Air Core Coax Chokes: Good, Bad and Ugly

Ugly coax choke, but it worked

To give you an idea of how effective a poorly made air core coax common mode choke can perform I'll refer you to one of the oldest articles in this blog. It's wound with spliced together lengths of RG58 I pulled out of a junk box. It's ugly yet it worked beautifully, in that it solved the common mode problem I had with an end-fed antenna.

Common mode chokes on antennas can be very beneficial. All kinds of ills result from RF currents flowing on the exterior of the coax, on both transmit and receive. However it is surprisingly easy not to recognize the problems since they are often blamed on other causes such as proximity to the antenna, a poor receiver and cheap consumer electronics, among others.

It is usually easy to install a common mode choke so there is no reason not to do so even if there are no obvious common mode ills present. The choke presents a high impedance on the coax outer shield surface to prevent conducted current on the coax. The impedance can be a resistance or a tuned circuit, with the former having the most predictable characteristics and being broad band and the latter being narrow band with relatively unpredictable characteristics.

I am not here to tell you how to choose or build a common mode choke. For that I'll refer you to the experts. Probably the best explanation, analysis and recommendations was written by K9YC. There you'll learn that the best chokes use large ferrite toroids, wound with the coax itself or even THHN electrical wire. Instead I'll speak to a commonly used and admittedly inferior solution: the air core coax choke -- in this article I'll often call it a "coax choke" for brevity.

I've used many over the years, sometimes because I was cheap and it's easy to make. It is nothing more complicated that the coax wound into a coil. The coil inductance itself has choking properties, as does any inductor. However most of the choking is courtesy of the distributed capacitance among the turns that in combination with the coil inductance forms a resonant circuit. If the circuit has a high impedance at the frequency of operation (resonant) it can work very well indeed.

Unfortunately it's easy to get it wrong or to have unreasonably founded expectations. Most hams don't have a two port VNA (vector network analyzer) or the expertise to use them properly. I don't have one either. The question is: can we rely on an unmeasured coax choke to be effective?

Before going further we have to review what we mean by effective. What effect? How much? Without a clearly defined problem to solve or an operating objective to be met there is the risk of doing too little or too much. At least with the latter you can be successful though perhaps at a price or time investment that isn't strictly necessary.

To study effectiveness let's briefly list the potential benefits of common mode chokes:

• Antenna impedance: Common mode on the coax will alter the antenna feed point impedance. This may be unnoticed since when we tune the antenna (e.g. beta or gamma match) that component of the impedance is accounted for. However that may be insufficient since the common mode impedance can vary widely with frequency, especially when a reactive choke such as a coax choke is employed.
• EMC (electromagnetic compatibility): RF travelling down the coax will radiate. In our dense neighbourhoods this places the "antenna" closer to us and our devices. There is an increased probability of EMI to those devices and received EMI from those devices. Expect the tower and other cables to join the fun since the common mode current will couple to those parallel conductors.
• Antenna pattern: Even a small amount of radiation from the coax can ruin otherwise good antenna directionality. When the F/B is 20 db or greater it takes very little undisciplined radiation from reducing that to 10 db or worse. Routing the coax perpendicular to the antenna reduces induced RF but does little to suppress conducted common mode RF.

Coax chokes are very attractive since they are cheap, easy to construct and can work well in select situations. Let's consider them with respect to the above metrics.

• Modest choking is sufficient to avoid difficulties with the antenna impedance. Once the choke impedance is above ~500 Ω further improvements have negligible impact.
• EMC is not a large problem in my station since the towers and antennas are far from the house and my neighbours are far away. This isn't typical of most hams. I can get away with a little leakage.
• Very little choking is needed to protect antenna gain. Directionality requires better choking. For an omni-directional antenna this is of low importance. I am more interested in gain than directionality since my primary interest is contesting where it can be beneficial to attract callers from directions other than where the antenna is pointing. 
• For low band receive antennas such as Beverages and vertical arrays the use of high impedance common mode chokes is mandatory. These require ferrite cores to cover both 160 and 80 meters. Besides which a coax choke at 1.8 MHz is quite large and difficult to build, and can actually be more expensive than a ferrite core choke.

A common impedance objective for common mode chokes is at least 5000 Ω, or 100× the nominal antenna feed point impedance. This is difficult to achieve with a coax choke. You can do it on a single band but may need to be optimized with the aid of a two-port VNA. They are narrow band since their resonance (LC tuned circuit) is sensitive to construction technique.

"Scramble wound" coax choke that I made in ~1987 for a TH6: bad!

The scramble wound choke pictured above replaced the original burned out BN86 balun on my TH6 ~30 years ago; the Hy-Gain voltage balun is in any case a poor choice for a common mode choke. Asking a coax choke to be effective across the 2:1 frequency range of a tri-band antenna is inordinately optimistic. I could say the choke "worked" in that I didn't have any obvious common mode problems. It is likely that it had a choking impedance of less than 500 Ω on one or more of the three bands. Depending on circumstances and expectations even that small an impedance can be considered effective.

There are impedance measurements for a variety of coax chokes available. Most hams would prefer to rely on those rather than do their own measurements, in the hope that they deliver the published performance. The important parameters are diameter, coax type (outer conductor OD) and winding style (solenoid, scramble wound, etc.). Scramble wound is the easiest (see picture above) but has unpredictable performance due to the unpredictable L and C values.

One resource I've turned to many times is the measurements of various air core and ferrite core coax chokes by G3TXQ (SK). I've extracted part of the table below since I cannot assume that the web site will be around forever now that he's passed on. This is the data by which I recently made chokes for my new 15 and 20 meter stacked yagis.

These are solenoid wound coax chokes. Always wind air core coax chokes in this manner and never use scramble winding. That's the only way to achieve predictability performance. Notice in the table how difficult it is to make a high impedance coax choke that covers more than one HF band. You can do reasonably well if, like me, your station and operating style can tolerate imperfection.

For the 15 meter yagis I used 5 turns and 6" diameter of LMR400UF. On the above chart you can see that I interpolated between two known designs to get one that has the diameter and turns count that I prefer for the chosen coax. The PVC strips and cable ties hold the turns in a solenoid form. Tape was used while winding the coax to keep the diameter consistent and to discipline the turns. A temporary form can be used if you have one of the desired diameter.

Solenoid wound single-band 15 meter coax choke: good!

Pay close attention to the bending radius specifications of the coax before winding your choke. I prefer to use RG213 or the ultra flex version of LMR400 since they are more flexible. Greater care must be take with foam dielectric coax to prevent the centre conductor from pushing through the foam and shorting to the outer conductor or altering the impedance.

These danger of excess or repeated bending can take months or years to manifest so build carefully and don't rely on a one time measurement. Avoid bending the coax more than once, especially with LMR400 with its solid centre conductor, since the minimum bend radius is far higher for multiple bends. Study the mechanical properties specs and use accordingly.

Here's the same choke installed on a 15 meter yagi. One of the PVC clamps doubles as a boom clamp. Yes, I do antenna and tower work in winter! Shortly after the picture was taken the antenna was trammed to 100' (32 m). The yagi is side mounted and fixed northeast as the lower yagi in the stack.

Speaking of stacking, if you use common mode chokes of any variety in a stack it is important to use the same choke (or current balun) on all yagis in the stack. Otherwise there will be a phase shift. Unless you compensate for the phase shift, gain and lobe formation for the stack will suffer. For these coax chokes I measure the length and type of coax, data which I'll use to ensure the yagis are fed in phase.

As much as I love coax chokes I avoid them for multi-band antennas and receive antennas. Almost all my tri-band yagis and lower frequency antennas have commercial ferrite core chokes that have high impedance across the bands the antennas cover.

My Beverage receiving antennas use binocular ferrite core 1:1 transformers at the feed points. It is also good practice to use them at the switching system ports and at intervals on the transmission line, especially if it parallels a Beverage.

Breaking up the system in this manner keeps common mode currents at bay thus protecting the high directionality of the antennas. You can also wind the coax on a suitable ferrite toroid. It is very difficult to achieve high choking impedance on the low bands with an air core coax choke.

In conclusion, go ahead and use a coax choke if it suits the application. Remember to wind them properly, use reliable specifications or measure them yourself with a two port VNA and try to limit their use to one band rather than two or more. But don't expect more from them than they can deliver.