First, let me send kudos to DjiPi, for his reception report in this post: http://www.digitalhome.ca/forum/showpost.php?p=703804&postcount=52

That was a significant amount of work and provides members of this forum with an excellent comparison of the Gray-Hoverman with a Screen Reflector, to the CM4221 clone.

DjiPi's results (post #322) would seem to confirm the EZNEC modeling results.

Copyright 2008, and distributed under the GPLv3 license:

I have now completed the optimization of the reflector screen for the SingleBay and the following graph shows the Net gain for:

1. Full Reflector Screen [Overall size is 30 x 40 inches]
2. Split Reflector Screen [Each section is 14.5 x 40 inches, with a separation of 1 inch]
3. 6 Pair Rod Reflectors
4. Channel Master CM-4221 for comparison

The spacing between the array and reflector is 100mm.


http://i258.photobucket.com/albums/hh262/autofils/NetGainSBayGray-HovermanOptimal.gif


This shows the additional gain that a optimal screen reflector provides, compared to the 6 pair rod reflectors.

For the reflectors screen, there is a slight increase in gain, if the reflector screen is split into two sections with a 1 inch separation.

For those that are looking for #9 Al wire, I found some on Ebay; just search for "Aluminum Ground wire".

If you try this antenna, please post your results to this forum.

First, let me send kudos to DjiPi, for his reception report in post #322.
That was a significant amount of work and provides members of this forum with an excellent comparison of the Gray-Hoverman with a Screen Reflector, to the CM4221 clone.

DjiPi's results (post #322) would seem to confirm the EZNEC modeling results.
(...)

Thank you Autofils, and you should be sent kudos as well for the modelling work that you have done, as well as any person that helped in any way:p. This helped me a lot (and convinced me) to build the SBay-Gray-Hoverman !

My wish now is to try the Double-Bay, with a similar reflector size adapted for the double-bay. I'll let everyone know the results. If you had some spare time to model the double-bay with optimal reflector size, that would be very appreciated. In the mean time I'll experiment.

Great work !

Hi t/s (or s/t),

I would suggest that you build yourself the Gray-Hoverman instead of the CM4221 clone. The Gray-Hoverman is a little bit better, so unless you want to build different antennas for fun or curiosity, this would be the shortest way to acheive your primary goal ;). By the way, where are you located to get 16 digital channels ?

Everybody: I rearranged the pictures of my album from my post #322, now comparing the channels is a lot easier, and the way I did it will allow me to post pictures when I'll compare the CM4221 vs Double-Bay Gray-Hoverman.

The EZNEC v3 modeling of the Double Bay Gray-Hoverman have been completed and are presented in this posting.

The graph below compares the Net Gain of the Double Bay Gray-Hoverman to the Channel Master CM-4228 for the following versions:

1. DoubleBay Array-Only (No reflector)
2. DoubleBay with 11 pair Colinear Rod reflectors
3. DoubleBay with Screen Reflector 30x75 inches
4. Channel Master CM-4228 (kq6qv Nec4 data for CM-4228)

http://i258.photobucket.com/albums/hh262/autofils/NetGainDoubleBayGray-Hoverman.gif

These results indicate the the DoubleBay Gray-Hoverman (Colinear Rod reflector and Full Screen reflector versions) are a bit better than the Channel Master CM-4228. :p

I initially thought it strange that the screen reflector had a "bump" in gain from Ch40, but when compared to the DoubleBay Array-only, you can see that "bump" is also present. I did not find any significant gain difference between full or split screen, as was found for the SingleBay.
The gain decreased for wider screens, so I believe 30x75 inch screen is very close to optimal for the DoubleBay.

If you are considering an outdoor installation, the collinear rod reflectors will have a much lower wind load.

If you try this Gray-Hoverman DoubleBay, please post your reception results.

These results indicate the the DoubleBay Gray-Hoverman (Colinear Rod reflector and Full Screen reflector versions) are a bit better than the Channel Master CM-4228.Now if we can come up with an easy-to-read set of schematics on the physical contruction we will have a real winner!

DjiPi, will you be fabricating the Double Bay Gray-Hoverman any time soon?

BTW, Autofils doesn't mind if we refer to it as the DBGH for short.

Given all the focused R&D about the Hoverman I was contacted by DjiPi about putting it all into its own thread. Now that we are seeing concrete successes I've moved all of the Hoverman-related posts here. :)

Now if we can come up with an easy-to-read set of schematics on the physical contruction we will have a real winner!
...


The dimensions for the SingleBay and DoubleBay, Gray-Hoverman with colinear rod reflectors, are specified in this post: http://www.digitalhome.ca/forum/showpost.php?p=694336&postcount=31

Copyright 2008, and distributed under the GPLv3 license:

Reflector screen for the singleBay is 30x40 inches overall. The split screen is two sections 14.5 x40 with a separation of 1 inch, between each vertical section.
Reflector screen for the DoubleBay is 30x75 inches.
The reflector screen used in the model was based on a wire grid structure of 2"(hor) x 1" (vert), same as CM-4221.
For indoor applications, tin-foil will work quite well.
For outdoor applications, the colinear rod reflectors will have a much lower wind load.

The reflector screen is centered with respect to the array(s) and the reflector spacing is the same as the colinear rod reflectors, at 100mm.

(...)
DjiPi, will you be fabricating the Double Bay Gray-Hoverman any time soon?
(...)
Yes, in fact the DBGH was the antenna I made first, but to report results on the SBGH, I took apart the second bay. Now it's time to put it all back together again and to install the reflector.

Just a note about my pictures on SBGH reception report: I just took out the CM4221 clone and put the SBGH pointing in the exact same direction, so I didn't try to pinpoint each channels for best results. That was a very rough comparison.

Autofils noted that channel 44 and 62 didn't seem to match the model, so I double-checked if I had mixed up the pictures. It appears that I didn't.

So, yes, I will be able to report results soon (I'm very excited now that I see that the performance would exceed in theory the CM4228 :D).

Thanks to Autofils, as well as DjiPi, Gerry B, Old Sparks, Kro, (and any others I may have missed) for your informative posts. I think many of us can't really appreciate the modeling work you have done here.
I have a partially completed 4228 clone in my workshop (which I started last summer! :( ), which I'm now thinking I might scrap and try my hand at your modified DBGH. I have very limited spare time so don't expect to see any results for a very long time ;), but I promise to post when I get done!!
The reflector screen used in the model was based on a wire grid structure of 2"(hor) x 1" (vert), same as CM-4221.
Unfortunately I only have 1" square construction screening to use as my reflector. What do you think I should expect since my screen material has twice as many vertical wires? If needed I can easily cut out the extra wires.

As well, perhaps I missed this, but do you have any idea what the overhead plot will look like for your modified DBGH ? Has anyone generated a reception pattern for this?

99gecko,

Your 1"x1" construction screen would be fine as is; no need to cut out any wires.
The reflector screen for outdoor applications, is generally made with the largest grid structure based on the highest operating frequency of the antenna, to minimize the wind load, so a 1"x1" will probably be a bit better.

I assume by overhead plot, you mean the polar plot. I have those results at 500 and 660Mhz for the DBGH, colinear and screen reflector. I'll get the photos posted soon and that will show the improved Front/Back ratio for the reflector screen compared to the 11 pair colinear rod reflector.

I look forward to your DBGH progress.
For me, the outdoor install has to wait until spring and at the rate we are getting snow here in Ottawa, it may well be a late spring this year. :(

Yes I meant polar plot.
at the rate we are getting snow here in Ottawa, it may well be a late spring this year.
If you lived in Ottawa during the winter of Feb 1971, this winter should be a breeze :). My parents took a photo of me standing on the snow looking in my bedroom window..... it was on the 2nd story !!

cheers

PolarPlots for the Gray-Hoverman were run on EZNEC v3 at two frequencies.
500MHz (Ch19) and 660MHz (Ch45).

Single Bay


(a) SBGH Collinear Rod Reflector vs Split-Screen Reflector

http://i258.photobucket.com/albums/hh262/autofils/SBayGray-Hoverman2pp.gif

(b) SBGH Split-Screen vs CM-4221


http://i258.photobucket.com/albums/hh262/autofils/SBGHSplitScreenvsCM4221-2.gif
Double Bay Polar Plots

(a) DBGH 11 pair Collinear Rod Reflector vs 30x75 Screen

http://i258.photobucket.com/albums/hh262/autofils/DBayGray-Hoverman2pp.gif

(b) DBGH 30x75 Screen vs CM-4228


http://i258.photobucket.com/albums/hh262/autofils/DBGHvsCM4228_2.gif

Thank you for your polar plots Autofils, very, very interesting...

Based on these plots for the DBGH, it indicates that there would be a significant gain at the back of the antenna for the refector rods: this is very interesting for me considering my location between Montréal and US.

I think that I will end up building the rod reflectors DBGH because in my situation, it would allow me to catch the Montreal stations with better gain, with little impact on the reception strength for the US stations. Does my reflexion make sense ?

But for the testing purpose of the antenna and report, I'll experiment with the full-splitted-grid first.

DjiPi.

DjiPi, what if you build it with the full reflector to test if the back door gives you the Montreal signals, then switch to the rods if its not good enough? :)

My own situation is very similar, with Mt. Seymour and SeaTac stations being almost 180deg apart, but I'm quite close to the Canadian ones so I might test to see if the full reflector's back door is generous enough.

DjiPi, what if you build it with the full reflector to test if the back door gives you the Montreal signals, then switch to the rods if its not good enough?

That is effectively my plan as I said before (well maybe not that clear ;)), but I'm curious about the rod reflectors gain pattern. My interrogation would imply the possibility of multipaths/problems, etc. resulting from the gain at the back. Wouldn't it be like building 2 CM4221 and putting them back to back ?

This is a correction for previous SBGH polar plots that I have posted on this forum.

On the AVS forum " New DIY antenna design from Slashdot", vman41 asked the following question...BTW, can someone explain to me why the polar plots from this NEC2 modeler are often asymetric? One would presume the geometry input for the antenna was idealized, so getting oddball lobes on side of the axis that aren't reflected on the other side is disconcerting.I had a quick look into this and found that it occurred only with the SBGH over the freq range 600 to 700Mhz, but I did not have any explanation. I sought help from old sparks, and he quickly identified the problem.

The SBGH modeling files did not have the voltage source located in the exact middle of the wire tag. (The DBGH files were correct and did not show the asymmetrical lobe)

Here is the link to the AVS forum for the explanation of the asymmetrical lobe, including my red face ...
I want to thank vman41 for spotting this and old sparks for the solution..

And here are the corrected SBGH polar plots.
( Stampeder: These should be added to the GPL web site with a short explanation)

http://i258.photobucket.com/albums/hh262/autofils/CorrectedSBGHRods-pp.gif
http://i258.photobucket.com/albums/hh262/autofils/CorrectedSBGHSplit_Screen-pp.gif

Hello!

Recently, I've been talking about antennas on AVSForum, where holl_ands posted a link over here about the time autofils started posting on AVSForum' Hoverman thread. I made a CM-4221 clone before discovering threads at AVS on that very topic, and find what you're discussing here very germaine.

One of the things I'm trying to do is understand what's driving antenna performance. Yagi's made sense - dipole, directors, reflector, H-polarization - so I made one and used it for a year or so. Eventually I read enough on AVS that I decided to reverse engineer a 4221 from a picture and overall dimensions. There was an error on the web page I used, so I scaled things up an inch. I found the error later, when I pulled dimensions from the GIF file Ken Nist used to calculate the gain curve on HDTV Primer.

My understanding of antennas is that as conductors in space, the E-field of EM radiation parallel to the conductor will set up a charge oscillation in the conductor, and a voltage oscillation at the ends. This will peak at wavelengths that can form standing waves in the conductor - the longest wavelength is 2x the element size, right?

Well the CM-4221 is based on an 8" dimension. The bays spacing and the wisker length are both 8", and the antenna is 4" deep, so reflector spacing is just under 4". The reflector is 20x36, so it extends about 2" beyond the elements. Here's what's interesting:
- the CM4221 peaks at Ch 66, a 7.5" 1/2 wave
- the 8" elements, when spread as shown, form a triangle 7.5" tall.
- the reflector is a little less than 4" away, close to 1/4 wave

My antenna may have been an accident, but these relationships aren't, and one can use this to change the channel gain curve. The next post after the predictions were some test antenna results that track the predictions. What's that mean here?

The Hoverman is a related antenna, and the dimensions fairly well predict the performance. A 10" bow-tie would have peak gain at channel 40, a 9" (mine) at ch 51. SBGH is coming in at Ch 49 with 10" bay spacing. The inner elements are a 5" horizontal spread, the top and bottom 10.5".

Since gain slopes slowly at channels below peak, all of these antennas, including the CM-4221, have good gain in the Ch 14-52 range, but the ones that peak in that range will naturally do better. That's one source of the SBGH design superiority over the CM-4221, but it's design intent; SBGH doesn't have much gain left at Ch 66; neither does mine.

The other thing I see is the work on reflector design - it makes a difference and you all see it in the modeling autofils has done. An advantage of arrays is that they have a large aperture - they grab signal over a large cross-section. It's one reason may people recommend the CM-4228 over high end Yagi's.

The SBGH optimum has a 30x40" screen. The element window is 24x30". By comparison, a CM4221 reflector is 20x36", element window 16.5x22". You see it clearly in the DBGH vs 4228 comparison. The added 60% reflector area makes part of the difference; design goals and poor bay ganging explain the rest.

My point is that you can't beat the laws of Physics; the Hoverman idea isn't magical, has no super powers, and will perform just as well as similarly designed, high performance antenna arrays.

Now FLIP THAT OVER to my perspective - more things to try.

I like to make things. Bow-tie antennas are a current interest. This is a new kind of bow-tie, so it's a new toy.

This whole Hoverman topic has resulted in improved performance, but it's also created a lot of discussion. The more we talk about this stuff, on this technically correct, empirical verified basis, the more we advance the art.

Have fun,
Frank

I'm sorry it couldn't have been more quantitative, but I don't really have anything to measure the signal.

As I was building it and the lack of workmanship was becoming apparent, I was wondering what the tolerance of the measurements were. Once I'd fastened the array on I tried it out (so this was before the reflector pieces), and it barely worked. Fortunately I persevered and it all worked on completion.


As for the eye thing, it was damn close. I can't believe how stupid I was. So can't agree more on your safety warning. The antenna is a big spikey thing that gets caught on everything as you move it around, and 'just being careful' isn't enough...

Nik,

Thanks for noting the DBGH response with the array-only.
That certainly "jives" with the modeling results which show the array-only with significant lower gain.
Here is the DBGH net gain plots, have a look.
http://www.digitalhome.ca/forum/showpost.php?p=707405&postcount=61

So even though most HDTV devices don't have a "signal-meter", just keen observational information is quite meaningful.
...thanks Nik

On Aiming Antenna for best HDTV Reception
=================================
I ran across an article that suggests a cheap way to help to aim an antenna for HDTV signals. As you probably know, the cliff-effect with digital signals, makes aiming your antenna a lot more difficult than aiming for analog channels. Well this article has a rally neat idea.... using a variable attenuator to reduce your signal to near the cliff-effect... and that way you aim for the optimum direction...have a read... its near the bottom of the page.
http://www.hometoys.com/htinews/jun07/articles/training/antennas.htm

Welcome to AVSers coming here - I've been an AVS member myself for quite a few years.

Digital Home is a business, and it relies on visits and memberships. AVS is a great forum, but we have an understandable bias towards this site. ;) While this is a Canadian site, the laws of antenna physics and electromagnetism are universal so everyone is welcome here!

I just want to encourage everyone who wants to discuss the Gray-Hoverman to use this thread since it is now linked from thousands of places around the Internet. With all due respect, DHC members like Autofils put a lot of effort into the project, and I put a lot of effort into getting it publicized, so your use of this thread is greatly appreciated. Home is where the heart is - right here. :)

Also check out our Build-It-Yourself thread in this forum, in which antenna design is also an encouraged topic.

Again, welcome and please help this site to remain a successful business in the way that AVS is.

Cheers, stampeder

AutoFis's latest modeling suggests that a deeper spacing between the zig-zags and the reflector rods could give us (here in Ottawa) an extra dB or so for our local channels. Fortunately this is an easy retrofit to perform.

Oops.. I misunderstood Mr.AutoFis on that one. The spacing back to the rods is fine as-is. His suggestion was actually to lengthen the horizontal tail ends of the zig-zag elements by 44mm, which shifts the response curve rather sharply down to the lower end of the UHF television band, adding perhaps another 0.4dB or so across them.

But at the expense of cutting off most everything above UHF 54 or so. I had forgotten that we have channels 60,65 locally, which do have occasional good programming on them, so I may forgo the extra 0.4dB.

But it's easy enough to pigtail some extra length onto those for experimentation, so in the end we'll likely try it both ways.

(and thanks to Stampeder for rescuing my photo posting from the mod queue!)

Cheers

I have decided to stop the "my NerdClub Ads" and put a stake in the ground to see if that gets the ball rolling.
This "ground stake" will use MediaFire and will have a simple membership criteria. If it does get established, then I would propose that it becomes a new forum and open to all with no membership rules.

You may not agree with this step; that's fine...but if I don't take some concrete first steps, this will never get started.
If it turns out that the NerdClub is a bad or non-workable idea; that's also fine with me.

The download links for MediaFire are:

NerdClub ReadMe:
http://www.mediafire.com/?uc1s9jtenj9

SBGH_GPLDesign Array_Only package
http://www.mediafire.com/?z5i1jbn4nyz

Regards
...Autofils