Original Gray-Hoverman Generation 2 Thread (Now Closed)

[DTV Student]

Hi Folks,

300ohm:
... What Ive found is that Arie Voors' front end to the Fortran 77 Lawrence Livermore Labs NEC engine behaves with common, similar features like most other Windows programs. For example, Ctrl-A to select all in Geometry Editor. It just takes some time and practice finding out where everything you want to do is. The more you use it, the more youll love it. Experiment ...

Yes indeed!!! Experiment! Experiment! Experiment! Arie's help files only hint at the power built into the Geometric Editor.

To avoid becoming totally frustrated with the Geometric Editor, keep it simple in the beginning. I think the best way to learn what this editor can do is by first drawing very simple antennas, like a dipole design (rabbit ears). Learn how to manipulate objects in a very simple design before attempting more complex designs. The idea is to build your skill set slowly. It does take time to get the hang of it. Become accustomed to switching back and forth between the 3D and 2D modes and watch that "Keep Connected" check box when moving a wire. Sometimes you need it and sometimes you don't. LOL... Learn how to undo your last edit.

The Ctrl-A method that 300ohm mentioned above selects all of the objects in a design for easy moving, rotating, etc, etc, or even deleting. Very often you might need to select a group of wires, but not all of the wires in a design. For example, you might want to move a complicated reflector structure a little closer to the active driver assembly. This is possible in the 2D modes by indicating a rectanglar region that includes both ends of each wire to combined into a larger group of elements. While in the 2D select object mode, left click, then drag and drop diagonally over the group of wires to be select as a temporary unit. A box will appear to show you what you are including. When you let up on the mouse button all of the wires with both end inside the retangular area will be selected and will turn red. One of the wires will appear thicker than the others. This is the "handle" that can be used to move the group by dragging and dropping it.

Good Luck,
DTV Student

[300ohm]

I have a feeling that Stampeder will probably move these non-GH specific modeling posts to the modeling thread, but here it goes anyway.

Last night I stumbled upon by accident a little known/unintended feature of 4nec2 that makes it very useful for importing data into a spreadsheet. The feature is really designed/intended for the wGNUplot program, but the exported data it produces works fine for other programs too. Once the data is into a spreadsheet, the amount of calculation and graphing that can be done on it is endless, heh.

After running the frequency sweep as normal, the gain, swr and impedence charts are produced. There is an option of plot then available on that screen. Click on it and choose what you want outputted to a plot file. Since I dont have wGNUplot.exe installed, it gives an error message, but it still creates a text file called plot.txt. It lies about where it exports the plot.txt file in the message, the actual file is in the C:\4nec2\plot directory. This must be renamed before running another plot option or it will be over-written with the new data. A simple copy and paste from the plot.txt file to my Open Office Calc spreadsheet program formats the data as pretty as you please, complete with column headings, heh. Should work just as fine in MS Excel or any other spreadsheet program. This feature is very very useful and saves many hours for calculating net gain if you have a lot of frequencies to deal with.

I was playing around with the .out file and came upon the plot.txt file and the titles and the data rang a bell. The 4nec2 program justs get better and better the more its dug into.

[ctgottapee]

so i'm having an impossible time finding 1/4-3/8" aluminum tubing. heck i'm having a hard time finding aluminum tube at all, or tubing that is less than 1/2".

autofills> would you care to run a standard GH10 model for me with 6guage copper elements and 1/2" aluminum tubing? apart from your new bent mesh design, it seems to be the most practical build parameters and i want to make sure i'm not building a dud.

while i could attempt to do so myself, i would take way more of your time peppering you with questions on how to run it, and this seems relevant to everyone who wants to build the colinear version.

the only smaller diameter tubing is the coiled stuff, generally expensive copper, and straightening 23 feet of it seems an impracticality...

i also looked into trying the composite lumber (plastic) for the base support, but the solid stuff is just way too heavy, it weights more than the mast would, and other variations are hollow and not very sturdy making them uselss really.

[Autofils]

Originally Posted by ctgottapee

autofills> would you care to run a standard GH10 model for me with 6guage copper elements and 1/2" aluminum tubing? apart from your new bent mesh design, it seems to be the most practical build parameters and i want to make sure i'm not building a dud.

ctgottapee,

If cost is an issue, or you just want to make a low-cost trial version, just use simple materials.
If it is for an attic install, the cheapest build would be using tin-foil on a cardboard sheet. For outdoor, grab a bunch of coat-hangers.

The original GH10 collinear model was with 7mm (~5/16") Alu diameter elements.

Because I had some concerns that some folks might get carried away and spend far too much money, I presented GH10 collinear model results (post #145) for Coat-hanger wire, 1/8" and 5/16" wire diameters to provide guidance in the effect of using different materials. The 4nec2 starts reporting warnings if the wire diam is set to 1/2" or more.

Post 145
http://www.digitalhome.ca/forum/show...&postcount=145


The excellent design and modeling of the curved screen design (post #191-192) was done by 300ohm, and used #10 copper wire.

Post 191
http://www.digitalhome.ca/forum/show...&postcount=191


If you could post a saved image of reception prediction by tvfool for your location, and indicate the stations that are "must-haves", 300ohm and I, can probably better serve you.

You might even find that a GH4 collinear would be quite adequate if all your "must-haves" are Rx(dBm) of - 85 or less ( i.e. less negative)

[ctgottapee]

thx for the reply autofills

i've already got the elements made...
-6guage bare copper ground wire seems to be the most practical way to get close to the 7mm specified in the design. bending soft copper tube (first you have to straighten the coil it comes in) doesn't seem practical, and lower guages of wire i've found do not come in solid cores, they are multiple lower guages combined into one jacket. i'm sure some could be ordered but i'm sure it would be pricey.
-any metallic tubing under 1/2" seems fairly nonexistent commercially as it doesn't have much use, apart from coiled copper tubing, which would be a nightmare to straighten 25 feet of, and is no longer so cheap if one did. so that is why i'm asking about using 1/2" tube for the reflectors as that is the smallest size of any type of metallic tube the homestores stock.

so i was asking if a modeling run would help verify this could work ok, or if i need to pursue ordering smaller tube somehow

i'm wanting the GH10 version as i don't think the mesh designs would last a year here, elements and wind are brutal.


and here is my tvfool numbers i posted before, i need all the help i can get
i've got a good range of test frequencies to test with, all dtv channels (and i don't expect vhf, but can test it)
[first number is compass degrees, second is miles to trans mitter, third is tvfool dbm]
vhf 09 191^/49m -096.1dbm 186-192mhz *adjacent ch
vhf 10 330^/45m -068.1dbm
uhf 18 199^/57m -107.3dbm 494-500mhz
uhf 22 199^/57m -109.5dbm 518-524mhz
uhf 26 133^/44m -095.4dbm
uhf 28 259^/39m -066.1dbm 554-560mhz
uhf 30 264^/54m -108.4dbm 566-572mhz
uhf 39 265^/55m -116.6dbm 620-626
uhf 41 148^/56m -097.5dbm 632-638mhz
uhf 42 219^/81m -123.2dbm *adjacent ch interference
uhf 44 217^/81m -125.8dbm
uhf 46 264^/56m -129.2dbm 500-506mhz (moves dt19)
uhf 48 178^/44m -119.2dmb 674-680mhz
uhf 51 237^/30m -098.6dbm
uhf 57 264^/54m -113.7dbm 536-542mhz (moves dt25)

[jdogmoney]

so that is why i'm asking about using 1/2" tube for the reflectors as that is the smallest size of any type of metallic tube the homestores stock.

I am acutually using 1/2" tubing for the reflectors as 300ohm said it would be fine (a couple pages back). It's the smallest I could find without looking through the phonebook for metal stocking companies. As I understand it, the only concerns about going as big as 1/2" are weight and windload, not performance.

BTW, the cheapest tubing I found was 1/2" metal conduit on the electrical isle at Lowe's. I bought 30' of it for about $8 and had 5' leftover.

[Autofils]

Originally Posted by ctgottapee

thx for the reply autofills

i've already got the elements made...
-6guage bare copper ground wire
-metallic 1/2"

i'm wanting the GH10 version as i don't think the mesh designs would last a year here, elements and wind are brutal.

ctgottapee,

Since you've got the elements made, here are the model results for the GH10 collinear rod reflectors (h=0.163), for array element wire diam #6, rod reflector wire diam =0.5 inches, and wire conductivity was set as Copper.
(The wire radius of VSource Tag1 was adjusted for AGT test result of 1.0(-0dB) at 585 Mhz.)

If you run a segment check, you will see warnings on all rod reflectors due to the 1/2" wire diam. This really means the model segmentation should be adjusted to eliminate those warnings. I did not spend any time trying to correct and the results are presented with those warnings.
( For 4nec2 students that wish to increase their modeling skills, the nec file and the warnings are attached)

Raw Gain




SWR



Tabular Data (dBi)
========================
Freq ..... Raw gain .....Net Gain
========================
470 ...... 13.51 ......... 12.74
510 ...... 15.27 ......... 14.92
550 ...... 15.56 ......... 15.25
590 ...... 15.93 ......... 15.83
630 ...... 15.34 ......... 15.04
========================



( For 4nec2 students.....nec file and segmentation warnings )

1. NEC FILE

---------------------------------------------------------------
CM SBGH Gen2 GH10 Collinear Rod reflectors
CM Modified for ctgottapee as folows:
CM - Wire diam Array = #6 (radius = 0.0020574meteres) VSource Tag1 adjusted for AGT 1.0(0dB)
CM - Wire diam Refelctor Rods = 0.5inch (radius = 0.00635 metes)
CM - Wire Conducivity in FR card set for Copper
CE
SY y=0.095 'FeedGap
SY x=0.094 'Length of Array Horizontal Stub
SY h=0.163 'Length of Array diagonal = (Sqrt(2) * h) = 0.231
SY s=0.094 'Reflector Spacing
SY g=0.05 'Reflector Gap
SY R1=0.227 'Collinear Rod Reflector Lengths
SY R2=0.370
SY R3=0.495
SY R4=0.268
SY R5=0.392
GW 1 5 0 0 0 0 y 0 0.00157
GW 2 10 0 y 0 0 y+h h 0.0020574
GW 3 10 0 y+h h 0 y 2h 0.0020574
GW 4 10 0 y 2h 0 y+h 3h 0.0020574
GW 5 10 0 y+h 3h 0 y+h+x 3h 0.0020574
GW 6 10 0 y 0 0 y+h -h 0.0020574
GW 7 10 0 y+h -h 0 y -2h 0.0020574
GW 8 10 0 y -2h 0 y+h -3h 0.0020574
GW 9 10 0 y+h -3h 0 y+h+x -3h 0.0020574
GW 10 10 0 0 0 0 -h h 0.0020574
GW 11 10 0 -h h 0 0 2h 0.0020574
GW 12 10 0 0 2h 0 -h 3h 0.0020574
GW 13 10 0 -h 3h 0 -(h+x) 3h 0.0020574
GW 14 10 0 0 0 0 -h -h 0.0020574
GW 15 10 0 -h -h 0 0 -2h 0.0020574
GW 16 10 0 0 -2h 0 -h -3h 0.0020574
GW 17 10 0 -h -3h 0 -(h+x) -3h 0.0020574
GW 18 18 -s (R1+y/2+g/2) -4.5h -s y/2+g/2 -4.5h 0.00635
GW 19 18 -s -(R1-y/2+g/2) -4.5h -s y/2-g/2 -4.5h 0.00635
GW 22 18 -s (R2+y/2+g/2) -3.5h -s y/2+g/2 -3.5h 0.00635
GW 23 18 -s -(R2-y/2+g/2) -3.5h -s y/2-g/2 -3.5h 0.00635
GW 30 18 -s (R3+y/2+g/2) -2.5h -s y/2+g/2 -2.5h 0.00635
GW 31 18 -s -(R3-y/2+g/2) -2.5h -s y/2-g/2 -2.5h 0.00635
GW 38 18 -s (R4+y/2+g/2) -1.5h -s y/2+g/2 -1.5h 0.00635
GW 39 18 -s -(R4-y/2+g/2) -1.5h -s y/2-g/2 -1.5h 0.00635
GW 46 18 -s (R5+y/2+g/2) -0.5h -s y/2+g/2 -0.5h 0.00635
GW 47 18 -s -(R5-y/2+g/2) -0.5h -s y/2-g/2 -0.5h 0.00635
GW 52 18 -s (R5+y/2+g/2) 0.5h -s y/2+g/2 0.5h 0.00635
GW 53 18 -s -(R5-y/2+g/2) 0.5h -s y/2-g/2 0.5h 0.00635
GW 60 18 -s (R4+y/2+g/2) 1.5h -s y/2+g/2 1.5h 0.00635
GW 61 18 -s -(R4-y/2+g/2) 1.5h -s y/2-g/2 1.5h 0.00635
GW 68 18 -s (R3+y/2+g/2) 2.5h -s y/2+g/2 2.5h 0.00635
GW 69 18 -s -(R3-y/2+g/2) 2.5h -s y/2-g/2 2.5h 0.00635
GW 76 18 -s (R2+y/2+g/2) 3.5h -s y/2+g/2 3.5h 0.00635
GW 77 18 -s -(R2-y/2+g/2) 3.5h -s y/2-g/2 3.5h 0.00635
GW 80 18 -s (R1+y/2+g/2) 4.5h -s y/2+g/2 4.5h 0.00635
GW 81 18 -s -(R1-y/2+g/2) 4.5h -s y/2-g/2 4.5h 0.00635
GE 0
LD 5 0 0 0 58000000
GN -1
EK
EX 0 1 3 00 1 0
FR 0 0 0 0 585 0
EN

-----------------------------------------------------------------------

2. Segmentation Warnings


GH10 - Modified for ctgottapee.nec wavelength=0.512 mtr.

Warn.: Wire 18 (tag 18), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 18 (tag 18), seg-len (0.013) below 2 * radius (0.013)
Warn.: Wire 19 (tag 19), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 19 (tag 19), seg-len (0.013) below 2 * radius (0.013)
Warn.: Wire 20 (tag 22), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 21 (tag 23), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 22 (tag 30), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 23 (tag 31), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 24 (tag 38), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 25 (tag 39), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 26 (tag 46), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 27 (tag 47), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 28 (tag 52), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 29 (tag 53), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 30 (tag 60), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 31 (tag 61), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 32 (tag 68), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 33 (tag 69), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 34 (tag 76), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 35 (tag 77), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 36 (tag 80), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 36 (tag 80), seg-len (0.013) below 2 * radius (0.013)
Warn.: Wire 37 (tag 81), radius (6.e-3) above wlen/100 (5.e-3)
Warn.: Wire 37 (tag 81), seg-len (0.013) below 2 * radius (0.013)


--------------------------------------------------------------------

[Autofils]

Originally Posted by ctgottapee

thx for the reply autofills

i've already got the elements made...
.................
and here is my tvfool numbers .....
uhf 18 199^/57m -107.3dbm 494-500mhz
uhf 22 199^/57m -109.5dbm 518-524mhz
.......
uhf 30 264^/54m -108.4dbm 566-572mhz
uhf 39 265^/55m -116.6dbm 620-626
......
uhf 42 219^/81m -123.2dbm *adjacent ch interference
uhf 44 217^/81m -125.8dbm
uhf 46 264^/56m -129.2dbm 500-506mhz (moves dt19)
uhf 48 178^/44m -119.2dmb 674-680mhz
........
uhf 57 264^/54m -113.7dbm 536-542mhz (moves dt25)

ctgottapee and all other antenna buffs,

Looks like you will be ready to test out your GH10 build soon.
Your TVfool data has a number of UHF stations that are in the range of -100dBm to -130 dBm.
These are just the right environment for real world tests on hi-gain antennas.

Using the "cliff effect" characteristic of OTA digital signals, you can easily make quite meaningful measurements by simply inserting attenuation at your TV set-end of your down-lead, until the digital signal disappears or becomes a frozen pic.

Knowing the predicted Rx(dBm) along with a terrain description at your location and the attenuation required to reach the "digital-cliff" ( with or without a preamp), would provide very useful reception data.

For those antenna buffs that build more than one antenna, this measurement technique would give very meaningful comparison data between antennas.

And for those members that are currently using highly rated commercial antennas, like the CM4221 or CM-4228, this measurement technique would provide excellent comparative data.

All you need for this type of test, is a set of 75 ohm attenuator pads.
Here is a link that sells 75 ohm in-line attenuator pads...
A set of 5 pads (1, 1, 3, 6, and 10 dB) provides up to 21 dB attenuation in 1 db increments.

http://www.mjsales.net/items.asp?Fam...=265&Cat2ID=93

If members of this forum could provide this type of measurement data we could build up a really significant data base, not only for the GH, but all other DIY antenna types and also those highly rated commercial types.

( This is only a suggestion,.... and I approve this message ...)

[300ohm]

The mast clamp is OK because its relatively small and its behind the reflectors and elements.

CM spartan 3 preamp with 300ohm input, mounting it right behind the center of the antenna and running a 6inch piece of 300ohm wire directly into it.

Thats what I use, the CM 0264 Spartan 3 preamp and it works great. You need to put a drip loop in it so youll be using more than 6 inches.

oh and why is 1/2" too big for modeling? is it i just a limit of the software?

Nope, not a software limitation. When the diameter of a wire is greater than a certain fraction of the wavelength, it starts to behave like a different kind of antenna. For example, if you were to make a UHF dipole with say 4 inch diameter tubing, the antenna would exhibit similiar behavior to that of a single bay bowtie. The software warns against that. Of course, the 4 inch diameter dipole would have tons more wind load than a common bow tie, one reason its not a popular design, heh.

Many Australian UHF antennas use 7/16 inch and 1/2 inch aluminum tubing because they need to make them parrot proof.

[mlord]

Originally Posted by ctgottapee

one final question. was thinking about using the CM spartan 3 preamp with 300ohm input, mounting it right behind the center of the antenna and running a 6inch piece of 300ohm wire directly into it. my thought is to reduce the balun loss, and coax loss. not quite sure if it equals out in comparison to using a bit better CM 7777.

Sounds good to me -- I'm going to try mine out initially with a simple Spartan 0264 pre-amp, and I expect it to have a better noise figure than a 7777+balun would.

If it looks good with that amp, then I may move it up the tower and connect to the Really Good Amp that my PR-8800 pair currently use.

No worries about the mast clamps, mate -- stuff like that is pretty minor and usually behind the primary reflectors anyway. On the ones here, it will likely be both behind and below the bottommost reflector rod.

Cheers

[Autofils]

Originally Posted by ctgottapee

i'm going to be using my PC and tuner card, so i'll actually be able to report the measured values. the card has a signal tester that reports signal strength in db and also the error correction type number most see on common displays. i can't be sure its methodology for determining this number, but it should be as accurate as the db pads.
.......
oh and why is 1/2" too big for modeling? is it i just a limit of the software?

ctgottapee,

What PC tuner card do you have? What is your ASTC tuner's sensitivity? ( I believe the FCC models assumes -83 dBm.)

If your tuner can measure signal strength of OTA digital signals,...that's great!! Knowing your tuner's sensitivity would give the overall antenna gain + tuner margin.

Since most folks don't have a tuner that actually measures signal strength, I suggested using attenuator pads to find the cliff-effect. This gives an overall measurement of antenna gain + tuner margin, and it would be a consistent measurement technique regardless of location and tuner capabilities, and would be useful for all members.
Of course, it would cost members some money and their time...and that's generally a big stumbling block for most folks.


Re 1/2" wire diameter:
Basically 300ohm's comment is correct. There is no software limitation. When you run segmentation checks on your nec model, there should be no errors and no warnings. Errors obviously must be corrected. Warnings re segmentation limits, simply means that the accuracy of the model is affected.

For those who have an interest in the detail, have a read of the NEC2 User manual pt3 at this link...
www.nec2.org/other/nec2prt3.pdf

[jdogmoney]

Originally Posted by John SE Texas

Hi, this is my first post. I have been reading Digital Home for about 3 months and thought it is about time for me to get my feet wet with the Gray Hoverman antenna.

Welcome, it's good to have you join John. I hope you find as much helpful information on the board as I have.

It looks like everybody's getting close to finishing their builds, it's going to be nice having the ability to compare data.

[300ohm]

..I have never done so much with so little..I live near zip 77626 and have at least 7 stations within 20 miles. At about 15 feet high...the antenna pulls them all in with near green to high green.

You have an unusual situation Ive never seen before. You have 2 channel 27's with about equal dbm located only about 65 degrees apart. Do you have ghosting on that channel ?

[ctgottapee]

one of my cards is the new dvico fusion7
http://www.fusionhdtv.co.kr/ENG/prod...TV7RTGold.aspx

and i see what you mean by the pads giving an equal measurement amongst us all, although reception conditions at our locations are varied enough that a signal measurement should suffice

many of the later cards have tools that measure signal strength, although i'm not sure how it is measured

ctgottapee,

What PC tuner card do you have? What is your ASTC tuner's sensitivity? ( I believe the FCC models assumes -83 dBm.)

If your tuner can measure signal strength of OTA digital signals,...that's great!! Knowing your tuner's sensitivity would give the overall antenna gain + tuner margin.

Since most folks don't have a tuner that actually measures signal strength, I suggested using attenuator pads to find the cliff-effect. This gives an overall measurement of antenna gain + tuner margin, and it would be a consistent measurement technique regardless of location and tuner capabilities, and would be useful for all members.
Of course, it would cost members some money and their time...and that's generally a big stumbling block for most folks.

[ctgottapee]

yes pci, there is a dual tuner pci-e version but it only has one input so it splits it for a 3db drop

pci is fine for 1080i or more. many have multiple cap cards on the pci bus. heck i know someone with 4 usb HD tuners.

which card is best depends upon many factors/tradeoffs. they all tend to have issues.

i have a MyHD, a pinnacle USB stick, and the fusion7
also had a fusion3 that was crap

one thing for sure, the ATI card is the absolute worst of the bunch, software and reception wise, so you can't go wrong replacing that thing.

the dvico fusion7 is the only card so far with the latest HD samsung 5th gen demod and new silicon tuner
[quote]Hmmm, that card is pci and not pci-e, correct ?. 1080i broadcasts will overpower the pci bus bandwidth. What other ATSC pc cards do you have, and which one do you like the most and why ? I have an old ATi NTSC pci TV card that Ill need to replace.[quote]