Antenna Cap Hats
Last Modified:
Contents: Basics; The Cap Hats; Building A Cap Hat; Efficiency Increase; Conclusions;
Cap hats are a mixed bag of tricks to be sure, but they can be effective in more ways than one.
Those of you who subscribe to the ARRL publication QEX, will probably recall the series of articles by Rudy Severns, N6LF. The articles contained a lot of empirical data with respect to vertical antennas, and their requisite ground plane requirements. If you haven't read the articles, you should, as the data is rather enlightening. Copies of the articles may be downloaded from Rudy's web site.
These articles were not aimed at the mobile operator, but the data does explain the ramifications of an inadequate ground plane under a vertical antenna. Suffice to say, the lossier the ground plane, the lower the efficiency, and that's exactly what we have in a mobile installation; a very lossy ground plane. I should point out that any vehicle is an inadequate ground plane at HF frequencies. Fact is, the body of the vehicle acts as a capacitance to the surface under the vehicle, which acts as the ground plane, albeit very lossy.
An important point needs to be made here. The body of the vehicle is a much better conductor of RF, than the surface under the vehicle. When we mount the antenna low, on a trailer hitch mount for example, a goodly portion of the return current is made to flow through the surface under the vehicle which increases ground losses. If we mount the antenna higher, atop the quarter panel say, more current returns through the body, so ground losses are reduced just like they are when we use elevated radials versus ground laid radials. Even so, ground losses in a mobile installation are much higher than those encountered in a typical base station installation.
A follow up article, written by Bob Zavel, W7SX, appeared in the July/August 2009 of QEX, entitled Maximizing Radiation Resistance in Vertical Antennas. Part of the article covers top loading. Top loading is a methodology which increases radiation resistance, hence efficiency, even if the ground plane is substandard; seemingly a ubiquitous vertical antenna shortcoming. This article is also a must read especially if your urban bound!
The conclusions at the end of Bob's article are well founded. Of specific importance are the following points; to paraphrase: The radiation resistance (Rr) of a vertical antenna is a function of the physical height (overall length), and the current distribution along that linear height; The efficiency of a fixed-height antenna can be optimized by orientating the maximum current point at the half way point (height) of the antenna; Series and parallel losses (ground losses and stray coupling losses respectively) are always present, with series losses the most severe; Lowering of ground (series) losses, and raising radiation resistance will result in higher efficiency, but the latter is easier to accomplish. These conclusions support the thought that reducing ground losses, and maximizing radiation resistance are the two paramount objectives in achieving maximum performance from a base station vertical. Or from an HF mobile antenna!
The mobile equivalent to top loading is the cap hat. As their name implies, cap hats add capacitance to the top portion of the antenna (the whip). Properly placed, they can effectively increase the electrical length, by moving the current maxima toward the center of the antenna.
Tech Talk: Whatever capacitance any given cap hat adds, is the same no matter where it is placed. However, whether or not a cap hat increases the effective length and/or increases the radiation resistance and/or increases overall losses, depends on where (how high above the coil) the cap hat is placed. For example, when placed too close to the loading coil, the capacitance can have a detrimental effect on the coil's Q, and will indeed produce an increase in the measured input impedance. For example, the left photo depicts a cap hat incorrectly installed. The input impedance and bandwidth will indeed increase in this example, however the changes are due to increased coil losses, and not by an increase in radiation resistance (Rr). Therefore, the following assumes the cap hat is mounted at the very top of the antenna, and thus the noted increase in input impedance is a positive one, not a negative one.
Because the effective electrical length is increased, the radiation resistance also increases, albeit slightly. Physically increasing the whip length will do the same, but there is a practical limit to how long an HF mobile antenna can be (height above ground). By adding a cap hat, the overall length can remain the same (or less if the cap hat is large enough), while radiation resistance increases. That's the good part! The bad part is, they increase wind loading and complexity.
Tech Talk: (I mention this in the Antenna Mounting article, but it bears repeating here.) There is a difference of opinion about where to add length (capacitive reactance) to a mobile antenna to increase efficiency; below the coil, or above the coil. The truth is, it depends. Under ideal conditions, we want to add the length above the coil (whip and/or cap hat), as this raises the current node higher than adding length below the coil (mast). However, the optimal location of the loading coil within the overall length (whatever it is) of an antenna, is reliant on several factors. They include (but aren't limited to) the combined ground losses, coil Q, and to a lessor degree, the effective diameter of the mast and/or whip. Generally speaking, as the ground losses increase, so does the optimal height of the coil. Thus, if the ground losses are excessive, one might see a slight improvement in field strength (you can't use the input impedance as a measurement) by increasing the mast length. If ground losses are relatively low, the best improvement in field strength will always occur by adding length (capacitance) to the whip; either by lengthening it, or by adding a cap hat to it.
The actual design of the cap hat is important too. If the cap hat is enclosed like the wheel shown at right, the effective length is increased approximately twice the cap hat's diameter. Cap hats without the peripheral connection, like the DX Engineering unit shown at left, are equivalent to approximately 60% of their diameter.
Cap hats typically increase bandwidth. The amount of the increase is based on a myriad of factors, most of which you can't measure directly. Here too, you have to be careful about applying a change in input impedance (SWR over any given bandwidth), as a positive indication. In fact, the bandwidth of an antenna with incorrectly mounted cap hat (directly atop the coil), will be greater than a correctly mounted one, and with far less efficiency as well.
There's another drawback which might not be apparent. No matter how well an antenna is built, it can be destroyed if you smack it hard enough, and it doesn't take a big limb in most cases. This is especially true if the mast holding the cap hat is solid, and not flexible like a whip. It's your basic, high school physics, lever law. Yet, every currently sold cap hat is designed to mount atop a solid mast. And, they're typically mounted too close to the coil to have a positive effect.
There's one potential problem depending on the design, and that's corona discharge from the tips of the cap hat. At the tip of an HF mobile antenna, the RF voltage can easily exceed 10 kilovolts. The upper left photo shows a stock DX Engineering cap hat. You'll notice the rods are tipped with little plastic protectors. The rods themselves are polished round, but because of their small diameter, it is possible to have corona discharge off the tips of the rods, especially if you run high power. You can get around this problem by purchasing three of the optional 48 inch rods, and bend them into circles as shown in the right photo. This also increases the effectiveness (capacitance) of the cap hat, but making it more susceptible to snagging by an errant limb.
Tech Talk: There are at least two antenna manufacturers who sell what amounts to a tubular cap hat. They're essentially a large diameter whip, although solidly constructed, and rather heavy. As noted above, they do indeed increase the capacitance above the coil. Unfortunately, a major portion of the added capacitance is within the coil's field. This fact reduces overall efficiency rather drastically. It's been shown by actual field strength measurements, that antennas equipped with tubular cap hats are less efficient than ones with just a whip of the same length.
As alluded to above, there are a few rules you have to follow if you want them to be worth the effort and hassle. The most important rule is, you have to mount the cap hat as far away from the coil as possible. That is, at the very end of the whip, but obviously that isn't an easy task for several reasons.
First, most cap hats are designed to be attached using the ubiquitous 3/8x24 threads. Because most whip material is too flimsy to support even a modest cap hat, too many folks mount theirs directly (and incorrectly) atop their coil housings as shown above. Since we want to maximize the gains a cap hat can give us, we need to find a way to make one light enough to be mounted at the top of the whip.
With the help of Ken Muggli, KØHL (who did the drawing and machining), I came up with an easy to make a cap hat and hub, which is light weight, has relatively low wind loading, yet is flexible enough to absorb some physical abuse without overly stressing the antenna it's attached to.
Ken Muggli is a first-rate draftsman, and a fine machinist as you can see. You can click on the photo at left (or drag and drop), and it will open up to full-size. It's 400 k, so it'll take a moment to load, but should print out in very fine detail. A completed hub is shown at right.
The hub is made from 6160 extruded aluminum bar, 1.5 inches in diameter,and just shy of 1.2 inches long. The peripheral holes are .125 (1/8) inches, and the set screws are all 10x32; six at 1/4 inch long, one 3/8 inch long. The hub itself weighs just one ounce. Material for the hub was purchased from Speedy Metals. The set screws were from Micro Fasteners.
The main support hole is .200 inches, and requires some explanation.
While Ken and I were playing with the basic designs, I went through about $50 worth of 1/8 inch, brass welding rod. While not sturdy enough for daily use, it is easy to solder and bend. This made the various designs easy to construct, but not always as effective as we would have liked. One very important attribute needed to be discovered up front, and that was how to support the cap hat effectively at highway speeds, yet offer some flexibility in case you snagged a low-hanging limb.
Tech Talk: Believe it or not, there is only one supplier of 102 inch, 17-7 stainless steel whips, no matter where you buy one. They start out life as rolled wire, .210 inches in diameter. The wire is straightened, and ground into the common size, and shape we all know. Starting at approximately 60 inches from the base, the wire is taper ground so the tip is .100 inches in diameter. A swaged brass 3/8x24 fitting is attached to the bottom, and a small chromed, brass tip is added at the other end.
It turns out, that right where the taper begins is an ideal place to attach a cap hat hub, hence the .200 inch diameter hole through the hub. If it is mounted any higher, the whip isn't strong enough to support the cap hat, and if it is any lower it isn't as effective as it can be. Although I hate to admit it, the exact corresponding spot was strictly by accident, and not by design, per sé.
One design goal (for me at least), was to construct a cap hat, which would allow 17 meter operation, without any loading. That is to say, with the Scorpion's coil completely collapsed. As luck would have it, a three loop design, mounted 60 inches away from the coil, met the design goal. The fact the whip's taper begins at 60 inches, was serendipitous!
I tried about a dozen designs trying to come up with the most effective design consistent with a minimum of wind loading, and of course light weight. I tried wheels, loops, cones, multiple shorting circles. I even did two solid disks, but the wind loading for those beasts was enormous! Some were better than others, but most had one drawback or another. So, what you see in the left photo, is the culmination of all of those designs.
One could use four loops instead of three, but the improvement (in capacitance) would be very slight (less than 10%). However, the wind loading would be about 33% more, and the .200 inch whip would have to be much larger.
The individual 17-7 stainless steel wires, .125 inches diameter (1/8") by 60 inches long, were purchased from Small Parts. The three wires, plus the hub, weigh a total of just 10.5 ounces. They're very springy, so care should be taken when installing the wires in the hub.
The final design is stable at highway speeds (75+). The angle the whip bends back, it roughly equivalent to just a 102 inch whip alone. Yes, there is more loading on the top of the antenna, but the to and fro oscillations commonly seen with 102 inch whips at highway speeds, is not evident with the design. This fact alone (I believe), lessens the overall stress on the antenna, but only time will tell.
What I was trying to do here was somewhat simple mechanically, but mathematically rather diverse. I could get technical at this point, but I suggest you read what Tom Rauch, W8JI, has to say here, as he does a much better job of explaining the intricacies of adding capacitance above the coil. Cap hats, in other words.
The main thrust wasn't necessarily to improve the radiation resistance, but I'm sure some improvement did occur as the input impedance increased 2 to 4 ohms across 80 through 17 meters. It was, in fact, an effort to reduce the overall height of the antenna to keep it under the proverbial 13.5 feet, yet maintain at least the same (if not better) efficiency level. From all indications, both measured and empirical, I've accomplished my goal.
Note there is no whip showing above the cap hat, and this is by design. During the empirical testing, it was discovered that any additional whip showing above the cap hat had very little effect on the resonant point, or the input impedance. There is no corona ball either, as the effective end(s) of the antenna are the rounded loops. To date, no corona discharge has been observed.
The increase in antenna efficiency garnered for any given cap hat installation is both difficult to measure or calculate, as the resistive losses (series and parallel) will be different for every installation. Further, the overall size of the cap hat, its height above the coil, and whether its ends are electrically connected, all play a part. While not definitive as a properly conducted field strength measurement, the following is at least an indication of increased efficiency.
The antenna in question is a Scorpion 680, mounted in the bed of my Honda Ridgeline, directly atop the main cross member which passes just behind the fuel tank. The 102 inch whip is your basic CB item. Adding in the QD, and the corona ball, its overall length 105 inches. The Scorpion's 3 inch diameter coil is wound with #10 silver plated wire, at 6 turns per inch.
On 80 meters with the cap hat installed, ≈70 turns show above the contact ring. With the whip, ≈80 show; On 40 meters with the cap hat installed, ≈16 turns show above the contact ring. With the whip, ≈20 show; On 20 meters with the cap hat installed, ≈4 turns show above the contact ring. With the whip ≈6.75 show; On 17 meters with the cap hat installed, Ø turns show above the contact ring (antenna fully collapsed). With the whip ≈4 turns show.
There are corresponding increases in the input impedance as well. Depending on the band, the unmatched input impedance increases 2 to 4 ohms. Obviously part of the increase is caused by the slight decrease in coil losses, but for the most part one can assume the radiation resistance is the major portion of the increase.
The measured receive signal strength difference between the cap hat configuration versus the whip, varied between ≈3 dB (80 meters), and ≈6 dB (20 and 17 meters). I believe this is proof-positive that the cap hat design herein, does indeed increase efficiency significantly enough, to offset the mechanical aspects of the design, let alone the reduction in antenna tip height above ground (≈13 feet versus ≈16 feet).
Cap hats aren't for everyone, but if you want the best performance (read that as highest efficiency), out of an HF mobile antenna, they are the only way to accomplish the goal, yet maintain some modicum of practicality. They have some drawbacks as any abbreviated system will have, but I am convinced the pluses outweigh the minuses.
For those who might ask, I did do several whip length comparisons as mentioned above. I used an MFJ-1956, 12 foot, telescoping whip. Once the antenna was at resonance, and the input impedance measured, I removed the cap hat, installed the whip, and extended it to the exact same resonant point. In most cases, the R value was slightly higher (two to four ohms) with the cap hat installed, when compared with the whip. This is close to the accuracy fuzz of the MFJ-259B antenna analyzer I was using. One might argue that the difference was capacitive loading to the body of the vehicle (extra loss), or a slight increase in radiation resistance (a little gain). Either argument is all but moot. What isn't moot, is the reduction in overall length, which for most folks is a worthy goal.
A full quarter wave vertical antenna (no loading coil), mounted on a vehicle, should have an input impedance, at resonance, of 36 ohms plus whatever ground, capacitive, or resistive losses are present. Using the aforementioned whip, it is possible to resonant the Scorpion 680 on both 20 and 17 meters, with the coil fully collapsed (fully shorted out). So resonated, the unmatched input impedance on 20 meters was 40 ohms, and on 17 meters, 39 ohms. These measured figures are very close to the theoretical input impedance, plus the calculated ground loss using the formulas published in the ARRL Antenna Handbook.
As stated above, the cap hat, when mounted 60 inches above the fully-collapsed coil, resonates the Scorpion antenna on 17 meters. The unmatched input impedance measures 43 ohms, or 4 ohms better than the equivalent whip. The reader can draw his/her own conclusions.