Canadian TV, Computing and Home Theatre Forums banner

161 - 180 of 361 Posts

·
Registered
Joined
·
2,421 Posts
Discussion Starter #161
im looking for a very simple to build #13 yagi design, up to 6 ft long max, with very high f/b, i already have 3/8" aluminum elements, and 8 ft long 3/4" square aluminum boom....

not too much to ask for ;)
 

·
Registered
Joined
·
3,396 Posts
How much forward Gain can we trade off for F/B Ratio??
Is 8 dBi enough...or 10 dBi...maybe 12 dBi...or more???

Do you really mean F/B (directly astern) or F/R (120-deg included angle)???
If we maximize the F/B, there will be a very deep null with a very large F/B,
but it will very narrow. Perhaps we should define a fairly small F/R included
angle which would include your problem station...

I didn't see much of anything in my archive that met the 6-ft max criteria....
Why not allow for the full 8-ft boom length???

I just restarted a Ch13 9-El FD Yagi with a 12 dBi Gain and 30 dB F/R Ratio
targets. I think my MLE algorithm is favoring Mismatch Loss a little TOO
much, so I "tweaked" it a bit to reduce ML and increase F/R "importance".

PS: Each run takes perhaps 12-hours and I'll have to readjust the
targets a few times to get a feel for what is REALIZABLE....
So might take all weekend....and then there is the HIGH Gain target
and the FULL BAND target....weeks later we'll know what's possible....

FYI: Better Ch13 5-EL FD Yagi's are ~40-in, Gain~9.5, F/B=24+ dB, good SWR.
Which outperforms K6STI style Swept Back 5-El Yagi....
Must redo the run spoiled by the power failure....and analyze a bunch of data
before I pick out the "best" for posting here...

I have only looked at FULL-BAND, Gain maximized, 7-El FD Yagi's so far....
 

·
Premium Member
Joined
·
1,121 Posts
Which outperforms K6STI style Swept Back 5-El Yagi....
Indeed; broadbanding for 7-13 seems to imply a sacrifice of gain VS a single-channel yagi for 13.

If stagger stacking works for him now that he has done proper grounding, maybe he can accept a little less F/B or F/R inherent in each antenna for more gain that he needs badly for WVNY.
 

·
Registered
Joined
·
2,421 Posts
Discussion Starter #164
i'm not sure what F/B is required, since it appears co-channel #13 is stronger than previously thought, or wanted #13 is weaker,

if i don't stack, i would guesstimate F/B in the order of 30+db, co-channel is at 171 deg., and adjacent #12 is at 65 deg., (from wanted #13),

i'm going simple here with straight elements only on top of boom, so i ruled out FD (mistake ???)

will be solely for #13, gain in 13.5+ dbi range, i guess i have no problem using the whole length of boom

rabbit73, the problem is ultimately i wont have the space to properly stack, also i dont have full confidence in the 5 element i modeled (i tried to optimize K6STI 5-element for #13) and built - i thought i would try another angle now ;)
 

·
Premium Member
Joined
·
1,121 Posts
With your co-channel interference at 171 deg. relative, that looks promising because a filter can't help you there.
i dont have full confidence in the 5 element i modeled
Neither do I, because you don't want a 7-13 antenna only for 13
.....wont have the space to properly stack.....i thought i would try another angle now
Understood---go for max gain and F/B with a single yagi.

I keep thinking of a WA5VJB "cheap yagi" for 13.

It uses direct 75 ohm feed with an unusual driven element that has 1/2 of a folded dipole on one side of the driven element and a straight element on the other. The theory is that without the parasitic elements the feed point impedance is 150 ohms, but when the parasitic elements are added, it comes down to the coax impedance.

There are two versions, one for 50 ohm feed, and the other for 75 ohm feed. The 75 ohm match is made by using slightly wider spacing between the elements.

It would not survive your weather conditions, but it would be interesting and inexpensive to build one for a test.
 

·
Registered
Joined
·
551 Posts
rabbit73: Is this what you are looking for? It is 300 ohms not 75.
CM Experimental Antenna Design
CM Yagi Ch.13
CE
GW 1 1 0 0 1.25 0 0 -1.25 0.24
GW 2 5 0 -12.8 1.25 0 0 1.25 0.156
GW 3 5 0 0 1.25 0 12.8 1.25 0.156
GW 4 1 0 12.8 1.25 0 12.8 -1.25 0.156
GW 5 5 0 12.8 -1.25 0 0 -1.25 0.156
GW 6 13 -9 -13.65 0 -9 13.65 0 0.156
GW 7 11 6.14 -12 0 6.14 12 0 0.156
GW 8 11 18 -11.5 0 18 11.5 0 0.156
GS 0 0 0.0254
GE 0
EK
LD 5 0 0 0 5.8e7 0
EX 0 1 1 0 1 0
GN -1
FR 0 1 0 0 213 0
XQ
EN
 

·
Registered
Joined
·
296 Posts




Code:
Channel 13 Yagi
Free Space Symmetric
210 213 216 MHz
8 6061-T6 wires, inches
x1 = 0
x2 = 6.514553
x3 = 10.47846
x4 = 20.53799
x5 = 38.41845
x6 = 58.12122
x7 = 78.43935
x8 = 95
y1 = 13.965
y2 = 14.10225
y3 = 12.6566
y4 = 12.13685
y5 = 11.69517
y6 = 11.52444
y7 = 11.62053
y8 = 11.01056
1  x1  -y1  0   x1  y1  0    0.3750
1  x2  -y2  0   x2  y2  0    0.3750
1  x3  -y3  0   x3  y3  0    0.3750
1  x4  -y4  0   x4  y4  0    0.3750
1  x5  -y5  0   x5  y5  0    0.3750
1  x6  -y6  0   x6  y6  0    0.3750
1  x7  -y7  0   x7  y7  0    0.3750
1  x8  -y8  0   x8  y8  0    0.3750
1 source
c = 9.139196
Wire 2, center 1 0 c pF
I ran this off pretty quickly. I think it could be further optimized, but it's a good starting point. This isn't in NEC format, but I think you can extract the dimensions.

The feed impedance is 75 ohms. Solder a 9.1-pF capacitor across the feedpoint (10 pF will work fine). This is the lowpass version of a hairpin match. Install two Fair-Rite 2643625202 ferrite sleeves, available from Mouser, at the feedpoint. I'd install two more sleeves a quarterwave down the coax as well.

Any design with such low backlobes is quite sensitive. Cut the elements accurately, keep them away from other conductors, mount the antenna in the clear, and route the coax in the element midplane.

I see that I forgot to use dBi. Sorry--I think in dBd. You'll have to translate.

Brian
 

·
Registered
Joined
·
2,421 Posts
Discussion Starter #169
thank you, can this be redesigned for standard 4:1 balun, which is what im used to and have available

thanks again
 

·
Registered
Joined
·
296 Posts
The antenna was specifically designed for 75-ohm feed. This eliminates the loss of a 75:300-ohm balun (and unwanted signal pickup, in the case of a halfwave coaxial balun). It avoids the complexity of a folded dipole driven element, including sharp tubing bends or interconnection weathering if you use separate conductors. If you substitute a folded dipole, you'll have to completely redesign the antenna because the pattern is very sensitive.

If for some reason your heart is set on 300-ohm feed, I recommend a horizontal folded dipole. It has some remarkable properties. For more, see

http://ham-radio.com/k6sti/100.htm

A version for High-VHF TV is here:

http://ham-radio.com/k6sti/vhftv.htm

Brian
 

·
Registered
Joined
·
2,421 Posts
Discussion Starter #171
i did a quick 4nec2, with dimensions only, not sure how to do the feed impedence


Code:
CM ch.#13 yagi. K6STI 75 ohm feed, september 2011
CM All elements Aluminum with O.D.=3/8".  SOURCE: Elem#2 middle segment.
CM .75" square aluminum Metal Boom, Insulated top mounts.
CM ALL MEASUREMENTS ARE IN INCHES.  Segments per AutoSegmenter.
CE
SY Gap=1.1		' Vary Gap Size until AvgGainTest = 1.0
' Following are Element positions along X-axis:
SY X1=0.00000		' Reflector Element (At X-axis Origin)
SY X2=6.514		' Center of Driven Dipole Element
SY X3=10.478		' Director #1
SY X4=20.538		' Director #2
SY X5=38.418		' Director #3
SY X6=58.121	'Director #4
SY X7=78.439	'Director #5
SY X8=95	'Director #6
' Following are Element Half Lengths:
SY L1=13.965		' Reflector
SY L2=14.102		' Driven Dipole Element
SY L3=12.656		' Director #1
SY L4=12.137		' Director #2
SY L5=11.695		' Director #3
SY L6=11.524	'Director #4
SY L7=11.620	'Director #5
SY L8=11.01	'Director #6
SY Relem=0.1875		' Radius (inches) for All Elements (O.D. = 3/8 inches)
' First Element is Passive Reflector
GW  1 27  0.0  -L1  0.0   0.0    L1  0.0  Relem
' Eight Passive Directors
GW  2 21  X3   -L3  0.0    X3    L3  0.0  Relem
GW  3 21  X4   -L4  0.0    X4    L4  0.0  Relem
GW  4 19  X5   -L5  0.0    X5    L5  0.0  Relem
GW	5	19	X6	-L6	0.0	X6	L6	0.0	Relem
GW	6	19	X7	-L7	0.0	X7	L7	0.0	Relem
GW	7	19	X8	-L8	0.0	X8	L8	0.0	Relem
' Driven Second Element with SOURCE in middle segment
GW	10	13	X2	-L2	0.0	X2	-Gap	0.0	Relem
GW	11	13	X2	Gap	0.0	X2	L2	0.0	Relem
GW	12	1	X2	-Gap	0.0	X2	Gap	0.0	Relem
GS 0 0 0.0254 ' All in inches.
GE 0
EK 0
LD 5 0 0 0 3.08e7    		' Aluminum Elements (6063-T832)
EX	0	12	1	00	1	0		' SOURCE at middle of Element #2
GN -1
FR,0,1,0,0,213
RP	0	37	73	1511	-180	0	5	5
 

·
Registered
Joined
·
7,360 Posts
not sure how to do the feed impedence
Balm, in 4nec2, impedance is a setting. From the main screen > Settings > Char-impedance. Set to 75 ohms.

Dont forget to set it back to 300 ohms for other antennas, heh.

In your case, I would definately go for the ferrite bead balun setup for the channel 13 antenna.

I'd install two more sleeves a quarterwave down the coax as well.
By sliding the second set of sleeves up and down, couldnt he fine tune the setup a bit ?
 

·
Registered
Joined
·
2,421 Posts
Discussion Starter #177
im going to try a single yagi...i think if its designed properly, it should work enough...
 

·
Registered
Joined
·
296 Posts
> i would like to further optimize it, maybe higher F/B, and lower SWR....


You won't be able to get a significantly better pattern unless you accept lower forward gain. You might be able to raise F/R a dB or two with the same gain. I only did a quick local optimization, but I wouldn't expect much more improvement than that even with a longer global optimization. I don't know whether it's possible to construct a Yagi with such low backlobes anyway. They rely on cancellation of element currents. When you subtract large numbers hoping to get a small one, the slightest error in a number will throw off the cancellation.

The SWR as designed is already extremely low, with insignificant mismatch loss. The gain figures in the pattern plots include conductor and mismatch losses.

Unlike wideband Yagis, the natural impedance of the channel 13 design is about 27 ohms with a resonant driven element. A matching network of some kind is necessary to transform to 75 ohms. It's possible to reoptimize the design for 75 ohms with no matching network, but you'll have to accept lower gain or an inferior pattern. A single shunt capacitor to raise the feedpoint resistance, coupled with a longer driven element to cancel the added reactance, is the simplest matching network I know of (it is equivalent to an L-network). If you use a folded dipole, you'll still need a capacitor unless you sacrifice performance.

The ferrite sleeves provide a broadband series impedance to unwanted current induced on the outside of the coax. The impedance of the specified Fair-Rite sleeves peaks around 100 MHz. It's down a bit at 200 MHz but should still be adequate. Perhaps you can find a more suitable sleeve for the High-VHF TV band. I selected that particular sleeve for use in the FM broadcast band.

It's possible to coil the coax to make a resonant trap that will be much more effective than ferrite sleeves in inhibiting unwanted shield current. I've tried to create one, but it was very awkward physically with my quick test setup and the stiff RG-6 I had on hand. I think this is the simplest, quickest, easiest, and cheapest form of current balun, but the dimensions of the coax coil must be correct.

If you want to stack this Yagi in the H-plane (one above the other for horizontal polarization) you'll need to redesign it. Mutual coupling between two antennas will destroy the pattern. You might be able to get away with E-plane stacking (side by side), but I'm not sure. Try one antenna first and see how well it works.

Brian
 

·
Registered
Joined
·
2,421 Posts
Discussion Starter #179
The SWR as designed is already extremely low, with insignificant mismatch loss. The gain figures in the pattern plots include conductor and mismatch losses
.

just so i understand correctly, the effective net gain is still very close to 11 dbd, even though the modeling SWR is well above 1.0
 

·
Registered
Joined
·
296 Posts
The SWR is very close to 1. Apparently you are not including the shunt capacitor in your model. You must do that to get accurate results.

Brian
 
161 - 180 of 361 Posts
Top