I should first mention that it's my own idea to call this antenna "Lazy H"; it may already have an appellation of which I'm unaware. That's how it looks to me: an "H" that's fallen over on its side. It is basically a stunted vertical dipole loaded at its ends (top and bottom) with a capacitance hat, configured so that the whole thing fits in a vertical plane. It takes up more space that the linear-loaded dipole, which is not a problem in my backyard.
This is the 2nd in my series of 3 candidate short, vertically-polarized antennas I am contemplating for 40 meters. The first was the linear-loaded vertical dipole and up next will be a linear-loaded loop. I am comparing these to each other and to other, full size candidates: in particular the delta loop and inverted-vee. My aim is good DX performance without making too much a statement in the neighbourhood. I do not have a tower in my plans.
At the side is a picture of the EZNEC model and the current distribution. (It's a bit askew so you can see the current profiles.) This is a plumber's delight style antenna since it pretty well has to be constructed from aluminum tubing. It stands 10 meters tall and the loading hats are 7.8 meters tip-to-tip. The horizontal elements can be shortened by making the antenna taller.
There is at least one commercial version of this antenna, the Sigma-40 by Force 12. It is substantially shorter (~7.3 meters) by virtue of the loading coils (doing a search I found that W0SJS took a picture of the feed and coils). I dislike loading coils since, of necessity, they must be located where the current is high, thus having a noticable impact on loss. Force 12 claims no more than 10% (~0.5 db), which is quite good if true.
Another concern I have is that by being short the average height of the Sigma-40 current distribution is several meters lower than in my model. This will increase ground losses and shadowing by obstacles (houses, wiring, local terrain, etc.). From reports it does work well for many people, but that is not enough to stop me from trying to do better. That antenna is also not an inexpensive choice for what it is and does.
My Lazy-H model places the antenna bottom 3 meters off the ground to keep the antenna out of hand reach. If you run more than QRP the tips of the horizontal elements are a serious shock hazard. At this height the top of the antenna stands 13 meters above grade. My choice would be to guy it, though it is possible to make it free-standing with a concrete base. The same is true of the vertical dipole.
Like the Sigma-40 I have chosen to feed my version at the center. This gives a good match to 50Ω but makes construction and tuning challenging since the vertical element will have to be split or fed with an out-of-easy-reach matching network. In either case it is vital to use a high-resistance current (common-mode) choke on the transmission line and to run it for a distance at a right angle to the plane of the antenna. Since this point is nearly 8 meters off the ground, I would have to run it across the roof of my two-story house, then down toward the basement shack. That might be a minor challenge.
Let's first take a look at what EZNEC has to tell us about the SWR across the 40 meter band. It has an even better SWR curve than the vertical dipole. As always I am focussed on CW, yet even so it does fine above 7.2 MHz. At the resonant frequency of 7.08 MHz the impedance is 53Ω.
The pattern is also quite good. It peaks a little higher than the vertical dipole (20°), but still outperforms it at 15° (see note at bottom). The additional gain is almost entirely due to lower ground losses, which are modeled at -3.9 db for this antenna, or about 1.5 db better than the linear-loaded vertical dipole.
The azimuth pattern is almost perfectly omnidirectional, which surprised me since the loading elements are not axially symmetric. I haven't bothered to look more deeply at this particular result although it does intrigue me. Perhaps those horizontal currents are cancelling in the far field.
After playing with various parameters I can say that this antenna is very sensitive to element diameters and lengths and small changes in height. For example, using wire for the loading elements their length must be increased by 5%. In the model the mast is 50 mm (2 inches) and the loading elements are 25 mm (1 inch). Actual construction will undoubtedly use different diameters plus tapered tubing. Because of the mentioned sensitivity I recommend redoing the model with actual tubing choices.
Tuning can be accomplished by adjusting the length of the mast (best choice) or the lower loading element. These can be adjusted without lowering the antenna if tapered tubing is used. The feed point will require breaking the mast with a dielectric or using a matching system such as the gamma match. One idea I have is to have the upper mast half slide into the lower one and insulated with a plastic sleeve. If you do this you may want to model the capacitance between the overlapping sections to avoid surprises. You will probably need to guy with antenna, so choose rope or kevlar, not metal.
Next up will be the linear-loaded loop. I plan to cover both square and delta loops. I'll get to this in a few days as time permits.
NOTE: I did not mention in the previous post that I am using "good" suburban ground in my model. If your urban environment is worse the performance of any vertical will also be worse. Keep that in mind when comparing to horizontally-polarized antennas since they are less sensitive to poor ground. I'll say more about these comparisons after I wrap up the series on short verticals for 40 meters.