Have you struggled with building quality ham radio antennas? High SWR? Problems with feedline? Well in this video I explore 5 common mistakes made by antenna builders and some tips to help you get the most from your efforts.
With the success of my 11 element 2m LFA, I decided to build a smaller 8 element out of spare material I had here at home whilst waiting for the propagation to pickup (note I'm still waiting on you E-Layer!).
The idea behind it was to build a yagi for either portable use, or to replace the 7 element YU7EF at the remote station to see how it performs. I chose the 8 element model detailed on G0SKC's website.
Justin had the details for using 6mm and 13mm tubing, however I only had 6.35mm and 10mm on hand, so I popped the figures into MMANA and rescaled for my material.
The antenna pattern looked good, note the near 30dB F/B ratio and reduced sidelobes. Comparing this to the 7 element I have at the remote station showed a significant improvement. Red is the LFA, Blue is the YU7EF on the remote station.
Anyway... enough theory. I managed to build this up in an afternoon. I took my time measuring and cutting the elements. Some photos.
Another good build requiring little to no adjustment to obtain a decent match. Now to find a place to use it!!
Further to my previous posts on the "hybrid" LFA...
I managed to finish construction over the weekend and install it on my tower. At the same time I also decided to take down my 6m yagi to convert to an LFA design too. So how does it perform?
That evening I ran some tests with Peter 7PD on WSPR. There seemed to be some tropo enhancement as usually I cannot hear him from my location. Tests seemed to show the antenna was working quite well. I also tuned to a local beacon and noted signal strengths pointing in different directions. In this area we have a lot of mountain ranges so it's difficult to get an accurate reading due to reflections/refraction from the beacon.
The most impressive bit about this antenna though that I noticed straight away was the lower noise I had overall on 2m. You may recall in my previous post I mentioned that in the ZL direction I had high levels of noise on 2m. Well, this antenna has certainly improved on that! Rough measurements on the old antenna indicated about 12dB of increased noise in Spectrum Lab when pointing at ZL as compared to a relatively quiet area. The noise level increase is now only 6dB over that same direction. In fact the noise now peaks between 30-60 degrees, which is perfect as there is nothing to beam at in that direction.
Below are some photos of the build. As detailed before, I just used three (all I could fit) ferrite cores of known impedance over the feedline and placed them as close to the feedpoint as possible. I then also slipped three further cores, around the bend of the coax underneath the boom. I also made a "ugly balun" of some sorts, 2.5 turns of coax... I also connected the centre of the driven element non-feed end to the boom using a small screw and some lugs.
Photos of build below.
Loop assembly with coax feed.
Close up of mounting method to boom/mast
Boom to centre of driven element. Note, another lug is present under the screw head to the right hand side of the photo, soldered to the centre screw lug.
Return loss come up pretty good.
Further to my previous post about the LFA, I have now completed construction on my "hybrid" version. The next step is to prepare a balun for the feed. But what balun (if any) is needed? In my previous yagi, the only "decoupling" I ever had was three clip on ferrite clamps of unknown impedance at the feedpoint. To make matters worse, I don't think my previous weatherproofing was that good. Nevertheless, I've learnt a bit in that time, and intend to make improvements on this antenna.
In researching baluns, I discovered there have been many different types of variations used on the LFA in particular. So many in fact, I don't think most designers/builders of these antennas ever research how effective their balun really is. Is it doing the proper job? I found too this article to be of interest. The author states that the full wave loop on these antennas have a natural balance, effectively eliminating a majority of current flow on nearby things such as the coax, or boom. To verify this in MMANA, I decided to add an extra wire, drooping down, but connected to the feedpoint, a 1/4 length long. Below is the YU7EF EF0211 (my old antenna).
Notice the current distribution. Vertical current components are shown in Red and horizontal in Blue. There is a significant amount of current flowing on the vertical wire down from our feedpoint. This would effectively be common mode current flowing on our coax (or mast), disrupting the pattern, and causing SWR issues. This wire (coax, mast) then becomes part of the antenna. Also notice how badly the current distribution is in the dipole. Below is the radiation pattern.
As can be seen, the antenna looks nothing like it used to. Gain is down, F/B is shot to pieces. Now lets compare with the hybrid LFA.
As is clearly seen, there is hardly any current flowing on the vertical wire now. The current distribution in the dipole is also more symmetrical.
There is a small amount of pattern distortion, but it is leaps and bounds better than the split dipole. In MMANA you can zoom in on the current. It takes a lot of zoom to see the smallest current flowing on the vertical wire.
Here is a comparison of current flowing on the boom.
As can be seen, significant current flows on the boom with the split dipole.
Whereas the split dipole has a lot of current flow on the boom, the full wave loop has very little. This seems to correlate with the linked article earlier. I'd call this a more symmetrical dipole feed. Interestingly here, Justin states that "the loop is not self-balancing, not in all real-world conditions in any case". Interesting.
So does that mean that a balun is not required with these antennas? This perhaps explains why so many hams who build these antennas see easy to obtain results and performance by just directly feeding with coax. However, for purists out there, a balun is not necessarily a bad thing.
The most common form of balun (and what is preferred on G0KSC's site) is a choke balun. This comes in many forms, the most common being the "Ugly balun". This consists of a few turns of the coax just after the feedpoint. The idea being that the inductance formed in the coil creates a RF choke. The problem with this type of balun is there are not many published test results of exactly "how many turns" one needs. If you get it right, an extraordinarily high impedance of typically unknown value can be obtained at the centre frequency of the antenna. In this case, it works really well, if you're lucky.
Another common method is sliding ferrites cores over the coax. The impedance at the centre frequency can be known using the ferrite manufacturer's datasheets. Sliding multiple ferrites increases the sum of the impedance, however a large number of ferrites is usually required.
Two other interesting "baluns" I found builders were using were the G0KSC hybrid, consisting of a bent piece of tubing, or flat bar connected to the centre conductor of the feedpoint and back 1/4 wavelength then attached to the boom. A small variation on this was also a piece of 1/4 wavelength stub piece coax (calculated for velocity factor) attached to the feedpoint, then the far end shorted to the boom. Both of these "baluns" are not really baluns at all. All these create is a "DC grounded" antenna, i.e. the centre conductor is connected to the shield. In my opinion all this does is limit the bandwidth of the antenna (could be a good thing?) and creates the DC path. Alternatively, a much easier method, at least with the LFA, is to directly connect the centre of the non feed end of the driven element directly to the boom. This effectively does the same thing as the two "baluns" mentioned above, albeit without the possibility for measurement errors.
So what will I do on my antenna? I think I will slide several ferrite cores of known impedance over the feedline to stop any small amount of common mode current that shows in the models. I'll also attach the centre of the far end of the driven element to the boom for a DC ground.
The amount of current radiated down the coax and on the boom of the LFA antenna doesn't concern me enough to warranty building a "proper" balun, such as the I0QM balun. This would only increase coax losses, and allow more places for water to get in. In the next post, I'll detail my results.
As per my previous post I have been looking at building a LFA yagi. Essentially this is a folded dipole loop turned on its side (i.e. mounted parallel with the rest of the elements). The antenna is essentially a 12.5 ohm antenna with the full wave loop stepping up the impedance to 50 ohms. This enabled direct feeding with coax (with the appropriate balun of course).
I contacted Justin G0KSC several times to get details on a 11 element design for 144MHz. I currently have a YU7EF 11 element - EF0211 design.
YU7EF has a EF0210LT design which has suppressed sidelobes to lower noise pickup. I wanted an antenna such as this to lower my noise level on 2m, but also have the "low noise" loop fed arrangement. I set about designing a yagi using MMANA. Now according to G0KSC, the loop cannot be accurately modelled in MMANA. However I decided to "have a play" anyway and build the design I came up with.
I call it a VK7HH 11 element "hybrid". Hybrid because it's a mix of the YU7 suppressed sidelobe antenna, and the G0 loop fed system. As I type this, I have the yagi being built in my shed, so results to come.
The link to the MMANA file is here. I might post the antenna dimensions on this page, if I get time in the future. As can be seen in the below free space plot, the antenna has very suppressed sidelobes. My goal was to keep these under 30dB. The antenna has a peak gain of 12.32dBd at 144.3MHz (design freq). This is probably 0.5-0.7dB less than the "optimum gain" for this length of boom. However I think there is more to gain in the compressed pattern of the antenna. The purpose of my antenna is for use on Sporadic E openings, so gain isn't a massive necessity either. This antenna also models well in a 4 bay stack, suitable for EME work with very small spacing for stacking. Full details below.
Gain = 12.32dBd (14.47dBi)
F/B = 31.1dB
SWR = 1:03:1 (144.3MHz)
Bandwidth = 1.6MHz <1.5 SWR
As can be seen, the pattern is exceptional for this antenna. All sidelobes and behind the antenna are less than 30dB down on the main lobe. So how should this antenna perform in theory?
Below is a MMANA (read not accurate apparently) comparison of my hybrid antenna (blue) the YU7EF EF0210LT (low sidelobe antenna) (red), my current YU7EF 11 element (green).
As is evident, both antennas beat my current antenna for sidelobe suppression and F/B ratio. Especially so on the elevation plot (right plot). My hybrid antenna has 0.5dB more gain than the EF0210LT for the longer boom and extra element, but 0.6dB down on my current antenna. I explained earlier in this article that isn't too much of a concern for me.
A lot can be taken away from this - at least in theory, antennas can be modeled in MMANA or similar, and optimized for the users particular needs. Below is a 4 stack array and the associated pattern.
Horizontal = 2.3m, Vertical = 2.4m. Different stacking distance may be suitable for extra gain, etc. Antenna can also be stacked in a two bay array.
Sometime ago I built a decently sized 6m yagi here at home - a YU7EF EF0606. This yagi has performed very well over the past few years, and I've had to do very little maintenance on it. I've worked all of mainland VK on Sporadic E, JA, ZL, Fiji and the Philippians. One thing that has eluded me though is USA on WSPR. N3IZN runs 6m beaming VK during the summer months here, but I have yet to decode Chris's signals. One potential issue is that this heading is directly over my house (and several other houses in the neighborhood, as well as a fish factory some 2km away).
I have the same issue with this fish factory on 2 metres. On that heading is directly toward ZL, and my noise floor increases dramatically in that direction.
I've been looking at Justin G0KSC designs of LFA antennas. These seem to offer superior noise rejection from nearby noise sources due to the utilization of a Loop Fed arrangement. I like experimenting with antennas, thus I think I'll give one of these antennas a build to evaluate performance. Below is the comparisons I made with G0KSC designs vs my YU7EF.
Below is both antennas modeled in MMANA - Note - Justin specifically says on some of his designs that the true SWR of the LFA arrangement on some antennas cannot be accurately modeled without using the NEC4 engine (something MMANA does not have). Most of his 6m designs however model perfectly fine in MMANA.
YU7EF EF0606 Free space
Gain = 9.63dBd
F/B = 27.9dB at 180 degrees.
This antenna has a distinct rear lobe which is not uniform. The elevation plot is also interesting. There are lobes present, for instance at 76 degrees which is 15dB down from the main lobe and off the back (21dB down at 135 degrees). Why is all of this handy? These elevation plots show us how much noise (read unwanted signals, TVs, solar inverters, DSL etc) are being picked up by the antenna from nearby sources.
A major stumbling block however is that all of these plots are in free space... what about over real ground? In my reading various articles and blogs, comparisons are made between antennas which are at different heights above ground. Sometimes an operator will say "my old yagi worked better for DX, even when it was closer to the ground".... This needs to be taken into account. For any yagi to be truly compared it needs to be on the same mast, at the same height, or very close to it.
I'm not too fussed... as long as performance is not degraded to the point where I notice it, I don't care. Nevertheless, I want to be able to increase my performance of my antenna system, at least in theory that is what the whole point of this article is about.
Lets have a look at YU7EF over real ground (plotted at my antenna height of 7m AGL). There will be some variation as the antenna sits above a shed roof which looks like a ground to the antenna.
Notice the main lobe in the elevation plane now sits at 11.5 degrees. There is a second lobe at 36 degrees (6.3dB down) and a third at 80 degrees (almost unity gain). Remember! The elevation plot in MMANA shows on half of the antenna (i.e. the flat horizontal line X axis is the centre point of the antenna - not the GND! Therefore if you flip the antenna on the X axis 180 degrees you will find the same pattern, or similar). Why is this important? Your yagi also has a main lobe at -11 degrees too - possibly beaming directly into your house!
If we now raise the antenna to the top of my mast (9m AGL), the main lobe starts to shift downwards to 9 degrees, and the 2nd and 3rd lobes gradually start to break up. This is another reason why antennas at different heights seem to "work better" than others. Your main lobe could be looking into that elusive E layer cloud that your old antenna didn't.
You'll also notice ground gain comes into play, and our antenna now has a gain of nearly 17dBi, or 14.85dB over a dipole. Getting back on track now... let's look at G0KSC and if the claims of a "lower noise" system is true, at least in theory.
I decided to model G0KSC's 6 element on a 6.4m boom design. This is a comparable yagi to my YU7EF which is on a 7m boom. Apparently this design is over 7 years old now, and he has newer designs available. I'm yet to receive an email back from him yet on these designs. Below is the free space MMANA plot.
Figures claimed on G0KSC's website, MMANA calculations are in bold:
FG = 11.63dBi at 50.150MHz
FG = 11.55dBi at 50.150MHz
F/B = 32.27dB at 50.150MHz
F/B = 28.2dB at 50.150MHz
I suspect the F/B is different due the modeling engine in the software. G0KSC includes EZNEC Pro/4 models of all his antennas, and these confirm the pattern and FG I see in MMANA - i.e. close enough.
Of particular interest now though is the pattern. The rear lobe is slightly better. in the elevation plane, the upper lobe at 76 degrees is now 17dB down, however off the rear of the antenna, it never falls below 26dB and peaks at 37db down! A marked improvement over the EF0606. Now over real ground at 7m.
The elevation plots look very similar to the EF0606. Off the rear of the antenna is a marked improvement however, and even the third lobe (straight up and down) is at least 2dB better than the EF0606.
Whilst this is a very basic theory based comparison of the yagis, the proof will be in real on air use of them. I do suffer from rain/dust static at times too, and it's proven that a loop fed system greatly reduces noise pickup. If you live in a country area, you may not notice a difference, however a residential area like mine, the difference may be substantial.
In Part 2 of this post, I think I will build the yagi and evaluate performance "on air". For this to happen, I may need to wait for Justin to email me back with his newer designs. He last told me - "Let me know what size you are looking at, a lot of the designs are 10 years old on G0KSC, a 5el 50MHZ LFA is now over 12dBi for example".