GO or NO-GO?

[Jlerch]

Ok, here's what I came up with, and I'm looking for a GO or NO-GO recomendation. (Plus, exactly who else could I show this to that might have any appreciation for the work involved

The center vertical is 1/4 wave long at 910mhz or 3.24 inch, measured from the top of the 90 degree SMA -> Coax connector

The 6 radials are also 1/4 wave long and a made from stainless steel spot welded to the outside of the sma connector. The hot glue is there for strain relief.

As Installed on my Foamy glider:

As modeled at an ALT of 120 feet over normal soil.

If I get the wiring done, I might get a flight in today...

[Mr.RC-Cam]

Looks good, but you need to flip it over so that the RF pattern is directed towards the ground rather than to the sky. Also, isn't 1/4 wave at 910Mhz more like 3.08 inches?

[Jlerch]

Looks good, but you need to flip it over so that the RF pattern is directed towards the ground rather than to the sky.

This is one of those compromise results. The airframe may have to make a water landing, and I'm trying to keep everything above the water line. Also, the airframe doesn't have landing gear, so its a belly landing every time. I couldn't come up with an attractive solution to either problem.

I finally gave up on the "point it down" approach when the modeling software clearly showed the free space radiation pattern as a big beach ball in either orientation. When modeled over ground, the pattern was essentially the same..

Of course, all this assumes I modeled it correctly, which is a HUGE assumption!

Also, isn't 1/4 wave at 910Mhz more like 3.08 inches?

Now there's a good question. Specifically, do we want the radiator to be 1/4 wave length based on the speed of light in the material the radiator is fabricated from, or the material the radiator will radiate into? The modeling software says 1 wave at 910mhz = 12.97 inches, online frequency to wavelength conversions agree. This makes 1/4 wave at 910mhz = 3.24 inch. However, is that the wavelength of 910 mhz in a vacuum, in standard atmospheric temperature and pressure, or in copper?

Finally, since your asking me, I have to assume you already know the answer. Me, I'm just a guy off the street that's taken an interest in all this and am doing my best to educate myself. So, I'm probably screwing this up in a phenomenal fashion!

[Mr.RC-Cam]

(1) You've built a ground plane type antenna, not a center fed dipole. So, typically there isn't good vertical symmetry (such as is normally seen on a dipole). With the mounting you show, much of your RF will go up into the sky, rather than to the ground. It should be flipped over if you want optimum aerial performance. I'm not saying it won't work, it just won't be ideal. Instead, a dipole might be best for your installation.

(2) The common method for determining quarter wavelength in copper medium is to divide 234 by MHz. This gives estimated 1/4 wave length in feet. Ideal length requires trimming the element during use.

[Jlerch]

(1) You've built a ground plane type antenna, not a center fed dipole. So, typically there isn't good vertical symmetry (such as is normally seen on a dipole). With the mounting you show, much of your RF will go up into the sky, rather than to the ground. It should be flipped over if you want optimum aerial performance. I'm not saying it won't work, it just won't be ideal. Instead, a dipole might be best for your installation.

I wish I knew what I was doing, because I'm doing something wrong as my ant models don't bear this rather obvious common sense solution. For instance, comparing the same two antennas in free space, one pointing down, the other pointing up

Same thing, flipped over pointing up:

Now same antennas, but modeled at an altitude of 120feet over "high accuracy" ground consisting of pastoral or heavy clay.

Pointing down:

Pointing Up:

In free space the two orientations appear identical, over earth, the two antennas seem to have the largest deviation based on constructive /destructive interference from the earth reflection.

(2) The common method for determining quarter wavelength in copper medium is to divide 234 by MHz. This gives estimated 1/4 wave length in feet. Ideal length requires trimming the element during use.

So what happens when the wavelength in air is different than the wave length in copper medium? At 910mhz, the difference is shy of an inch on a full wave length. Its got to be a huge difference down in the 10s of MHZ range..

To be honest, It still makes my head hurt to think that a piece of wire 1/2 wave length long can have a measured zero volts on each end and tens or hundreds of volts measured in the center..

[Mr.RC-Cam]

Here's a drawing that shows how a quarter wave vertical differs from a half wave dipole:

[Jlerch]

Here's a drawing that shows how a quarter wave vertical differs from a half wave dipole:

The quarter wave monopole is depicted on an infinite ground plane. I have a finite ground plane, and that ground plane is only 1/4 wave in radius. I wonder how such a small diameter 'ground plane' will affect the situtaion.. Time to fire up the simulator again and increase the radius of my 'ground plane' by several wavelengths to see if I get your Infinite ground plane radiation pattern...

BTW, if I wanted to use a simple half wave center fed dipole. Only the center conductor would be attached to the center antenna, what completes the circuit?

[Mr.RC-Cam]

if I wanted to use a simple half wave center fed dipole. Only the center conductor would be attached to the center antenna, what completes the circuit?

See drawing at the top of the page: http://en.wikipedia.org/wiki/Dipole_antenna

[Mr.RC-Cam]

The quarter wave monopole is depicted on an infinite ground plane. I have a finite ground plane, ...

That's a very good point. I finally had time to dig through my antenna books and it looks like your simulations are moving along. The ground-plane antenna, with 45 degree angled radials, should work out fine. In free space, it should behave much like a half-wave antenna.

One thing that doesn't seem quite right in your model is the number of RF sources. I see you are using six, but I think it should be set to one for this design. Back when I played with EZNEC, there was ground plane antenna in the user library, so try to find one and use it as a proven starting point. I recall that EZNEC had problems with angled radials at the time, but maybe that has been fixed?

Also, for a sanity check you could compare your simulations to these MININEC plots (if yours match, then you know you are good to go): http://www.qsl.net/df3lp/projects/vertical/index.html

Finally, here's a online calculator I found today for estimating element length on a ground plane antenna design (but keep in mind that the angle of the radials, and proximity to ground, will affect resonant length):

http://www.csgnetwork.com/antennagpcalc.html

Good luck!

[Jlerch]

One thing that doesn't seem quite right in your model is the number of RF sources. I see you are using six, but I think it should be set to one for this design. Back when I played with EZNEC, there was ground plane antenna in the user library, so try to find one and use it as a proven starting point. I recall that EZNEC had problems with angled radials at the time, but maybe that has been fixed?

With a Homer Simpson Esque "DOH" I finally realized how sources were working in the simulator, which is NOT how I thought they worked..

After reading the Stinking instructions in the help file, I worked out the following simulations for a single RF source ant.

In free space:

1 inch above normal soil

10 feet above normal soil

100 feet above normal soil

1000 feet above normal soil

I still don't get why the sim says 3.24" = 1/4 wave, but clearly prefers the 1/4 wave Ant height to be 3.08"

At this point, I'm going to stop playing with the sim and get back to building.. I will admit, loop antennas are starting to look interesting to simulate

Thanks again for your help!

[Mr.RC-Cam]

At this point, I'm going to stop playing with the sim and get back to building.

I can empathize with that. I'm an ex-simulating fellow from bygone days. My partner and I had the responsibility of performing all the design simulations for a complex medical imaging system. We used Daisy System workstations (early Mentor stuff) that cost more than a house at the time. It was exciting at first, but the novelty soon wore off. All I can say is that after a few months, I was fully cured of ever wanting to simulate again. Now I do my best simulations with solder and wire cutters.

[Jlerch]

I can empathize with that. I'm an ex-simulating fellow from bygone days. My partner and I had the responsibility of performing all the design simulations for a complex medical imaging system. We used Daisy System workstations (early Mentor stuff) that cost more than a house at the time. It was exciting at first, but the novelty soon wore off. All I can say is that after a few months, I was fully cured of ever wanting to simulate again. Now I do my best simulations with solder and wire cutters.

That reminds me of a good friend that landed a job with "Harris Corp" back in the late 80s. It sounded so cool on paper, he was going to be building a military Tank warfare simulator. In practice, he sat behind a terminal typing numeric values into what he described as a three dimensional data base, describing each cell's details such as altitude, collision data, color, etc. I thought that sounded kind neat until he told me they only rendered the terrain every 2 weeks, then they spent the next week debugging the existing terrain, the week after that building new terrain, re-render, repeat.

I'm pretty certain he hated that job! :-)

[Terry]

Yep, get out in the field and test it

Terry

[Mr.RC-Cam]

Yep, get out in the field and test it

I agree. A few years ago a lot of us on this forum built Tx mounted ground plane antennas after yb2normal reported good results with his. On my model, I found that it worked much better pointing down, rather than up. So, it would be interesting to hear some new reports on the orientation's performance.

[Jlerch]

Ok, Here is a follow up.

In a nut shell, this setup is much better than I had before on the other airframe. If its because of the antenna or just the way it was installed, I don't know, but its working great.

For the first time I was able to record a whole 10 minute flight with out my ground recorder giving the video input the finger after a loss of signal. I still don't trust the system yet, that will take a little time, and the only complaint I have is the little white foamy airframe is darn near impossible to find when transitioning from FPV to direct observation.

Once I get my Eagle Tree OSD, and get it working, I'll start stretching the legs on this airframe. Having flown out of range of the 2.4ghz Spektrum radio once on the previous airframe, I am leery to get too far away at this time.

Till then, here is the maiden flight as seen by the ground recorder.. (also, I seem to really get a kick out of seeing the wing and aileron surfaces while flying! )

And here is a video tour of the setup

As soon as the upload is complete, I share a sunset airborne beach video that actually made my grandmother cry when I had her on the glasses!

[Mr.RC-Cam]

It's doing a nice job! I'm envious of your flying field -- lots of clear space away from civilization.

[Jlerch]

It's doing a nice job! I'm envious of your flying field -- lots of clear space away from civilization.

Here's the beach video. The ground station is located at the large concrete building seen at the 1:35 second mark. Not the most desirable position, and it showed a few times.

Apologies for how jerky this one turned out, I was out near the water flying direct, while the glasses and ground recorder were passed around the family.

I look forward to getting out and trying again with a proper setup, should be fun

[Terry]

Seems to be working great, well done

Terry

[W3FJW-Ron]

Just a ? on the modeling program. Is there an input for the velocity factor or type of wire (copper, steel) ? If so, that could make a difference in the length of the antenna 3.24 vs 3.08. Seems to me though that since RF only travels on the surface of the radiator it shouldn't make that much of a difference.

Which leads to another ?. Has anyone ever tried thick driven elements such as 1/8 or 1/4 brass or copper tubing vs the wire elements most commonly used?

Edited May 11, 2009 by W3FJW-Ron

[n3m1s1s]

i really like it but im a total newb lol so is there any chance you could make a few of them and ship them to the UK at a reasonable fee? t,i,a