5
$\begingroup$

Making a common mode choke out of bifilar windings has proven rather difficult. For me the difficulty lies in getting the impedance to remain a constant 50 ohms from on the HF bands. I've tried 14 gauge enameled and I've tried 18 gauge PTFE. The wires seem very sensitive to any kind of change in position.

I've been modeling the choke after the one shown on this tutorial. Also what's the deal with that design? It's like someone took a 4:1 guanella and just cut the connection that makes it 4:1. Turning it into to parallel 1:1 chokes.

Is there any disadvantage to simply using coax windings instead of bifilar? Can I just wrap 11 or 12 turns of RG-316 around a toroid? How much more choking impedance do you get from a bifilar choke?

The application here primarily on the 40m band. I was looking to build a simple in-line choke (SO-239 on both ends) as well as choke "balun" (SO-239 on one end and balanced terminals on the other). I was also entertaining the idea of a choke "unun" for use with an elevated quarter wave vertical. I would have a top lug connected to the center conductor; the shield would split off to two side lugs.

$\endgroup$
6
$\begingroup$

In most multi-band balun applications, there is rarely a need to maintain a perfect 50 ohm impedance within the balun. The feedpoint or input impedance is varying widely so another impedance bump in the mix typically has no detrimental effect.

I highly recommend the use of coaxial wound over bifilar style for a 1:1 balun. Comparatively, the coaxial and bifilar design can be made to have equivalent common mode choking.

The bifilar style will couple some of the energy traveling on the parallel lines into the core - much like getting a parallel transmission line too close to metal objects. This creates an unnecessary loss and it promotes heating of the core due to the complex permeability of the core material.

It is also very difficult to fabricate a true bifilar 50 ohm line. The design is so touchy that even minor changes in the wire coating thickness or failure to keep the lines adjacent will cause impedance discontinuities. Due to this marginal construction, the voltage withstand rating of this type of transmission line is much lower than what can be attained through the application of a coaxial transmission line.

The coaxial wound style will keep all but the common mode current from creating a flux in the core. This is the ideal behavior of a transmission line transformer of this type. Do estimate the losses in this length of transmission line in order to project the effect on core heating. If heating of the core due to coax loss is a design constraint, consider using higher impedance coaxial cable (e.g. 75 ohm) as this will reduce the loss compared to the same size/class of 50 ohm cable.

Selection of the correct core material is essential. Generally, type 31 is suitable for the lower HF bands and type 43 for the upper HF bands. Look at the complex permeability curves and use a material where the $\mu$''s is greater than $\mu$'s for the frequencies of interest.

You may find some practical information regarding the number of turns of coax vs the core material on the G3TXQ (now an SK) site. Here are some examples of his analysis:

enter image description here

enter image description here

The K9YC 2018 Cookbook also features type 31 material balun designs.

$\endgroup$
  • $\begingroup$ Jim has a new PDF revised in 2019, linked to in my comment above. $\endgroup$ – Mike Waters Mar 11 at 15:46
  • 1
    $\begingroup$ @MikeWaters Thanks, Mike but in your link, chapter 7 is not included. My link is his referenced chapter 7 contents specific to choking baluns. $\endgroup$ – Glenn W9IQ Mar 11 at 15:55
  • $\begingroup$ Yours was the correct one. From Jim: "I no longer recommend coax wound through multiple cores. I'ts simply not practical to wind chokes that way and get anything close to the same result every time -- turns must go through the core in the same order, a scrambled turn cancels a turn, turn diameter matters a lot, and so on. ... $\endgroup$ – Mike Waters Mar 11 at 16:07
  • 1
    $\begingroup$ how important is it to use RG400 over something like RG316? RG400 is about $2/ft whereas RG316 is about 60 cents/ft. $\endgroup$ – Paul Mar 11 at 18:57
  • 1
    $\begingroup$ @Paul The RG-400 is generally lower loss and handles more power although these specs can vary greatly between manufacturers. Also watch bend radius specs. So it will depend on your application and the specific supplier of the coax. If you care to supplement your question with application details, I might be able to update my answer with some of those calculations for you. $\endgroup$ – Glenn W9IQ Mar 11 at 21:07
4
$\begingroup$

The advantage of using coax for a 1:1 choke-balun is the constant Z0 of 50 ohms. As can be seen in the following graphs, the only time a 1:1 choke-balun accomplishes a 1:1 transformation is when it sees 50 ohms at its output. Any other impedance at the output causes an impedance transformation because the SWR is not 1:1. enter image description here

$\endgroup$
  • $\begingroup$ Nice. And to clarify, this is true of either a choke wound with coax or a choke using bifilar windings, correct? $\endgroup$ – Mike Waters Mar 15 at 1:07
  • 1
    $\begingroup$ @MikeW, yes assuming the bifilar windings have a Z0 of 50 ohms which is sometimes a rash assumption sans measurements. The beauty of 50 ohm coax is that an assumption of 50 ohms is not rash. $\endgroup$ – Cecil - W5DXP Mar 15 at 2:23
  • $\begingroup$ This is a very important and useful answer. It illustrates the fact that this is a "transmission line transformer" whose behavior depends on both the operating frequency in terms of the lengths of the transmission lines comprising the transformer and the impedances at the two ports. $\endgroup$ – Brian K1LI Mar 15 at 12:59
2
$\begingroup$

I strongly recommend "Building and Using Baluns and Ununs" by Jerry Sevick, W2FMI. I have built dozens of coaxial- and bifilar-wound transformers and chokes for use in high-power amplifiers and antenna systems based on the easy-to-understand information in this book. Specific mention is made of how to make reproducible bifilar-wound units over a wide range of impedance levels. They differ in important ways from the units shown in the "splash screen" of the video you cite.

It's certainly easier to use coax to build a 1:1 choke for use in a coax-based system. Rather than adding separate units to choke currents on the shield, I find it cheaper and more convenient to wind as many turns of the feedline as possible through stacks of "large" diameter ferrite toroids. This also avoids the possibility of water-intrusion and other failures that can occur at the coax-coax joints needed to insert separate units. K9AY describes this configuration in, "Ham's Guide to RFI, Ferrites, Baluns, and Audio Interfacing."

In the category of "better is the enemy of good enough," it's important to understand the problem you're trying to solve. While operating recently in a DX location, I had to make do with a heavily compromised antenna installation. Observing unstable SWR readings and watching the mouse wander all over the screen of my laptop when transmitting, I surmised that RF was traveling down the outside of the coax shield. Simply winding a few turns of the coax feedline into a coil lying on the ground solved the problem.

$\endgroup$
  • $\begingroup$ Actually, not everything in Jerry's books are accurate. See the comments by Dave Cuthbert and Dale Hunt there, as well as Owen Duffy, ex-VK1OD (there are other reliable sources). K9YC's 2018 publication is the best, either in print or on the Web. But this is not to say that W2FMI's books are worthless. $\endgroup$ – Mike Waters Mar 15 at 17:09
  • $\begingroup$ Having said what I did about W2FMI's books, I have never seen the book you mention, Brian. Sorry if I came across as a little harsh. +1 for your answer. :-) $\endgroup$ – Mike Waters Mar 17 at 12:20
  • $\begingroup$ @MikeWaters No worries, Mike. $\endgroup$ – Brian K1LI Mar 17 at 19:22
  • $\begingroup$ @MikeWaters The book I cited doesn't deal with the issues surrounding off-center-fed dipoles. Fifteen years after the date of Sevick's book, during which new materials may have been introduced, VK1OD seems to criticize misuse of Sevick's measurements, not the measurements themselves. $\endgroup$ – Brian K1LI Mar 17 at 19:56
  • 1
    $\begingroup$ aw man i forgot how long and wordy his dam videos are. i end up just skipping to the part where it's done. there are a lot of people that follow these videos religiously. and if there is a any misinformation in there it can just lead to more confusion when people start regurgitating the contents of that video onto other forums. $\endgroup$ – Paul Mar 19 at 14:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.