !!
 
#14 ·
1964 Improved Hoverman Patent using Reflectors

I finally found the Hoverman patent ( 8 Sept1964) that includes reflectors.

You can view the second patent at http://www.pat2pdf.org/404.html.
Just enter the following patent number 3148371

He describes two designs with 4 rod reflectors; full wavelength and co-linear half-wavelength reflectors.

The driven array is same as original (56" dual segments with 8 subsections of 7")

Reflector spacing is 3.5"

Full Wavelength Reflectors:
Top and bottom 29"
The two middle 24"

Half Wavelength Co-Linear Reflectors
Top and bottom 14"
The two middle 10"

The above dimensions are for UHF range Ch14 to Ch35, as claimed in the patent.

He gives design equations for shifting the range, and suggests Ch35-58 and Ch58-83, although this range 58-83 is not applicable now, as TV on UHF only goes to Ch69 and after 2009 will only go to Ch51.

An Interesting read. :p:p
 
#16 ·
GerryB,

The 1964 Hoverman patents states " a conventional two-conductor 300-ohm transmission line may be connected to the driven element".
So it is probably close to 300 ohm; like your folded dipole.

Does your preamp use the ATF-34143 or ATF-54143? That seems quite popular with Radio Hams.
I found an Ebay seller with a package of 5 for both types.
Also would your preamp be useful for VHF Ch7 (172-176Mhz)?

I had a look at some of the app notes, and that circuitry is very tricky indeed and a bit beyond my electronic skills,
I'd be interested in getting one or two to play with, and would cover your costs.

Also if you would like to try a Hoverman array to compare with your dipole antenna, just pm me and I'll get it to you.

...Regards
 
#17 ·
I've updated my 'using NEC 2 to model the Hoverman' webpage at:

http://www.qsl.net/va3rr/hdtv/hoverman.htm

I've included two NEC files developed by Autofils which make some interesting modifications to the original hoverman with rod reflectors. I may try to build his UHF version come springtime as it would have much less windloading than an antenna that uses a screen reflector.

Russ
 
#18 ·
4nec2 Modeling Results for Modified Hoverman UHF [Ch 14-51]

Thanks to old sparks, I was able to experiment with variations on the Hoverman patent 3148371 using the 4nec2 modeling program.

For my modified Hoverman, the 4nec2 model shows a fairly even gain over the Ch 14-51 range with a very good swr.
(see attached composite jpg).

The pertinent dimensions of this modified Hoverman are as follows:

**** ARRAY ****
1. Length of diagonal sub-sections = 180mm
2. Length of horizontal sub-sections = 127mm
3. Feed Point Gap = 45mm


**** Co-linear Rod Reflectors ****
1. Length of Top and bottom pairs = 330mm
2. Length of middle reflector pairs = 270mm
3. Spacing between Array and reflectors = 114mm
4. Gap between co-linear pair elements = 20mm

Total overall antenna dimensions are 762mm x 680mm x 114mm

My thanks to old sparks for the original nec files.

If the actual results follow this modeling results, this should be an excellent full range UHF antenna. It's very easy to make, very light with very low wind loading for outdoors installation.

The results for the VHF-UHF modified Hoverman are not nearly as good; only about 6dB gain at Ch7; so the long term goal of getting WWNY with a preamp will probably be better using a Ch7 yagi design.


 
#19 ·
Hoverman impedance matching (long)

Hi guys, I'm a lurker/new member and have read lots of good stuff here but I need some help. I hope you don't mind the length of this post.

The Hoverman seems like a great DIY project and I am going to build it. But there is a problem.

In the days of 300 ohm inputs and 300 ohm flat twin lead, a single Hoverman was a perfect match for the TV imputs of the day. But that was then and this is now (75 ohm).

From what I understand, using a balun to impedance match results in a very large loss of signal at UHF and in effect defeats the benefit of a good antenna design like the Hoverman. Also, home made baluns made from shielded lead are bandwidth limited and will impede the wide bandwidth features of the Hoverman. There must be a better way.

I was thinking that by paralleling four Hovermans the resultant impedance would be 75 ohms and there would be no need to impedance match at all. I could then go direct to 75 ohm shielded lead and be done with it. So my first question.... Is that true?

My second question is what kind of problems would exist with that kind of a set up? In other words how should the array be wired so as not to screw up the natural benefits of the Hoverman?

I did a bit of surfing and found the following antenna. It seems to be a double Hoverman (It sure would be nice if somone could model and optimize this design with dimensions provided to all!)
The website where I found the pic does not describe the impedance of the pictured design. So now to my third and fourth questions. Is it 150 ohms? Can it be made a native 75 ohms?

My fifth question is how would one connect two of them to get phasing right, not interfere with the fundamental antenna design and get down to a 75 ohm impedance? Is a simple parallel connection the right way to go?

My last question is about an active impedance match. Are there any VHF/UHF or UHF antenna amps out there that have high enough input impedance that they can be used for an antenna with any impedance (300 or 150 or 75 ohm) and still provide a clean 75 ohms out for modern TV (or HDTV) inputs? If such an amp existed I could build any version of the Hoverman and still get a good match at my shielded cable and RF inputs.

Suggestions? Comments?

Thanks
 
#20 ·
oneoldude, To partially responds to your questions, I believe that effectively hooking up four Hovermans would results in 75 ohm. However, you would have to unbalance the output from the antenna array by using ferrite beads or create a choke balun (several turns of RG-6 on an air core).

I wonder if making such an array of Hovermans would result in high directivity from the resulting antenna.
 
#21 ·
I was also wondering about directivity - that many Hovermans could create a much different beam pattern, depending on whether they are side-by-side, stacked, or a mixture.

Another thought that isn't purely antenna-related is that you could run a single Hoverman into a preamp that takes 300 ohm twinlead input but gives 75 ohm coaxial output at a high signal level:
  • The Channel Master 0064 Spartan III does that with an advertised gain of 23dB in the UHF band.
  • The Winegard AP8283 and AP8783 both do that with an advertised gain of 28dB in the UHF band.
 
#22 ·
Modeling the Double Hoverman

From "oneoldude's" posting #264 showing a pic of a double vertical Hoverman, with 7 reflectors, I found a web site for the SuperG-1483 that gave minimal info on the dimensions, but enough to be able to get a ballpark to start on modeling it.

http://www.summitsource.com/antenna...-part-super-g1483-with-coax-cable-p-6356.html

Boom length = 60 inches
Max Width = 30 inches
Depth = 5 inches
UHF Ch 14-69 with 35- 60 mi range; Avg Gain 11dB
Number of Elements = 18

If you consider that the reflectors are co-linear, then reflectors are 7x2=14 elements and each array has 2 elements; giving a total of 14+4=18 elements.
The total width of the biggest co-linear reflector is 30 inches.
The maximum reflector to array spacing is 5 inches, probably something less.

Since the total height is 60 inches the height of each array + gap between arrays is 60 inches. Scaling from the picture, looks like the gap between arrays is 3-4.5 inches, with each array at about 27.75 to 28.5 inches.

This would result in a length of 6.54 to 6.71" for the array's diagonal elements.

Thus I have a starting point for the model.

A couple of points about this SuperG-1483:
The estimate for the length of the array diagonal elements is shorter than my modified Hoverman. This is probably due to their claimed UHF range up to Ch 69. With the coming changes to the frequency spectrum, need only to go up to Ch 52!

The SuperG-1483 only has 7 co-linear reflectors. I suspect that if both arrays had 8 co-linear reflectors and a larger gap between the arrays, there would be more gain. The modeling results will tell the tale.
Once I have completed the modeling, I'll post the results. Probably in about a week or less.

Concerning signal loss of 300 to 75 ohm baluns, there is a wealth of info at hdtvprimer.com. The following link indicates a loss of 0.2 to 2dB over the UHF range for the bauln. ChannelMaster seems the best.
There is also a discussion about cable loss and the use of preamps that is an interesting read.

http://www.hdtvprimer.com/ANTENNAS/basics.html
 
#24 ·
Thank You Autofils

Thanks for your post and the work you are doing. It will be most interesting to compare the results of these antennas against the commercially available ones.

I am waiting with anticipation for your results.

These antennas can be easily made from 1/2" rigid copper pipe and standard copper connectors and end caps soldered together. Would there be a benefit of using such materials over simple aluminum, or copper wire? While they would be heavier than the wire types, they should stand up well in heavy wind like we get here in Florida from time to time.

Thanks again.
 
#25 ·
Building a Hoverman

oneoldude,

The Hoverman is a very simple antenna to build. If you plan to install it outside with a rotor, copper pipes might be a tad too heavy.
I would advise using #9 Alumininum wire as mentioned in the original patent. It is light, but quite rigid, and should resist corrosion.

The effect of using larger diameter wires, or pipes would give a slight gain increase over 1/8" aluminum wire, in the order of +0.1 to 0.3 dBi.

I'm in the process of optimizing the Double Hoverman. I find that as I get more experience with the modeling I get a better understanding of this Hoverman design and have made further improvements in the single bay Hoverman. that I posted earlier (Post #252)

I also got EZNEC v3 modeling program and will be able to compare the Hoverman to CM-4221 antenna, as modeled by kq6qv using his CM4221a.ez file.

Note: kq6qv models with the NEC-4 model program. EZNEC v3 is NEC-2, but it can run the CM4221a.ez file, so the comparison of Hoverman to CM4221, using the EZNEC v3 program will be valid.

If you have an interest in NEC-2, here is the wiki page...
http://en.wikipedia.org/wiki/Numerical_Electromagnetics_Code

A lot of interesting things are coming !!
 
#26 ·
Building A Hoverman

Autofils,

I have built a prototype improved Hoverman (no reflector) with 10 Ga. copper wire. It is way too flimsy to put up on my mast. So I am looking for some sturdier material. I do not know what type of business sells the large diameter Al wire so I have not found any yet.

As my name implies, I am an old dude. It is not easy (or safe) for me to climb my tower. I have to remove a dead rotor and an ancient UHF-VHF monster from RS in order to install what I will build. I would like to get it right the first time.

Re the question about copper pipe, I understand that due to end correction, copper pipe might need shorter elements than those made of thinner wire. Also, the larger dia. copper pipe might result in a wider bandwidth.

One of the major network stations here will be using ch. 7 for its HD signal. Everyone else is in the UHF band. I do not mind loosing the sole VHF station, but what the heck. If I can get it, great.

I am in Sarasota, Fl South of Tampa. I am in between two antenna farms. One is 40 miles away NNE and the other is 57 miles away SE. A friend has a CM4221 and is able to get over 50 channels (including the -1, -2, -3, versions). So I am looking to have some fun with the hoverman designs.

Thanks for your help.
 
#27 ·
Oneoldude,

I've almost completed the Hoverman modeling, so some of your questions will likely be answered when I post the model results in a couple of days.

The model has used #9 Al wire, and I have purchased a 40 ft roll on Ebay. So you could do a google or Ebay serach for 40ft roll Aluminum ground wire. It is about 1/8" diam. If Al wiring is allowed for house wiring in Florida, Home Depot likely has it. Here in Ontario, Aluminum wiring is not allowed, so I went the Ebay route.

If you decide to use copper pipe, just send me a pm via this forum stating your pipe's diameter and I can run the model with that info. That way, you would have the optimum dimensions for your Hoverman build.

The model results shows that the Hoverman has a bit more gain than the CM4221 at the low UHF end - Ch14-30 range, so your reception results compared with your friend's CM4221 would be a great way to determine "real-world" results vs the model results.

The model also shows that the co-linear rod reflectors are most important, as the gain characteristics of the array-only is not that great.

Regarding Ch7, the model also shows the Hoverman has a gain of 3.29 dBi at 177Mhz (Ch 7) which is a more than a simple Ch 7 dipole, so if that station is close, maybe you'll get it with the Hoverman.
 
#28 ·
Building A Hoverman

AUTOFILS,

It will be interesting to compare the results of the Hoverman and the CM42221. My coverage area is shown here:

http://ota.winegarddirect.com/?address=2105+mcintosh+road,+sarasota,+fl+34232

Interstingly, my friend's 42221 gets better reception pointed Southerly rather than Norththerly! He gets all the Northern stations except for 10-1 (24) which is distant, weak and NW and 8-1 (which is actually 7 and VHF). He gets all that with the antenna aimed at the SSE! It seems the 42221 is getting signal off the side and behind its screen. Will wonders never cease? When he points in a Northerly (NNE) direction he gets none of the Southerly channels. Hmm.

I find the Winegard site interesting because it gives the station signal strength (I guess at my position) and the required antenna gain (the assumptions they make on that should be interesting). I also find that negative antenna gain is an interesting concept. When they show a negative gain I suppose they mean the antenna needs less gain than a simple wire antenna with a gain of 1. Hmmm.

In regards to Channel 8-1 (7) the Winegard site shows signal strength of 59 dBu and a required antenna gain of -22.5 dB (no reference given). So the Hoverman with a gain of 3.29 dBi might capture the channel. What do you think?

I am sure the co-linear rod reflectors will get the job done. But that beggs the question as to wether the same or better results could be obtained with a grid type reflector. Actually, the grid might be easier to make with hardware cloth somewhat like the 42221. Hmmm.

The results you will present will be interesting. I am looking forward to seeing them.

Thanks
 
#29 ·
Improved Hoverman called GRAY-HOVERMAN

Well I have finally completed the modeling of the Hoverman UHF antenna.
I started with the original patent by D.R. Hoverman (# 3148371 granted on Sept8 1964).

The initial modeling results of the original patent with 4 pair of collinear reflectors, indicated the gain was less than the highly recommended commercial antenna by Channel Master CM-4221.

Since the CM-4221 uses a screen reflector of 20"x36", it seemed reasonable to expect better gain with the Hoverman, using a screen reflector. However, I liked the Hoverman with rod reflectors, since it is easier to build and has much less wind loading. So I experimented with adding more collinear reflectors to simulate a screen reflector and to my amazement found that by simply adding two more pair of collinear reflectors, a significant improvement resulted in the gain characteristics as well as a modest improvement in the F/B ratio.

I have dubbed this as the "Gray-Hoverman" antenna.

Using the EZNEC v3.0 modeling program, I was able to compare this 6 pair reflector array (Single and Double bays) with the CM-4221, using the cm4221a.ez modeling file by kq6qv, available at his hdtvprimer.com website.

Summary Results
===============

SingleBay Gray-Hoverman is better than CM-4221 over Ch 14-44 range !!
DoubleBay Gray-Hoverman up to +2.3 dB over SingleBay

The comparisons of gain are shown in the Gain vs Freq photo.
EZNEC v3.0 results are only given in tabular form, so I made a best-fit graph, which shows the EZNEC v3.0 results for:
1. Hoverman Array Only
2. SingleBay Gray-Hoverman with 6 pair collinear reflectors
3. DoubleBay Gray-Hoverman (vertical stacked) with 11 pair collinear reflectors
4. Channel Master CM-4221



Photos of the Polar plot at 500 and 660Mhz are shown as well as the swr plot, as calculated by EZNEC v3, for the Original Hoverman patent, SingleBay Gray-Hoverman and Channel Master CM-4221.



Interesting to note the polar plots of the Original patent vs SingleBay. The increase in gain is significant and you can see the improvement in the F/B ratio. The beamwidth of the SingleBay is 38.2 degrees vs 47.2 for the CM-4221.

I have also include a photo of the Gray-Hoverman dimensions. The Double Bay uses the same reflector and array dimensions. The spacing between the arrays is 127mm.
The model used elements made with #9 Aluminum solid wire (1/8" diameter).
I'd recommend using #9 Aluminum solid wire, as it is light but rigid and somewhat corrosion resistant.

You could also use wire coat hangers. Probably need 8 hangers for a SingleBay. Remember to scrape and clean the wire at the connection points.

Dimensions for the SingleBay Gray-Hoverman Antenna Copyright 2008, and distributed under the GPLv3 license



Dimensions for DoubleBay Gray-Hoverman Antenna Copyright 2008, and distributed under the GPLv3 license


On my prototype build, I used a 1" OD pvc pipe as the boom.
The rod reflectors and array are attached to insulators (70x25mm) made from plastic cutting board. (The plastic cutting board, purchased at a dollar store, was 24x14cm and 6mm thick. It was cut using a hand bench saw)
The array was attached to brackets, made from galvanized metal strapping, that went around the pvc pipe and was formed to mount the insulator at the reflector spacing distance of 100mm.

Please post your reception results, if you try this Gray-Hoverman antenna.

Regards
...Autofils
 
#35 ·
(...)
The model used elements made with #9 Aluminum solid wire (1/8" diameter).
I'd recommend using #9 Aluminum solid wire, as it is light but rigid and somewhat corrosion resistant.

You could also use wire coat hangers. Probably need 8 hangers for a SingleBay. Remember to scrape and clean the wire at the connection points.
(...)
I just went shopping and I found at Home Depot some "hanging wires", gauge 12 at 7.00$ (100 feet) which is roughly the same (a little bit thicker) as coat hangers. I think this wire is rust proof, however not stated on the package.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top