Last Modified: April 2, 2014
Contents: Basics; How They Work; Important Considerations; Efficiencies;
Flexibility doesn't go hand-in-hand with efficiency
Auto-coupler/whip combinations are a mixed bag of tricks. They do offer near instant retune which is an attractive attribute for some folks. The tradeoff is poor efficiency when compared to almost any full-sized screwdriver antenna, or about on par with most short, stubby ones. When couplers are used in a mobile environment with electrically short antennas, the resulting high level of RF voltage (≈10kV) at the output of the coupler can and does cause problems. Not so much because someone might grab the antenna when you are transmitting, although this is a concern, but because the antenna components readily available are not designed to handle this much electromotive stress (see Important Considerations below).
Auto-couplers are basically decade-switched LC networks. As a result, they have a design limitation; they cannot adequately match loads close to the feed line's impedance (near 50 ohms resistive). For example, if the radiator the auto-coupler is feeding, represents a 1/4 wave vertical at the operating frequency, it may not be able to present a low impedance to the transceiver in question. At what reactance point this occurs, is difficult to measure or calculate. However, the typical scenario is that the coupler appears to tune properly at low power, but once full power is used the coupler attempts to retune. If this occurs in your installation, knowing the problem exists is half the cure.
It should be noted that auto-couplers cannot be driven at their maximum rating during the tuning process. Most contain circuitry which bypasses the coupler if the drive level is exceeded, but this isn't always the case. Once it does tune, it still may not be able to handle its full power rating. For example, when driving low impedance loads (<16 ohms), the throughput losses are very high; typically greater than 6 dB. While some auto-couplers, like the SG235, can indeed match a 12 foot whip on 80 meters, the internal losses limit the input power to less than 100 watts (it is rated 250 CW). Continued use will eventually destroy the auto-coupler. This is true of Icom's AH-4 as well. While it will tune an 8 foot CB whip to 40 meters, the throughput losses are rather high, and the components will get warm during long transmissions.
Top band operation (160 meters) is all but impossible. For example, a 102" CB whip has an impedance of about -j8,000 ohms at 1.9 MHz. In order to match this impedance, the requisite series L would be huge! There's another reason which is more significant. As noted above, the RF output voltage may exceed 10 kv. This leads to all sorts of high-voltage design problems. Some manufacturers address this by using several relays in series. At the other end of the spectrum antenna currents can exceed the relay's current capability, but this is rare and much easier to deal with.
The most popular auto-couplers are the Icom AH-4 (shown above right), the various SGC models, the various LDG models, the Yaesu FT-40, the Alinco EDX-2, and the various MFJ models. There are at least five other after-market manufacturers and/or resellers advertising in the pages of QST.
A little closer to home, there are two major differences which need to be mentioned. Some units are designed to match coax loads, and thus have a limited SWR matching range, typically 10:1 or less. An example is the LDG AT-7000 (rated 12 to 1,600 ohms) shown in the left photo.
There seems to be as much confusion about internal couplers as there are external ones. Auto-couplers built into modern transceivers are only good for mismatched loads up to ≈3:1 SWR. Therefore, they're only suitable for extending the bandwidth of a monoband mobile antenna, but even this fact is questionable. For example, on 75 meters a decent quality mobile antenna when properly matched will have a 3:1 bandwidth of about 15 kHz. Most built-in couplers will double this bandwidth, and a few might even triple it. In either case, it is a questionable improvement.
If you want to know more about matching, read the highlighted article.
How They Work
They're called auto-couplers because that is what they do. They don't tune the antenna per sé, they simply perform a conjugate match between the impedance of the feed line and the source (transmitter). In simple terms conjugate means to join together or couple, hence the name coupler. They accomplish this by using a series of inductors and capacitors. The right photo shows the inner works of an SGC coupler. There are 38 relays, 12 inductors, and about 100 capacitors, all controlled by a Motorola CPU and proprietary algorithm.
With two exceptions, auto-couplers are all LC or Pi configuration and are configured as low-pass networks (series L, shunt C). Relays are used to cascade switch the inductors and capacitors in and out of the circuit, and can place the C (shunt reactance) at the output or input (both in a Pi configuration) to match higher and lower than line Z impedances. If you don't understand how all of this gets accomplished, Chapter 17 of the ARRL Handbook is a good resource. If you want a more detailed explanation, Dr. Walt Maxwell's "Another Look at Reflections" (available to ARRL members on-line) is a must-read.
Again in simple terms, a coupler is a matching network which transforms the antennas input impedance to that of the transmission line. Due in part to the large reactances we're dealing with, auto-couplers used in a mobile installation must be mounted as close to the antenna as possible. That is to say, inches, not feet!
You cannot use coax for this connection either, as even a one foot piece of coax will reduce the efficiency by 30% or more. The reason? Coax has about 25 pf of capacitance per foot. The capacitance of a typical HF antenna ranges from 20 pf to about 45 pf depending on its length and frequency of operation. Since our auto-coupled antenna is essentially a base loaded vertical, placing 25 pf to ground will shunt a large portion of the RF to ground. Here is a more detailed article by Owen Duffy, VK1OD. It better describes the reduction in efficiency caused by shunt capacitance. This interaction should not be confused with using shunt reactances to match a low Z HF antennas to a 50 ohm feed. That is a different animal altogether.
The robustness of the RF ground is a major consideration. Classic examples of poor RF grounding are when the coupler cannot find a match, and/or is resetting itself during transmissions. One inch braid may work if the ground lead is short (less than six inches or so). It is important to remember that the ground side impedance must be much lower than the radiating element side. If not, the coupler will have a hard time deciding whether to match the radiating element or the body of the vehicle. This is especially important on the higher HF bands, and particularly on 6 meters.
The connections need to be very robust as well. The liberal use of I/O star washers and stainless steel fender washers is a must. Most importantly, the ground for the coupler cannot be coincident with any other ground! Failure to adhere to this will cause you serious RFI problems, and may indeed cause damage to your auto-coupler.
In the right photo is the insulator off of an old GE Master ballmount. You can see the burned-out groove between the edge of the (removed) ball and one of the mounting bolts. By the way, coating the insulator with car wax exacerbates the problem. If it needs weatherproofing, use a good quality, high voltage lacquer, which you can buy it at most hardware stores for about $4 for a 12 ounce spray can. As an alternative, use Rain-X.
At left, is a GeoTool® insulator designed specifically for use with an auto-coupler. It was featured in a March 2004 QST article written by Bob Lewis, AA4PB, a frequent contributor to eham.net. He describes how he conquered his high voltage problem, using the GeoTool insulator as shown.
The insulator shown at right comes from Breedlove, and sells for $40. Machined with standard 3/8x24 threads top and bottom, it is good for well over 20 kv. It even has wiring connections for both ends as well. If you use an auto-coupler and whip combination, these are the insulators to use.
Again, high RF voltages are the major drawback to auto couplers. It requires that the mounting hardware, especially the base insulators, to be able to withstand 10 to 15 kilovolts. Therefore, standard ballmounts, and nylon washers will typically fail post-haste, especially so if moisture (or car polish film) is present.
As note above, the efficiency of an auto-coupler/whip isn't all it could be. They are after all, roughly equivalent to a base-loaded antenna. However, there are ways to increase their efficiency. Obviously, increasing the overall length by using a mast could be done, but there are height limitations to be considered. A good alternative is a cap hat, but this raises other concerns.
In order to be effective, the cap hat must be mounted at the very top of the antenna. This is difficult to do when using a whip make from 17-7 stainless steel—they're just too flexible! Therefore, a solid mast is the only answer. Aluminum rod is fairly inexpensive, and available from Commerce Metals, Speedy Metals, Metals Depot, and similar suppliers. It requires drilling and tapping, but any good metal shop can do that for you.
A good rod size to shoot for is 5 feet in length, and 5/8" OD. Add a 5 foot cap hat as shown in the photo at right (described in the article), and the electrically equivalent length will be nearly 12 feet! However, there is one more consideration, and that is a stout mounting arrangement. How this is accomplished would be reliant on a lot of variables—the vehicle, and the mounting position to name two. These issues, however, should be weighed against the doubling of the overall efficiency!