eHam.net - Amateur Radio (Ham Radio) Community

Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net



[Articles Home]  [Add Article]  

How a 1:1 Guanella-Balun (Current-Balun) Operates

jsodus (KM3K) on July 10, 2010
View comments about this article!

How a 1:1 Guanella-balun (current-balun) operates.

In the spirit of the hobby, I pass along what I have learned and don’t profess to know all the answers.

1. The primary reason for this work is to explain how a particular Transmission-Line-Transformer (TLT) known as a 1:1 Guanella-balun operates. Three other commonly used names to describe this balun are: “current-balun”, “isolation-balun”, and “choke-balun”; I’d suggest committing all four names to memory.
The 1:1 Guanella-balun is the basic building block for other types of types of Guanella-baluns or ununs (a Guanella-unun is just a Guanella-balun with the grounds in different places; a similar statement for Ruthroff-ununs cannot be made).
Also, I’ve included ‘background’ info, which I thought may help in understanding.

2. About the word “balun”….
It means “balanced-to-unbalanced”; Guanella (a Swiss engineer in 1944) needed to match a balanced (push-pull) amplifier to an unbalanced load; hence the name.
However, these days we generally are going unbalanced-amplifier to balanced-antenna load.
So, “unbal” would be a more appropriate term but we are stuck with balun.

3. It is most important, for the time being, to absolutely and completely put out of mind the concept of the conventional transformer and its flux-linkages.

4a. This section explains the very first 1:1 Guanella-balun.
As part of his development work in 1944, Guanella coiled a transmission-line.
The historical record, that I’ve accessed, is meager on the details of the transmission-line that he used but it is reasonable to presume he used coax (Heaviside had patented coax in 1880 & Bell Labs began making coax in 1929).
In coiling the line, Guanella intentionally formed a choke, whose conductor was the exterior surface of the coax’s shield. Recall that skin-effect is operative on a coax-shield at radio-frequencies, so the transmission-line current is on the shield’s inside surface and the harmful “common-mode currents” and so-called “antenna-current” are on the shield’s outside surface.
It is the choke that limits these harmful currents; these currents do cause flux in the core.
As Guanella showed on his schematic, this choke is in parallel with the transmission-line.
This choke serves to isolate the input from the output and so suppresses any conventional transformer current and allows transmission-line current to flow.
So that’s it; we have a 1:1 balun made with an air-core and, in that core, there are no flux-lines from the transmission-line currents.

4b. The seminal work by Joe Reisert W1JR in 1978 enhanced the 1:1 balun idea further by winding that coil of coaxial transmission-line onto a toroid-core. Here also, by virtue of the coax’s inherent shielding feature, we certainly can see there are no flux-lines introduced into the toroid-core when just balanced transmission-line currents are flowing.
W1JR also invented a unique way to wind the coil and here is a link to a site showing that winding-technique:
http://www.balundesigns.com/servlet/the-63/balun-designs-1-cln-1-Isolation/Detail.
Notice the absence of a thick plastic jacket; easier to wind and can put on more turns.

4c. Without giving up much in performance, one can modify W1JR’s coax design to use a two-wire transmission-line where the two-wires are held together to maintain a desired characteristic impedance.
The site called up by this link shows an example:
http://www.balundesigns.com/servlet/the-98/isolation-choke-balun-common/Detail.
A paragraph later in this work has some verbiage about the characteristics of two-wire lines.

4d. Another ingenious style “current-balun” is W2DU’s beaded-balun, where coax-cable is passed through ferrite-beads to form chokes; although simple in concept, proper core permeability choice is needed to prevent cores from breaking due to excess heating of the cores.

5. A different 1:1 balun is a “Ruthroff-balun”; named after its Bell-Labs inventor in 1959.
Another name for the “Ruthroff-balun” is “voltage-balun”.
As I see it, the voltage-balun generally is inferior to the Guanella-balun.

6a. We can make a 1:4 Guanella-balun by connecting two identical 1:1 Guanella baluns such that the inputs are in parallel and the outputs are series connected per either the ARRL’s 2007 Handbook or Antenna-book; a comment that neither book has any text on the importance of the transmission-line’s characteristic-impedance.

6b. Performance is enhanced at higher frequencies if the balun’s transmission-line characteristic-impedance is made to an optimum level.
If matching a 50-ohm input to a 200-ohm load, let characteristic-impedance = 100-ohms.
If matching a 75-ohm input to a 300-ohm load, let characteristic-impedance = 150-ohms.

6c. The Guanella-balun can be reversed.
So, if matching an unbalanced 50-ohm input to a balanced 12.5-ohm load for a Yagi beam, let the balun’s transmission-line have a characteristic-impedance of 25-ohms.

6d. Sevick’s book “Transmission-Line-Transformers” 2nd edition to 4th edition has multiple pages detailing how to make unique value characteristic-impedance coax (his Table 4.1 ranges from 12.5-ohm to 35-ohm) and two-wire lines (his Figure 5.2 ranges from ~32-ohm to ~175-ohm). Also provided are schematics/pictures/details for the surprisingly simple test-equipment he used in testing his cables and baluns.

6e. In instances where there may be a space or weight or cost problem with a two-core 1:4 Guanella-balun, the two-core 1:4 Guanella-balun can be wound with just one-core if the user is willing to trade frequency performance; the exact trade-offs depend on if the load is floating or grounded at its mid-point.

7a. Overall, the 1:1 balun is the most frequently used balun.

7b. Second place in usage is the previously mentioned 1:4 balun (for example, 50-ohm-to-200-ohm), which can be a Guanella or Ruthroff design; strong arguments can be made for each design approach.

7c. The reader should be aware there is a myriad of other impedance-ratio possibilities made up from combinations of current-balun and/or voltage-baluns; one use is to match vertical-antennas.

8a. In a TLT, energy is transferred from input to output by means of a transmission-line
Although hams are very adept in using transmission-lines like coaxial-cable or two-wire ladder-line to move the transmitter’s output to an antenna, in a balun it is a difficult to think of the conductors as transmission-lines. One tends to think they are just two wires wound around a core.

8b. Concerning a transmission-line, the US-Army manual TM11-675 (Aug 1951) has that “…the greater portion (about 90%) of the transmitted energy is in the electromagnetic waves that the line conductors guide through the space between them. In general, less than 10% of the transmitted energy is actually in the conductors of which a well-designed line is comprised.”
(Comment by KM3K: I have not seen these statements anywhere else.)

8c. In a transmission-line, there are two conductors (spaced as in two-wire-line or as in coax).
There is an electric-field between the two wires.
Each wire has its own magnetic-field.
These fields are related in a way that only the Divine Creator fully understands but Heaviside, drawing on Maxwell’s work, was able to express mathematically and we are not going there at all.
For two-wire-line, the energy is confined to the immediate region around those wires.
For coax, the energy is confined between the center-wire and the shield (assuming a good shield).
The electromagnetic signals we will be using will be Transverse-ElectroMagnetic (TEM) waves; that is, the standard HF and VHF ham-band stuff.

9. The TLT can be made in several different ways:
a) coiling loops of coax (which means we have an air-core),
b) winding either coax or two-wire on a toroid-core or ferrite-rod.
To quote Dr. Sevick W2FMI (SK) about the TLT, “….it is both a choke (a lumped element) and a transmission-line (a distributed element).”
The toroid-core has enhanced performance over the other methods, although the other methods have appealing attributes like simplicity of winding and cost.

10. In his seminal ARRL 1985 article, Roy Lewallen W7EL introduced the terms of “current-balun” and “voltage-balun”.
Here is a link; it is a good read…. http://www.eznec.com/Amateur/Articles/Baluns.pdf
Both baluns work by the transmission-line mode.
The current-balun (Guanella) puts equal current into/out-of its output terminals; highly desired feature.
The voltage-balun (Ruthroff) puts equal magnitude voltages at its output terminals.

11. It is important to comment on the impressive frequency range of a well-designed current-balun.
Earlier in this work, I provided a link to a toroid wound with coax; that commercially produced 1:1 current-balun at 50-ohms, wound on a toroid-core with coax, has a passband from 1 to 54 MHz with very, very low insertion-loss and very high choking-impedance.

12. Moving energy in a balun from input to output by the transmission-line mode on a toroid-core means that theoretically there is no flux in the toroid-core.
Now, in actuality, because we cannot make a perfect transmission-line, we can expect there will be some accidental flux but nowhere near the magnitude from a conventional transformer.
It is easy to understand why there is no flux for well made coaxial-cable.
But it may not be so clear for two-wire transmission-line.
So here is why there is no flux in the toroid-core for a theoretical two-wire transmission-line:
a) the current in each wire is equal but going in opposite directions; should be 180 deg out-of-phase.
b) this means the magnetic-fields caused by those currents are also opposite and cancel each other out everywhere except in the very immediate region between the two wires.
c) because the magnetic-fields largely cancel everywhere, there is no net flux in the core, which in turn means no concerns about core saturation and all the problems that can bring.

13a. If the balun-designer is not careful, it is possible for the balun to conduct conventional transformer currents rather than the desired transmission-line currents.
Winding the transmission-line on the toroid-core makes a choke; hence the name “choke-balun”.
The choke has a reactance and it serves to isolate the input from the output and works to prevent conventional transformer currents; we want that reactance to be at least 5 times (some authors say 10x) the load resistance.
Although more turns makes more reactance, it adds parasitic capacitance between the windings and that harms the high frequency responses.

13b. In a voltage-balun, in spite of good design efforts, invariably at the lower frequencies there will be conventional transformer currents and attendant flux-linkages in the core; this causes the low frequency range to suffer performance (typically higher insertion-loss).

14. If the 1:1 current-balun is used in a 50-ohm system, then the transmission-line’s characteristic-impedance used in the balun is 50-ohm.
One factor in the high-frequency limit of the balun is determined by how close to 50-ohm is that transmission-line’s characteristic-impedance; parasitic generally are absorbed but can be another factor.

15. Key uses for the 1:1 current-balun:
a) is to marginalize the “inverted-L current” in the transmission-line feeding a dipole-antenna.
This will prevent a radiating Feedline and prevent distorting the antenna’s radiation pattern.
There will be some flux in the core but it’ll be minor due to the small current causing it.
This problem is covered at length in the 2007 ARRL Antenna-Book.
b) drive balanced antennas (e.g. dipole or Yagi) with equal currents.

16. A quick comment about notation “1:4” and “4:1”; which is right?.
The reader may see impedance-ratios written by an author as 1:4 or as 4:1.
From my reading, I noted the only consistency is inconsistency in notation (even with the same author).
So my own rule is to read the text and glean from it what is the author’s meaning.

I hope this was of some interest and help to my fellow hams.

73 Jerry Sodus KM3K; FN10je; member of the South-Mountain-Radio-Amateurs Club

Member Comments:
This article has expired. No more comments may be added.
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KH6DC on July 10, 2010 Mail this to a friend!
thanks for the in depth breakdown of voltage and current baluns. Real informative and great learning/refresher.

73, Delwyn, KH6DC
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KC8Y on July 10, 2010 Mail this to a friend!
Super article on baluns. Viewed the "...seminal ARRL 1985 article, Roy Lewallen W7EL..."

great information on baluns.

KEN, KC8Y

 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by WA2JJH on July 10, 2010 Mail this to a friend!
Outstanding post.


Khon-Ken Province Thailand. Be back soon ;)
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by TANAKASAN on July 10, 2010 Mail this to a friend!
If I have a filter with a 300 ohm input and output impedance do I use a current or a voltage balun to match it to 50 ohms?

Tanakasan
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by N4KC on July 10, 2010 Mail this to a friend!
Very good post! And thanks for the links. I was not familiar with Roy Lewallen's article, but it is an excellent piece. His explanation of common mode current on the feedline is worth the time it takes to read the article, just by itself.

I remember way back when as more people began using coax to feed their dipoles, they kept hearing of the necessity of a balun to make the transition from unbalanced feedline to balanced antenna. So everyone rushed out and bought or built voltage baluns for that purpose. Forcing equal voltage on each side of the feedline seemed like a fine idea. That's balance, right? And balance is good.

But the antennas just did not seem to work very well. Nobody gave thought about what forcing equal voltage across varied impedances might do! And could not understand why simply running a balanced antenna to an unbalanced coaxial feedline worked better than that ugly balun thingy. And many still don't understand why balanced feedline works better still.

Ladder line/open wire line unfortunatley still has that old stigma of radiating, causing pattern skew, RFI, and RF in the shack. People think that, with no shield to "hold the RF inside," it cannot work as well as coax and has to radiate. Roy's article does a good job of explaining how imbalance can happen with open wire...as well as with coax...and what can be done about it, including employing a good current balun.

Good stuff, all the way around...

Don Keith N4KC
www.n4kc.com
www.donkeith.com
www.n4kc.blogspot.com

 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on July 10, 2010 Mail this to a friend!
Towards answering TANAKASAN's 300-ohm filter question, really first some more details need to be supplied, such as:
a) tell us more about what will be driving and loading the filter...impedance?...balanced or not?
b) is the filter's input and/or output balanced or not?
c) what is the filter's expected bandwidth in the passband and stopband?
73 Jerry KM3K
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W8JI on July 10, 2010 Mail this to a friend!
Very good article, but something of note is:

"4b. The seminal work by Joe Reisert W1JR in 1978 enhanced the 1:1 balun idea further by winding that coil of coaxial transmission-line onto a toroid-core. Here also, by virtue of the coax’s inherent shielding feature, we certainly can see there are no flux-lines introduced into the toroid-core when just balanced transmission-line currents are flowing.
W1JR also invented a unique way to wind the coil and here is a link to a site showing that winding-technique:
http://www.balundesigns.com/servlet/the-63/balun-designs-1-cln-1-Isolation/Detail.
Notice the absence of a thick plastic jacket; easier to wind and can put on more turns."

There is no practical improvement between using the Reisert winding technique and a normal winding technique. It moves things around a little to use the fancy cross-back winding, but it really doesn't improve anything over the normal operating range and can actually hurt things.

http://www.w8ji.com/toroid_balun_winding.htm

73 Tom
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by AD7WB on July 10, 2010 Mail this to a friend!
Thanks for the write-up.

Can anyone expand on items 6a) and 6b)? It seems to me that the input differential voltages and currents equal the output differential voltages and currents for each of the two 1:1 devices that make up the 4:1 device. If that is so, shouldn't the impedance of the individual 1:1 devices' transmission line be the same as the input impedance?

Lance
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by AD7WB on July 10, 2010 Mail this to a friend!
Ok, never mind, I thought a little more and since the two 1:1 device inputs are in parallel the inputs should be 100 ohms to match the 50 ohm source. And the outputs should be 100 ohms because they are 1:1. And the transmission line windings should be as well.

Thanks again for the article,

Lance
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by K0IC on July 10, 2010 Mail this to a friend!
Thanks for defining art and science in a balun.

I have a 4:1 current balun to feed my 180-foot lazy NVIS antenna wound around my garage. I ground one end to earth ground. I am not sure if I am grounding the right end or if it makes any difference?

Someday I hope to compare a lazy NVIS antenna with a Hustler six band vertical with four 100-foot ground wires. I say one needs both antennas depending on the skip angles. On twenty meters I can work coast-to-coast with my 180-foot lazy NVIS antenna. I assume I will need a gas lightning arrestor with the Hustler six band vertical. The house needs to be re-shingled so it might be a few months before this happens.
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by N5TGL on July 10, 2010 Mail this to a friend!
Well done! Thanks for doing this writeup, lots to digest here.
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KG4TKC on July 10, 2010 Mail this to a friend!
Thank you KM3K for an excellent article,you deserve a 'job well done' for this piece.
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by K9FV on July 10, 2010 Mail this to a friend!

Thank you Tom for the link: /www.w8ji.com/toroid_balun_winding.htm and work you've done. You just answered a question I was getting ready to ask.

Also, THANK YOU to the author of the article - well written with good info.

73 de Ken H>
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on July 10, 2010 Mail this to a friend!
To Tom W8JI,

Thank you for the compliment about the balun article.

I'd like to comment about these two sentences---"There is no practical improvement between using the Reisert winding technique and a normal winding technique. It moves things around a little to use the fancy cross-back winding, but it really doesn't improve anything over the normal operating range and can actually hurt things."

Dr. Sevick made baluns with both winding-techniques and, after testing them, he wrote "...no difference in performance at HF was noticed..."; he also wrote, "The main advantage in the HF band with the crossover winding is purely mechanical. Having the input and the output connections on opposite sides of the toroid is not only more convenient, but it also offers a much stronger method of mounting."

I've since gone to your web-site and am still pondering over it.
To that end, I need to ask though about the first photo.
I will write that, when I first saw that picture, I winced so hard that I almost got whip-lash.
Consider putting up a warning-sign. :)
I winced because, in my experience, bunching-up turns and leaving two gaps as done in that photo generally is disastrous; something to be absolutely avoided.
But, when I saw the plastic-tie to secure the bunched-up turns, then I thought the man is intentionally using such poor coil-winding technique to make a point.
So, I'd like to ask, "Towards what purpose were the windings bunched-up such as they are?"

73 Jerry KM3K
 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KX0R on July 10, 2010 Mail this to a friend!
What a great article! The Guanella balun is indeed a nice tool to become familiar with. It opens up many possibilities. At low power these things are almost trivial to use!

Not very long ago I discovered how to make simple 1:9 Guanella transformers. The ones I make are not baluns - they are intentionally balanced-to-balanced devices. We could call them balbal's or just balanced transformers.

I wanted to use open wire line to feed a balanced antenna measuring about 60 ohms. For several years I've been modelling and building Bird Yagi antennas for Field Day - these simple antennas have natural balanced feeds of 60 to 70 ohms, with very wide frequency coverage, when built with 2 elements (driven and reflector). I like them because they are easily built with wires and can be suspended by ropes in trees.

I wanted to minimize loss since I run 5W at my 1B-battery FD station, and my feeds are over 100 feet long. Low-loss coax is not only expensive, but it's cumbersome and too heavy to hang on the feeds of my 2-band Bird Yagi in the pine tree tops. So I made up some open-wire line from #18 stranded wire and Home-Depot polyethylene irrigation tubing - best feedline I've ever used. I spaced my line about 3", so it's 500 to 600 ohms.

With a 1:9 transformer at the Bird Yagi's 60-ohm feed, it transforms to 540 ohms. This is balanced-to-balanced. The open-wire feedline is practically a flat line, virtually lossless, and so very easy to match at the transmitter end. There I use a BLT (balanced line tuner), so no balun is required. The BLT's tune so easily, with broad, un-critical settings, that I know the balanced lines are behaving. The BLT's also are almost lossless going from 50 ohms to 500 or 600 ohms resistive. There's only about 3 feet of coax in the system - from the rig to the BLT.

The 1:9 transformers are very easy to make for low power - just make three 1:1 Guanella units, connect the input sides in parallel, and connect the outputs in series. Ideally the impedance of the individual windings should be about 180 ohms if you're going from 60 ohms to 540 ohms. In practice you can get away with a lower impedance winding, but the bandwidth of the transformer is reduced. Since my units work on 7 MHz and 14 Mhz for the dual-band Bird Yagi, the impedance error in the transformers is no big deal. I tested a couple of my 1:9 units back-to-back and the loss was measurable but trivial, about 0.2 db per unit. I used a couple of #22 teflon wires side-by-side for windings, 10 simple turns on #43 ferrite cores, about 3/4" OD, and these things work great with no attempt at optimization! They're pretty flat from about 4 MHz to over 30 Mhz - no doubt they can be improved.

Building a high-power 1:9 Guanella transformer is not trivial - the voltages are high, the winding impedances should be close to 150 ohms, and the 3 cores tend to be large - but someone's going to figure out how to do a really good one. Then there will be no excuse for using expensive coax from your station to your antenna(s) - instead we'll put a 1:9 at the station and a 9:1 at the antenna, and we'll use balanced open-wire-line in between - it's cheap and more power is delivered to the load. Balanced antennas fed with balanced line are really nice!

I encourage readers to play around with Guanella transformers - they are easy to make as long as you don't get too idealistic - you can learn a lot just by winding some cores, connecting them up, and doing some easy measurements. You'll be impressed.

Carey
KX0R

 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by WA8FOZ on July 11, 2010 Mail this to a friend!
Terrific article!
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by N2EY on July 12, 2010 Mail this to a friend!
Jerry,

WELL DONE! A lot of good info in a small space. I'm sure there are small things that could be tweaked but the info is all there - or is contained in links and references.

Couple of comments:

1) I'm still on the fence about the two different winding methods and their advantages/disadvantages. I see W8JI's point about the different measurements but also see your point about how winding technique can make a difference.

2) Would like to see a side-by-side comparison between comparable amateur-type "coax wound on a toroid" 1:1 baluns and "lots of ferrite beads on a straight piece of coax" baluns. By comparable I mean size/weight/cost.

3) Suggestion for a companion article (not necessarily by you): "When do you really need a balun?"

73 es TNX de Jim, N2EY

 
How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KL7AJ on July 12, 2010 Mail this to a friend!
One of the limitations of the transmission line transformer is the fact that it generally only operates in "perfect square" impedance ratios, i.e, 1:1, 1:4, 1:9,
1:16, 1:25, etc. Jerry Sevick, W3FMI got around this limitation with some clever "Texas Two-step" modification of the basic transformer so as to get "odd" impedance ratios. These "half-tapped" transformers aren't perfect, but they do work very well for most ham applications (and some commercial and scienrific applications as well).

Eric
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by LU2DFM on July 13, 2010 Mail this to a friend!
Looking at the pictures in Tom's article, the only question I have is that the second balun looks like it has 1 or 2 turns more than the cross-coiled (first picture) one, and this could affect the comparison.

Aside from that, I think the explanation is very clear and I think correct:

"The rule is simple. With any parallel L and C (neglecting losses), impedance always increases unless the parallel capacitive reactance (in ohms) is less than half the value of the inductance (in ohms). With a choke winding impedance of j300 with an added parallel -j150 capacitive reactance, the combined impedance would be -j300. Impedance is the same value, only the sign changes! Any value of capacitive shunt reactance higher than -j150 (less capacitance value) would actually increase choke isolation!"

So, the neater winding in the commercial product has the effect of reducing parallel capacitance (by spreading the turns), leading to a reduced common mode total impedance and pushing up the frequency where performance peaks.

Also, by Tom's results, the bunch winding method seems to provide more bandwidth.

73 de Fer
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W8JI on July 15, 2010 Mail this to a friend!
It's very plain how the crossed-back winding works, and why it is not the improvement suggested. Many people, not just me, have proved this with measurements. It moves peak isolation up in frequency, but also reduces the peak isolation. At HF in a balun case it does virtually nothing helpful at all.

73 Tom

 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W8JI on July 15, 2010 Mail this to a friend!
Jerry,

You said:

"I winced because, in my experience, bunching-up turns and leaving two gaps as done in that photo generally is disastrous; something to be absolutely avoided.
But, when I saw the plastic-tie to secure the bunched-up turns, then I thought the man is intentionally using such poor coil-winding technique to make a point.
So, I'd like to ask, "Towards what purpose were the windings bunched-up such as they are?""

I don't know what you are measuring (because your response was mostly non-technical innuendo and "drama") but here is how it actually works:

1.) When windings are bunched on a toroid core the effect is the impedance of the winding as a general rule INCREASES. The only time impedance would decrease would be when the winding is above the self-resonant frequency.

2.) The proposed advantage by the Reisert technique comes from minimizing shunt capacitance from end-to-end on the windings. As such, and because bunching the windings IMPROVED impedance, I bunched them.

3.) **You** said Sevik measured the Reisert winding style and found no electrical improvement. I know of several others that have also measured the same, so why would anyone assume my data is out of line when even you supplied supporting data for my conclusions? :-)

4.) I used the optimum impedance spacings for both winding styles, because that is what I felt answered the actual question.

You may personally not like bunching the wires, but a few minutes with an impedance meter measuring below the SRF of any toroid inductor will show the impedance **increases** when the winding is "bunched". This is because it minimizes flux leakage.

In a typical T94-2 core the impedance ratio is about 30% or more **higher** when turns are bunched as a opposed to being evenly spread over the core area.

The only deleterious effects I am aware of are that heating is more concentrated in a smaller core area, although thermal conductivity disperses the heat so that isn't much of a factor. The second effect is flux density is higher in the core area under the winding with more concentrated turns (hand in hand with heat) but that isn't an issue either unless the core is near saturation limits.

All of this can be easily proved with a few minutes of work, and quantified in real numbers for any particular core.

Despite dramatic personal innuendo to the contrary, my data was fair. I actually optimized the Reisert winding style I used. My data also agrees with many other people's independent results, that there is no real electrical advantage.

Now the real kicker is some popular baluns are not actually baluns at all. They actually guarantee a balanced system will be UNbalanced by the addition of the very balun that is supposed to improve the balance. We can't always trust what we copy from books, because sometimes books have mistakes. It's really a good thing when we point out these mistakes by doing real meaningful measurements.

:-)

73 Tom













 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on July 15, 2010 Mail this to a friend!
Hi Tom W8JI,

My posting was directed solely to the winding-technique.
Apologies for the "drama"; no intent that way at all.
I just reported my spontaneous reaction to the photo.

Please notice that I was not questioning the test-data.
But, as long as the topic has come up now, I think it would be interesting to compare that data to the graph of common-mode impedance associated with the article's link to the BalunDesigns "Reisert" balun, which stacked two cores.
BTW, I have no connection in any way to BalunDesigns; I merely make use of their public data accessible from the Internet.

To repeat from an earlier posting, Sevick wrote any advantage from Reisert's method is strictly mechanical.

Here is my background with coil-winding technique (none of which so far involves winding Guanella-baluns):
In my 27+ years as a crystal-filter design-engineer, I've probably designed maybe 4,000 coils (mix of toroids, pot-core, solenoid).
Those designs, when put into production, multiplied into production of perhaps many hundreds of thousands of coils.
With production numbers like that, the designs have to be right.

As with any design-work, associated bench-work is done; it was there that I learned to avoid separating windings per that photo.
As best as I can recall, two problems were "stability over temperature" and "coupling to/from other circuitry".
I wish I could recall other difficulties from my poor memory but too many years have gone by and, of course, my engineering-notes are the property of that employer, so I no longer have access to them.
Bottom-line gut-reaction is I remember it as not a good technique and that caused the red-flag to go up.

No issues with anything else in your posting.

:) 73 Jerry KM3K
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W8JI on July 15, 2010 Mail this to a friend!
Hi Jerry,

You wrote:

<<As with any design-work, associated bench-work is done; it was there that I learned to avoid separating windings per that photo.
As best as I can recall, two problems were "stability over temperature" and "coupling to/from other circuitry". >>

I've run into stability issues also, using toroids in high-Q narrow band systems. Actually this would apply to any reactance-critical system, like your application with filters (assuming the transformers are not just broadband coupling systems). We are exactly on the same page there, although my experience is phasing systems and reactance compensation is high-Q systems.

As for coupling problems, you and I are also on the same page. I read about toroids not having flux leakage, and being "self-shielded", and wonder where that comes from. I can squeeze or spread the turns and change inductance substantially on many cores, and that all indicates significant flux leakage.

I think the subtle point you initially overlooked is the leakage reactance is not important at all in this application. After all, the series connected "input and output" windings are wound on the very same core!!!

In your filter application you would never have an input transformer, followed by a crystal or crystals,coupled to the output winding! In that case spacing turns evenly makes sense.

In this application the input and output shares a single core, and any minor flux leakage variation has nothing to impact. There is no second transformer, especially one on another side of a very high attenuation filter component!

In a balun or common mode isolation device application, we only care about total series impedance and and shunt reactance to the case or surroundings. Magnetic flux leakage is not a concern unless the core is open-ended. The electric field is the dominant problem.

As such the system is improved in a few ways by using a compacted winding:

1.) Stray capacitance is reduced by moving open ends further apart. This moves SRF up, which of course could be deleterious if we are already operating below SRF, but it also allows us to have more inductance for the same SRF...which is good for choking.

2.) Bunching turns increases winding inductance because it increases turn-to-turn magnetic flux coupling. This moves SRF down, the opposite of effect 1. This is actually a significantly larger effect with most cores when compared to 1.

In high power systems, when a core is near capacity for heat (the usual limit) or near flux density limits (rare for high duty cycles like we use), we probably do not want to compact the winding. In an application where we worry about coupling from one coil to another isolated coil, we would want to spread the turns evenly to keep flux pulled into the endless circular magnetic path.

Not in this case though. Don't be alarmed at the winding style changes between systems coupled at a distance after passing components making very deep notches, and this application where the two "coils" share a common core. :-)

73 Tom

 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by N2EY on July 16, 2010 Mail this to a friend!
ISTM that the basic difference between the Ruthroff "voltage balun" concept and the Guanella "current balun" concept, *as applied to amateur HF baluns* is this:

The Ruthroff voltage balun designs are such that a large part (or all) of the energy being transferred from the unbalanced side to the balanced side, or vice-versa, has to go through the magnetic balun core. Whether the core is air, ferrite, or powdered iron, a solenoid or a toroid doesn't matter, lots of the RF has to go through it in a Ruthroff voltage balun.

The Guanella current balun designs are such that they don't require a large part of the energy being transferred from the unbalanced side to the balanced side, or vice-versa, to go through the magnetic balun core. The core only exists to reduce common-mode RF.

For example, consider the classic 4:1 Rothroff balun consisting of a bifilar wire winding on a powdered-iron toroid core, with the windings connected series-aiding and the center-tap so produced being grounded. It's the kind of balun commonly supplied and used with amateur-radio HF Transmatches of many kinds for at least 40 years.

In such a balun, the unbalanced side is connected to ground and one ungrounded winding end, while the balanced side is connected across both ungrounded winding ends. The 4:1 ratio results from the 2:1 turns ratio squared. The only way for the desired RF energy to get from one ungrounded winding end to the other is through the balun winding and core. Otherwise there's no path.

The Rothroff balun under consideration deals with common-mode *voltages* on the balanced side by shorting them to ground. This happens because the bifilar winding impedances essentially/hopefully cancel each other out in the common mode.

Now consider a Guanella balun that consists of a coax cable winding on a powdered-iron or ferrite toroid core. It's the kind of balun shown in the links.

In such a balun, the desired RF energy gets from one end to the other *inside* the coax. The balun core has no effect on that RF because it's outside the coax.

The Guanella balun under consideration deals with common-mode *currents* on the balanced side by presenting them with a very high impedance to ground. This happens because the common-mode currents have to deal with the impedance produced by winding the coax on the toroid core.

You can see how this would play out in real life. A Rothroff balun could have low common-mode voltages but high common-mode currents, while a Guanella balun could have the opposite.

Which balun to use depends on the application.

----

Some fun facts:

- Coax was in use by hams even before WW2. There was even a QST article (Feb 1938, by W2BZR) about making your own air-insulated rigid coax from copper pipe and common hand tools.

- In the 1950s B&W sold air-core balun coils for HF amateur use. (QST, June 1957, article by W6EBY shows a typical application.) Heath made a kit (B-1) based on them. They consisted of a pair of solenoid coils with bifilar windings, which could be connected for 1:1 or 4:1 ratio. By the 1960s hams were using "iron core" (mostly powdered iron) baluns for such applications.

It would be really interesting to see the characteristics of those old B&W air-core balun coils measured with modern techniques...


73 de Jim, N2EY
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by K7JBQ on July 16, 2010 Mail this to a friend!
Jim,

Thank you. I'd forgotten why some of us who have been licensed for half a century or more habitually refer to "balun coils," and why younger hams look at us like we're idiots.

73,
Bill
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W4AEE on July 21, 2010 Mail this to a friend!
An easy, small and inexpensive, 1:1, HF, common-mode current balun for 100 - 200W (perhaps more, depending on SWR) can be made from a short length of good quality RG58/U (I use flexible "Davis RF" brand) and a giant, ferrite bead of #31-mix material. The bead I use is a Fair-Rite Products Corp. p/n 2631102002, available from Mouser Electronics, their p/n 623-263-1102002 @ $1.90 ea. The bead has dimensions of 1.02" OD, 0.505" ID, and 1.125" long. Pass the coax through the balun as many times as possible, maybe using a little lubricant to get that last turn through. Five turns are usually possible. Position the choke about a foot or so below the antenna feedpoint. This choke balun works fine here for 80 - 20m. at SWR's of 5:1 (Higher freq and greater SWR more have not been tried). If more impedance is needed (e.g. for 160m.), put two cores side by side and pass each coax turn through both. Use good black plastic cable ties to keep it all together. A QRP balun (est. 50W or less) with higher impedance is possible by substituting RG174/U miniature coax. 73, Mike, W4AEE
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on August 1, 2010 Mail this to a friend!
Hello Jim N2EY,
I’d been meaning to get these replies to your posting much sooner but “stuff” kept coming up; oh well!!!
So anyway, here we go.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: ISTM that the basic difference between the Ruthroff "voltage balun" concept and the Guanella "current balun" concept, *as applied to amateur HF baluns* is this:

The Ruthroff voltage balun designs are such that a large part (or all) of the energy being transferred from the unbalanced side to the balanced side, or vice-versa, has to go through the magnetic balun core.
Whether the core is air, ferrite, or powdered iron, a solenoid or a toroid doesn't matter, lots of the RF has to go through it in a Ruthroff voltage balun.

KM3K:This is only true when the coil’s inductive-reactance is too low to prevent the undesired and potentially harmful conventional-transformer-current from flowing.
We can expect both style baluns at low-frequency to have conventional-transformer-current in (more so with the Ruthroff-balun) and there is little that can be done to prevent it at low-frequencies.
The coil’s inductive-reactance can be increased by:
a) adding more turns, or
b) by using a higher permeability, or
c) by increasing the core’s cross-section area (for example, stacking cores).
There are design compromises to choices ‘a’ and ‘b’.
It is intended for both the Guanella (current) and Ruthroff (voltage) baluns to transfer energy by transmission-line mode.
------------------------------------------------------------------------------------------------------------------------------------
N2EY: The Guanella current balun designs are such that they don't require a large part of the energy being transferred from the unbalanced side to the balanced side, or vice-versa, to go through the magnetic balun core.
The core only exists to reduce common-mode RF.

KM3K: IMHO the word “only” should be removed; Sevick has a plot showing a core improves the overall frequency-response but he makes no attempt to explain the mechanism to account for this improvement. It'd be nice to know what the cause is.
------------------------------------------------------------------------------------------------------------------------------------
N2EY: For example, consider the classic 4:1 Rothroff balun that consists of a bifilar wire winding on a powdered-iron toroid core, with the windings connected series-aiding and the center-tap so produced being grounded.
It's the kind of balun commonly supplied and used with amateur-radio HF Transmatches of many kinds for at least 40 years.

KM3K: What’s a Transmatch?? The ARRL-Handbook and Antenna-Book only say it’s an antenna-tuner.
However, about the powdered-iron toroid in a tuner, Sevick W2FMI (SK), in his “Transmission-Line-Transformers” 2nd, 3rd, & 4th editions pages 9-20 & 9-21, writes about the danger of using a powdered-iron core in a tuner.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: In such a balun, the unbalanced side is connected to ground and one ungrounded winding end, while the balanced side is connected across both ungrounded winding ends.
The 4:1 ratio results from the 2:1 turns ratio squared.
The only way for the desired RF energy to get from one ungrounded winding end to the other is through the balun winding and core.
Otherwise there's no path.

KM3K: The way I read it, you’ve described the Ruthroff-balun’s low-frequency model, which is in play only when the undesired conventional-transformer-current flow is present due to the winding’s having insufficient inductive-reactance.
Hopefully, at some higher frequency when the windings have sufficient inductive-reactance, the device works entirely in the more efficient (lower loss) transmission-line mode (no flux-linkages) and so the 1:4 ratio change comes from voltage summation.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: The Rothroff balun under consideration deals with common-mode *voltages* on the balanced side by shorting them to ground.

KM3K: Actually, the common-mode-current is reduced by the winding’s inductive-reactance (think Ohm’s law).
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: This happens because the bifilar winding impedances essentially cancel each other out in the common mode.

KM3K: please see the previous comment.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: Now consider the Guanella balun that consists of a coax cable winding on a powdered-iron or ferrite toroid core. It's the kind of balun you showed in the links.

In such a balun, the desired RF energy gets from one end to the other *inside*the coax.
The balun core has no effect because it's outside the coax.

KM3K: As noted above, the core, for reason(s) not explained by W2FMI (SK), is needed to enhance the balun’s efficiency over a greater frequency range.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: The Guanella balun under consideration deals with common-mode *currents* on the balanced side by presenting them with a very high impedance to ground.

KM3K: strike-out “on the balanced side” and IMHO you’ve got it.
------------------------------------------------------------------------------------------------------------------------------------
N2EY: This happens because the common-mode currents have to deal with the impedance produced by winding the coax on the toroid core.

KM3K: yes, indeed they do.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: You can see how this would play out in real life. A Ruthroff balun could have low common-mode voltages but high common-mode currents, while a Guanella balun could have the opposite.

KM3K: I don’t understand the sentences.
-----------------------------------------------------------------------------------------------------------------------------------
N2EY: Which balun to use depends on the application.

KM3K: For a one-time application, unless compelled by size, cost, or weight, my choice is a Guanella(current)-balun.
Sure, there may be a Ruthroff(voltage)-balun design that’ll work but I think I’d usually ask myself if it is worth the hassle of figuring out if it’ll work or not (unless for just the fun of it to see if I could make something special).
If I can afford it, I’d go with the safety and expected superior performance of the Guanella(current)-balun and move on.
-----------------------------------------------------------------------------------------------------------------------------------
So there are my opinions; lots of stuff there.
73 Jerry KM3K
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W6EM on August 17, 2010 Mail this to a friend!
While I agree with what you've said with respect to the current transformer balun(s) under discussion here, I don't with respect to your description of what amounts to an "autotransformer" voltage balun.

Jim, N2EY, is correct in his explanation. Common-mode, of especially longer wavelength low frequency noise voltages can appear in phase across both windings connected to a dipole. As such, they cancel in both windings that are grounded at their connection, where they are connected in series from transformer action. That is, the grounded point where also the incoming coax shield is also connected.

The coax is connected to the top of the third coil. Thus, there are two coils seen by the incoming coax, and two coils seen by the antenna.

Your claim of no transformer action in the case of a voltage balun makes no sense. In order for the bottom coil to have induced voltage it must be induced through transformer action of the coils and the core. Why? Because its other end is grounded to the coax shield and possibly another ground at the antenna.

Self impedance is important in that case, yes. In order to minimize losses. Self-impedance should be at least 10 times the value of the port impedance to minimize core losses. Therefore, the two coils on the coax side should have a combined self-impedance of at least 500-750 ohms. I usually shoot for twice that amount of reactance in choice of core.

The three winding voltage balun is an autotransformer, plain and simple. With its advantages and disadvantages. A current transformer it is not.

What I distinctly do not like about CT baluns is the lack of effective DC or low frequency grounding of both sides of the antenna (at the antenna) that comes built-into the voltage balun. Static charges can build up or be sent back through a current balun to terminate in whatever protection your receiver input has from the ungrounded leg of your antenna. Not something I relish, especially living in a high lightning incidence area. While a choke balun does afford additional common mode reactance that will help suppress lightning current flow assumed common to both sides, it does nothing to prevent static charge build up on antennas that a voltage balun does automatically.

Another thing not discussed here is the perturbation of the fields of two conductors of a balanced feedline (two wires) when wrapped about a core. Could this be why many insist on using bifilar windings about a core? Twisting or transposition as it used to be called, is the fundamental reason why pairs are twisted about each other on lengthy transmission lines. To improve balance with respect to ground and minimize coupling. Ask any telecomm engineer with prior-to-fiber experience.

Perhaps, for small lengths about cores it isn't all that important or germaine to this discussion, but if one is truly concerned about maintaining current balance without core involvement, then I suspect it would be. Balanced line wrapped about itself is hardly balanced.

Lee
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on August 17, 2010 Mail this to a friend!
To Lee W6EM,
I'd like to reply but I'm having trouble linking your comments to the specific material with which you wish to take exception.
From my seat, it is obvious there is a lot of material that preceded your posting; a photographic memory I don't have, so it is too tedious to sort thru it all to match up the issues and hope I got it right.
As an example, you've written to me, "Your claim of no transformer action in the case of a voltage balun makes no sense."; it'd help me if perhaps you could quote where I gave that impression and I'd be happy to engage in dialogue.
Same principle relative to my postings applies to the remainder of your post.
Dialogue is good because I did write that I don't know all the answers.
Looking forward to your reply.
73 Jerry KM3K
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by W6EM on August 17, 2010 Mail this to a friend!
Perhaps this will help.

Jim, N2EK: "The Ruthroff voltage balun designs are such that a large part (or all) of the energy being transferred from the unbalanced side to the balanced side, or vice-versa, has to go through the magnetic balun core.
Whether the core is air, ferrite, or powdered iron, a solenoid or a toroid doesn't matter, lots of the RF has to go through it in a Ruthroff voltage balun."

This is exactly how a voltage balun works, and what prompted my response.

Now, your response, I take exception to. The above is a correct description assuming there is plenty of self-impedance in the winding. 10X or more of the source impedance.

Only true when the inductive reactance is too low? Nope, not the case. If its too low, losses go way up.
Transformer performance is desired in a voltage balun.

"KM3K:This is only true when the coil’s inductive-reactance is too low to prevent the undesired and potentially harmful conventional-transformer-current from flowing.
We can expect both style baluns at low-frequency to have conventional-transformer-current in (more so with the Ruthroff-balun) and there is little that can be done to prevent it at low-frequencies.

Ideally, in a CT, the flux is minimum in the core since the winding fluxes, if 180 out of phase and equal, cancel out one another in the core or air, as the case may be.

"The coil’s inductive-reactance can be increased by:
a) adding more turns, or
b) by using a higher permeability, or
c) by increasing the core’s cross-section area (for example, stacking cores).
There are design compromises to choices ‘a’ and ‘b’.
It is intended for both the Guanella (current) and Ruthroff (voltage) baluns to transfer energy by transmission-line mode."

I take issue with your claim of transferring energy in a voltage balun by transmission-line mode. Doesn't happen. It happens by transformer action.

You may be able to transfer energy through the transmission line wrapped as a choke inductor even if sufficient flux and coupling isn't there to achieve maximal field attenuation about the windings from transformer action between the two paths. So, there's really no obvious failure to a CT balun to force equal and opposite currents in a transmission line. Unless you perform current sniffing at the antenna, you won't see it.

Does that help?

As to my explanation of the benefits of a voltage balun on an otherwise ungrounded half dipole, that wasn't mentioned before.

73.

Lee
W6EM
 
RE: How a 1:1 Guanella-Balun (Current-Balun) Operates  
by KM3K on August 22, 2010 Mail this to a friend!
Hi Lee W6EM,
I'd need some answers to a couple questions before I can write an appropriate response:
1. The abbreviation "CT" is not familiar to me; for my benefit, please elaborate what it means. Thanks.
2a. Are the voltage-baluns, about which you are making comments, the 1:1 impedance-ratio or the 1:4 impedance-ratio or both or are you possibly writing about auto-transformers?
BTW, the auto-transformer works by flux-linkages and not by the transmission-line mode.
2b. The 1:4 has a schematic different from the 1:1, which has several different schematics and attendant winding-methods; so, if it is 1:1, which schematic/winding-method?
73 Jerry KM3K
 
Email Subscription
You are not subscribed to discussions on this article.

Subscribe!
My Subscriptions
Subscriptions Help

Other Recent Articles
Propagation Forecast Bulletin #30 de K7RA:
Amateur Radio Enthusiasts Hosting Event In OKC:
ARNewsline Report 1928 -- July 25 2014:
DX News -- ARRL DX Bulletin #30:
Just Ahead in Radiosport: