Modeling hose clamps, element mounts and gamma matches with NEC2?
Scott
exact needs using NEC Win Plus+ (NEC2). I carefully remodeled the 4
element 12 meter Yagi from the ARRL handbook. Since I am reusing a old 4
element Yagi, I am accurately entering taper schedule. I am so happy
with the design, I want to ensure the real world model is close as
possible to what I've modeled and this has raised a few questions:
1) I want to use hose clamps so the elements are adjustable. Should I
try to model these? Common sense tells me I would need to slightly
shorten the elements...
2) What about the element to boom mounting plate? The plate is a 3.5"
wide piece of aluminum with a 3/4" tube on it so that the 5/8" element
passes through it. I modeled the 3.5" X 3/4" inner element piece, what
what about the rest of the plate? This seems like it also might cause me
to need to shorten the element length as well?
3) Lastly, I am going to use the gamma match that came with the antenna.
I have 2 questions about this:
a) Do I need to compensate the driven element length because of the
"shorting strap" and extra material added by the gamma itself? I just
have a 50 Ohm source in the middle of the driven element and that's it.
b) The center of the gamma rod mounts offset from the center because of
the mounting plate (it mounts 1.75" from the center). Should I try to
rework this gamma match so that it taps in the center of the boom and
element? Or can I offset the element in the mount so that the place
where the gamma rod connects is the physically the center of the element
length (I don't think so, but I'll ask)
Here is the manual for the antenna I am using:
http://www.majestic-comm.com/assembly/SUPPORT/M104C.PDF
It shows how everything fits together.
Thanks!
Scott.
Wes Stewart
wrote:
|I have spent a great deal of time designing a 4 element yagi to meet my
|exact needs using NEC Win Plus+ (NEC2). I carefully remodeled the 4
|element 12 meter Yagi from the ARRL handbook. Since I am reusing a old 4
|element Yagi, I am accurately entering taper schedule. I am so happy
|with the design, I want to ensure the real world model is close as
|possible to what I've modeled and this has raised a few questions:
|
|1) I want to use hose clamps so the elements are adjustable. Should I
|try to model these? Common sense tells me I would need to slightly
|shorten the elements...
Don't bother.
|
|2) What about the element to boom mounting plate? The plate is a 3.5"
|wide piece of aluminum with a 3/4" tube on it so that the 5/8" element
|passes through it. I modeled the 3.5" X 3/4" inner element piece, what
|what about the rest of the plate? This seems like it also might cause me
|to need to shorten the element length as well?
Leeson in "Physical Design of Yagi Antennas" has info on calculating
the equivalent diameter of various element to boom clamps.
You must consider not only the effects of the plate but that of the
diameter of the boom. The data summarized are a bit complicated but
I'll try to explain them here.
Let W = width of the mounting plate (3.5" in your case)
Let T = thickness of the mounting plate (?)
Let D = diameter of the element. (3/4")
Then:
A = (pi*D^2)/4 + T*W
P = pi*D + 2*(W + T)
the equivalent diameter of this is then:
D_equiv = ((A/pi)^.5 + P /(2*pi))
So you plop this into your model as the diameter of the element with
the length equal to the length of the PLATE. Whatever 3/4" tubing (if
any) that extends beyond the plate should be modeled as---you guessed
it--- 3/4" tubing. If you are using the mounting method in the
Antenna Handbook then this should do it. If you are using what is
shown in your reference, it looks like the "plate" is really one-sided
as contrasted to what Leeson uses. There is probably some difference
but I seriously doubt that it is significant.
A rule of thumb (from Lawson's "Yagi Antenna Design") for the plate
clamp is that the effect of the boom diameter is to reduce the
electrical length of the elements by 6% of the boom diameter, which
again is insignificant.
|
|3) Lastly, I am going to use the gamma match that came with the antenna.
|I have 2 questions about this:
| a) Do I need to compensate the driven element length because of the
|"shorting strap" and extra material added by the gamma itself? I just
|have a 50 Ohm source in the middle of the driven element and that's it.
No, the gamma adjustment procedure will take care of this. The gain
and F/B are not a strong function of the driven element length. You
don't need to be particularly accurate in determining the feedpoint
impedance since you will be taking care of that with the matching
system and gammas (IMHO) are not particularly easy to model.
|
| b) The center of the gamma rod mounts offset from the center because of
|the mounting plate (it mounts 1.75" from the center). Should I try to
|rework this gamma match so that it taps in the center of the boom and
|element? Or can I offset the element in the mount so that the place
|where the gamma rod connects is the physically the center of the element
|length (I don't think so, but I'll ask)
I would prefer to have the coax connector mounted on the boom. That
is the electrical center of the element. Don't try to fudge the
element off-center.
Have fun,
Wes
Dan Richardson
wrote:
>So you plop this into your model as the diameter of the element with
>the length equal to the length of the PLATE.
Wes,
The only problem with that is NEC2 doesn't like stepped diameter
elements (wires) and will not model properly.
BTW: Antenna Model (a MININEC program) will handle that okay.
Danny,
K6MHE
Scott
correction algorithm. I'm sureWes was taking this into account.
Scott.
Roy Lewallen
>
> The only problem with that is NEC2 doesn't like stepped diameter
> elements (wires) and will not model properly.
>
> BTW: Antenna Model (a MININEC program) will handle that okay.
Some NEC-2 based programs, like EZNEC, also handle stepped diameter
elements for common Yagis. They use a technique described by Dave Leeson
in _Physical Design of Yagi Antennas_ to substitute an equivalent
constant diameter element for each tapered one. The substitution works
very well, and actual pattern measurements of beams analyzed with EZNEC
have corresponded very closely to predictions. You can, if you wish,
manually apply Leeson's method to your model before running it on NEC-2,
but it can be tedious. Programs like EZNEC do it automatically. But if
you run a tapered Yagi on NEC-2 without the correction, you're just
about sure to get a very wrong answer for the front/back ratio.
MININEC has its problems with Yagis, too. It produces an error when
modeling Yagis with a relatively large number of elements, such as used
for VHF or UHF beams. (The error probably exists with shorter beams but
is too small to cause a problem.) The effect is that the pattern is
about correct, but at an offset frequency. Quite a bit of effort by a
number of people went into tracking this down, but to my knowledge no
one found the cause. The original authors of MININEC also couldn't find
the cause of this problem. I believe Brian Beezley, K6STI, claimed to
have corrected it in his MININEC-based programs, but don't know of any
objective evaluation of the correction's effectiveness, and he kept all
details proprietary. I don't know whether the Antenna Model folks have
found a way around it or not.
Roy Lewallen, W7EL
Scott
strange the formula doesn't take the length of the plate into account? Does
it assume the plate is 3.5" X 3.5"? Time to drop this into my picture perfect
model and see the damage!
Scott.
Scott
up about 100Khz. I can tweak it back down, but I hope this holds when its on the
roof!
Scott.
Wes Stewart
wrote:
|Wes - I calculated right about 2" (my plate thickness = 1/8"). I find it
|strange the formula doesn't take the length of the plate into account? Does
|it assume the plate is 3.5" X 3.5"? Time to drop this into my picture perfect
|model and see the damage!
No. It's calculating an equivalent diameter for that length of
element, whatever the length of it.
If your plate is 3.5" in length then you model that portion of the
element as 3.5" long, 2" in 'diameter'.
Wes
Scott
(the length of the plate in the boom's axis) of the mount plate is ignored in
calculating the diameter.
Scott.
Wes Stewart
wrote:
|I understand what you meant and that is how I did it, but I am saying the length
|(the length of the plate in the boom's axis) of the mount plate is ignored in
|calculating the diameter.
|
|Scott.
What we have here is a failure to communicate.
The "width" is the dimension along the axis of the boom. The "length"
is the dimension along the axis of the element.
Scott
R P Haviland
following:
1: Get the perormance you want with average element diameter, at design
height and best guess ground
2: For each element without the others present, find the resonant frequency
3: Build the elements, tapered, hose clamps, mounting plates, etc
4. One at a time, mount the elements on the boom, and measure the resonant
frequency with a grid dip meter coupled to the element, inductively near the
center, a very small capacitor near the high voltage point, usually the end
6. Adjust the element length to get the same resonant frequency as in 2,
repeat for all elements
7. Mount the elements.
Performance will be as near to calculated as you can get. Major difference
will be the difference between calculated and actual ground conditions
bob w4mb