-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
Modeling my Ford Explorer
- Thread starter The DB
- Start date
The 4Nec2 charts always show the highest and lowest gain figures for a plot on the chart as a range. All of mine always have.
I haven't done more with the Ford Explorer model I made. Everything I have with models on that object has been presented here in this thread.
If you were talking about that small car, I haven't done anything with the model of the small car object in quite some time. I want a more realistic and average car model to use in the future.
The DB
I haven't done more with the Ford Explorer model I made. Everything I have with models on that object has been presented here in this thread.
If you were talking about that small car, I haven't done anything with the model of the small car object in quite some time. I want a more realistic and average car model to use in the future.
The DB
Recently I've been working with my small model, and the main thing that amazed me was other than the different pattern it presents from my Bronco model...is that the match is so bad even at resonance.
I noted earlier that I've also made a bigger model that is 18' feet long and 7' feet tall for the back and it is somewhat similar to your Explorer. I kept the skeleton design.
I just now compared my big model vs. my small model and I attached an overlay below. I see the bigger model producing less gain than the smaller model, and that really surprises me.
There is something really screwy going on with my mobile models or I'm just confused. Help me out.
How does the gain for your Explorer compare to your small model with both measured at the same maximum workable angle...like I used in my overlay at 23* degrees.
Last edited:
I just took my old Bronco model and removed all the wires from the back. No I did not make any real checks on the model, but I did not get any error messages either. I did the models a couple of days ago, but I just now compared them...while I was writing post #33.
Here is the small model at maximum of 30* degree of angle. I added a pattern at 23* degrees by moving the angle down. I did this to compare the gain at the same degrees the 18' foot Bronco produced.
I checked the model for segment and geometry errors and found none.
I did not tune this model to resonance, and the match is very bad with a 102" whip. So, that may make the difference in gain we see.
Last edited:
I made a new SUV modeling object today. It is much like the object I used above, only better. It was a case of me thinking I could make a better object than I used above, so I did.
So what are the differences?
First, this object isn't just a bunch of boxes like the previous model, like actual vehicles, there is some curve to it as shown here with a front on view.
Two, it's size is customizable.
I can use this base model for any number of different vehicles if I desire, I just need to take a few measurement and enter them in. All of these measurements are in inches. The model is designed in such a way that when I put the new information in the model adjusts itself automatically. It should work with any vehicle that has an SUV type shape to it...
And finally, there are fewer wires, which means it takes less time for nec2 to process this model.
Here is a link to the file itself if anyone wants to play with it.
The DB
So what are the differences?
First, this object isn't just a bunch of boxes like the previous model, like actual vehicles, there is some curve to it as shown here with a front on view.
Two, it's size is customizable.
I can use this base model for any number of different vehicles if I desire, I just need to take a few measurement and enter them in. All of these measurements are in inches. The model is designed in such a way that when I put the new information in the model adjusts itself automatically. It should work with any vehicle that has an SUV type shape to it...
And finally, there are fewer wires, which means it takes less time for nec2 to process this model.
Here is a link to the file itself if anyone wants to play with it.
The DB
DB, your new model really looks nice, but how does it perform with an antenna mounted?
I am also curious what difference you see with this new one compared to the object it replaced?
Can you cut this new one down to make a model like M0GVZ posted here:
Good video explaining mobile antenna placement.
How do you think these two models would compare in performance?
I am also curious what difference you see with this new one compared to the object it replaced?
Can you cut this new one down to make a model like M0GVZ posted here:
Good video explaining mobile antenna placement.
How do you think these two models would compare in performance?
I would need certain dimensions for that model, but if I had them then sure. I think my new one would have a lower SWR in general, performance would be pretty close, although the model he posted would have slightly higher gain. Both of these differences would be do to the same reason, namely the curves in my model. I think many of the differences would be the same as the comparison to my older model below.
M0GVZ, would you have access to that model, and if so could I get a copy of it for comparison purposes?
A few things to note, lets start with the gain plot comparison. These are all with a center mounted antenna.
In this first plot the front of the vehicles are to the left.
In this second plot the front of the vehicles are towards the bottom.
As we can see, the older model does provide more gain. That makes sense, it has more "metal" immediately under the antenna, and that metal is all flat. It does have the disadvantage of more gain going to the back of the vehicle. The new model has more consistent gain in all directions, and actually has a slight advantage in gain towards the front, although it is small enough that you won't notice it.
EDIT: I just noticed something, a difference between these models. The models are actually facing opposite directions in space. On one the front is towards the Positive Y, and on the other the front is towards the Negative Y. Unfortunately this means without some editing I can't easily overlay the results from these models as the results will be backwards from each other. The above results aren't so bad because one has a good omni pattern, but if I move the antenna on the models the results will not line up properly...
Now for the model data.
The old model is on the left, the new one is on the right. Both have lengths tuned to the low SWR point.
As we can see, the new model antenna has a lower SWR than the old one. I am reasonably sure that comes from the curvature in the model that didn't exist in the old model. Part of its ground plane is angled down, and what happens to impedance when we angle radials down on a ground plane antenna? There is also the fact that there is less roof on that model, namely because of the curves elsewhere on the model. This also has a lowering effect on impedance, especially when the bottom of the vehicle is about the same size overall as the other. Another thing to notice here is the radiation efficiency. The old model is more efficient. Again, this is because of the roof being completely flat. If you look above, this results in the larger overall gain circle for the older antenna.
In my opinion, the model that takes into account that there is some curvature to the roof and sides of a vehicle will be the more accurate model to play with.
The DB
Last edited:
DB, I may have misunderstood you when you said your model was designed to be easily re-modeled in size. I was suggesting a smaller version of your Explorer center mounted like M0GVS posted...not necessarily the exact same model.
Initially, I was seeing my smaller models showing better performance than my larger Suburban models, and the models I posted below still show a little better gain that my larger Suburban model not posted as yet. In the last few days I have been checking my mobile models over and found some inconsistencies between models that I did over the course of some time and it is not easy fixing, because there are so many wires.
Below I posted a model of my Bronco with an all metal top cover in the back. The comparison here was to model with and without a load to match a typical CB steel whip (102"/108") rather than having to shorten the radiator to see resonance...like I think you did on your examples above, at least that is the way I read your words.
My impedance values with a regular 102" whip initially did something similar to your model, its matching values were off the wall and fixing the model to resonance did not fix the match in the smaller mobile I was using. However, I did get the impedance close to 90 ohms at resonance like you did on making the radiator about 86" inches tall.
I have not made a judgment about lowering the SWR because I've not yet started reworking my 17'-18' Suburban model with a much more complex grid. In general I think you are right however. Thus far my larger models do not seem to need any matching at all in some mounting locations.
You are totally correct on one thing for sure...this is very tedious work.
As a note, now I'm also using Auto Segmentation in order to maintain some consistency in modeling...even though Roy Lewallen tells us...the use of either Conservative or Minimum Recommended segmentation often does not produce best results, and my Average Gain reports support that idea. BTW I'm not really getting good AG results with all these models I've done in Free Space, but I haven't developed a standard method as yet either.
It will probably never be possible for me to add curves, I'm already close to the maximum segments as it is now. The only way I can advance that idea is to keep using the auto segments and not experimenting with segment using the Average Gain results as a guide post. But I will say when the match gets good, however I do it, generally the AG is good too.
The first model is the model with matching and I included all the reports for you to consider. I also added the load information I used at the bottom of the Antenna Wire Descriptions.
The second model is with no matching added. It was just set close to resonance similar to what you did and our results are similar in impedance and SWR values. On this point I guess I could say my smaller Bronco and your Explorer model, even though very different, show about the same results and that surprises me.
I'm curious if center mounted is the best location for the typical highway use.
DB, one of these days hopefully I will be on top of this business like you are, and never fail to see my expectations fulfilled and never be surprised at any results. For me, that is a sure sign of a guy that knows his stuff.
Keep up the good work on mobiles...the least understood topic in antennas IMO.
Last edited:
Here are my models compared between my Bronco wMetalTop 103" Ct vs. Suburban Center Top 108" inches.
Here we don't see much difference in the gain and angle between these somewhat remarkably different models at least in overall length. You may notice that my smaller model shows a bit more gain at a 3* degree higher angle...which probably has to do with the difference in radiator length. IMO, at these low heights above Earth it is very difficult to easily achieve added gain. As antennas get closer to the Earth they tend to show less gain at higher angles, so the range of difference in gain is also very small close to the Earth.
These models are the same height to the base at a 17" inches of ground clearance and the top of the bodies are 72" inches high.
Do these models give you pause to consider the standard fare for what we have been hearing about CB mobile patterns for years?
Here we don't see much difference in the gain and angle between these somewhat remarkably different models at least in overall length. You may notice that my smaller model shows a bit more gain at a 3* degree higher angle...which probably has to do with the difference in radiator length. IMO, at these low heights above Earth it is very difficult to easily achieve added gain. As antennas get closer to the Earth they tend to show less gain at higher angles, so the range of difference in gain is also very small close to the Earth.
These models are the same height to the base at a 17" inches of ground clearance and the top of the bodies are 72" inches high.
Do these models give you pause to consider the standard fare for what we have been hearing about CB mobile patterns for years?
I think that is what I was talking about. For a fair comparison I would want to adjust the size of my model to match his.
Luckily changing the dimensions of the model in question is not hard, I just need to know the dimensions. I have modeled the entire object 6 inches higher, simulating a six inch lift kit on the vehicle, and that resulted in slightly more gain. Not a surprising result, was just playing.
In the comparison above I simply tuned the antenna lengths to low SWR. It doesn't surprise me that different lengths are required on different vehicle objects, and even on different parts of the same vehicle object. We are effectively changing the size of the ground plan after all... So far it hasn't made that much of a difference in length on the explorer objects I have made. As far as I can tell, working as intended. The good thing about this antenna is it is very wide banded, both for you and for me, so you can generally get away with a mount and the antenna, and your SWR will generally be low enough, even if it isn't at its lowest point. These models are, however, an example where the low SWR point is not the resonant point.
I have actually had thoughts about doing things different with wires and their segments. I don't really have a good way to explain what I want to experiment with, but it will require no auto segmentation. Essentially, the idea behind it is there are certain parts of the model that benefit more from more segments than others, so in those areas you can use wires with a higher segment count for a given length, and other wires on other areas of the antenna can have fewer segments per the same given length. I don't want to get to deep into it at the moment as I haven't completely thought it through as of yet. What can I say, it is a work in progress.
Honestly, I wish you would start using 4nec2. I know you are an EZnec guy, but with 4nec2 you could at least play with more of the stuff I am doing, and perhaps get a better idea at some of the things I am trying to do. I'm not saying to stop using EZnec by any means, I am just saying that I think it would benefit you to become more accustomed to 4nec2.
The best answer to this is it depends. For example, what are the goals of the operator? If you want the best omni-like coverage, yes, the middle of the roof seems to be the best place on this model. It is also the place of the lowest natural SWR I have as of yet found on this modeling object, as well as the point that has the best radiation efficiency. If you want some gain to the front, moving the antenna towards the back will definitely help with that, in my models giving about 2 dB gain to the front and a 3 dB loss to the back.
Thank you. I wouldn't say I never fail to see my expectations fulfilled and I am never surprised from results. When it comes to expectations, I don't model with expectations unless I am trying to demonstrate something for someone. It is just that more often than not, when I see something I don't expect, I can generally come up with an explanation that I hadn't considered. It is the occasional strange results that I learn from, so I am actually excited when they happen, but as I continue to learn more they are getting fewer and further between.
I guess that is my ultimate secret, no matter how much you think you know, don't stop learning/exploring.
The DB
I just recently raised the ground clearance for my mobiles from 14" inches to 17" inches and it messed up the tuning. I also added some more wires to the doors for example and that changed the tune. I added a grill and some wires to the hood areas and that changed the tune and performance for the model a bit. I'm a little surprised a 1/4 wave antenna works at all on a mobile as a ground plane it looks so big and with so many wires. I will post more about my ideas on this in another post however.
I don't find it hard to change dimensions either. I just add or delete whole sections of the Suburban to the front and back as necessary to change the overall length. It just takes entering the wire #, what you want to do...and then just click "OK" Eznec does it automatically, but the rule "good input means good results" really applies here...so be careful.
DB that said, even a mistake is not really bad because Eznec records every change made to a model during a session and you can do and undo all of your changes just by click the button. The only math I have to do is the typical sort of dimension work one would do just building an antenna kit.
Of course getting good dimensions over the Internet and these forums is a chore all by itself and I don't think Eznec will help with that. Eznec provides some very useful modifying tools to connect, copy, move, raise and lower the whole model are parts with a click of the OK button. I don't have to do all the formulas and coding, I just enter the wire numbers and the change parameters and Exnec automatically makes radials, top hats, coils, helix, loops, etc. A calculator does come in handy ever now and then however.
These utilities allow me to not have to use my fingers, toes, slide rule, and abacus anymore to do the geometry and math.
Like you note...I also find these mobiles very broad banded. Is this what we might see when the radiation efficiency (RE) is so low? Or, could these 1/4 wave mobile antennas show a low RE because the antenna are low to the Earth? I'm not sure.
Do you think this situation has anything to do with these antenna being so broad banded?
We started to talk about this a while back, but the idea fizzled out apparently.
I use to use as a standard approach to a model by making the segments all as equal in segment length as possible. Within the segment limitations for my version of Eznec at 500 total segment...I tended to use 2", 3", 6" inches for the segment length...and I never looked back. That was a mistake.
Reason being most of those models did not fare so well when using the Average Gain Test with the model in Free Space and in 3-D mode for Plot type.
When I started testing my old models over real Earth and then converting them to Free Space...I could see the differences you describe above. In most cases, my having all the segment lengths for the wires equal...does not appear to help for accuracy in modeling.
After talking to you about using auto-segmentation...I began to use the feature in Eznec. I'm still not clear on your use with 4Nec2 but with Eznec the feature insures that the model uses the same routine for the use of segments in the whole model to start but I don't claim to know the bases for how it works. Regardless, after my model is showing a good match and is also error free...then I might fiddle with the segments on wires that have more than one segment per wire...to see if I can improve the gain without messing up the match. I do this in Free Space mode so I can use the Average Gain Test as a guide post.
I don't think that will happen...I have trouble seeing and it is getting worse. I don't have the time or energy and I think I'm still learning stuff using Eznec.
Like you note...I also find these mobiles very broad banded. Is this what we might see when the radiation efficiency (RE) is so low? Or, could these 1/4 wave mobile antennas show a low RE because the antenna are low to the Earth? I'm not sure.
That is a very good question.
The broad banded antenna is helping to show this effect, but that is not why SWR and resonance doesn't necessarily happen at the same point. It takes more than resonance to figure out SWR.
This was in reference to some parts of a model benefiting from having more segments than other parts of the antenna. I don't remember talking about this specifically. I do remember the discussion about high segment counts vs lower segment counts, and where some lower amounts are just not enough to maintain accuracy. I don't recall having any discussion where some parts of the antenna benefited more from a higher segment count than others.
The DB
just curious if u could model a p/u truck s10/ranger size. with a 7 foot top load mount on right side of the bed .and a 5 foot baseloaded mounted on the roof.im thinking the baseload should show slightly better given its mounting height advantage