[TenTec] 238 "Tuner" Balun

Rick at dj0ip.de Rick at dj0ip.de
Thu Dec 28 16:03:40 EST 2017

Hi Jim,

Happy Holidays!
Hope you are not affected by the fire.

You are an expert on theory but you have often said that you do not like openwire and openwire fed dipoles.
I am assuming you have not used them very much for that reason.

I doubt all of this theory will help Bob in his immediate quest to find a balun for his matchbox, though it is certainly good advice to keep learning.

Having used openwire-fed dipoles for about 50 years, here is some of my practical experience:
Erect a dipole of any convenient length.
Feed it with openwire (I prefer homebrew ladderline over WIREMAN windowline).

Using an ASYmmetrical matchbox, such as the Model 238-x, or any of the multitude of T-network matchboxes, insert a balun between the feedline and its coaxial output.

Key-down 1kW of power (after first matching, of course), and then simply touch the PL-259 on the matchbox side of the balun... or the case of the matchbox, whatever.  

Do this for each and every type of balun you can build, borrow, or buy.
You soon learn what works and what does not.  You 'FEEL' the difference.
OR.. after 15 years of running this with OMNI V/VI/VI+, you soon see which ones trip the circuit breaker and which ones don't.

I blew a small fortune on high-power commercial baluns which not only did now work, but often burned up.
Eventually I quit buying, read RF-Ham, studied Steve's Rainbow Chart on RF Chokes and began building my own.
Since then I have never burned up another balun.

What works (as tuner balun): 
A 1:1 Guanella balun.

What kinda works:
A very good 1:1 Maxwell balun.   Depends on how well the balance is on the feedline/antenna.  If poor, it won't work.
A very good dual-core 4:1 balun, if and only if you have high impedance at the matchbox end of the feedline and a good balance in the antenna system.  

What never works:
A voltage balun, whether 4:1 or 1:1 - except of course at low power.  
A 4:1 Guanella balun wound on a single core.  This one has no CMI and should never be used in RF antennas.

(Definition of "works" is not just the ability to make QSOs but also the ability to do so without getting RF burns.)

Bob asked what kind of balun to use with his matchbox.
I have given him clear instructions on how to build one that I 'KNOW'  works because I've use it often.
It will only cost about $15 to $30, depending on how much power he plans to run.
That's not theory.  It's experience.

Steve suggested I should use #52 instead of #43.
It is probably a good suggestion.  I will try this... in the field, not in the lab.

I spend a lot more time on the low bands than on the high bands and most of the time I have used a beam on the high bands anyway, so I have always placed more focus on 80/40.  Thus my choice of #43.

One thing we should always do when using openwire is twist the feedline about 1 turn per meter/yard, being careful not to short the two lines together if the feedline is not of insulated wire.  This helps reduce common mode current (CMC) pickup on the feedline.   It won't help if the imbalance is in the antenna itself, but now days the feedline almost always picks up CMC from consumer products in our neighbourhood so we should always twist the feedline.

In 2013 I spent a few hundred hours running CMC tests on dipoles (in the field, not in the lab).
I only took measurements on 40m and above.  About 1000 measurements.  All well documented.
I did this on standard dipoles, as well as OCF-dipoles.
I tested several kinds of baluns.

At the end of the test I concluded there was more CMC on the feedline on 40m than any of the other bands, but with good baluns or combination of bad balun and good choke, I was able to reduce the CMC to noise level on my RF Ammeter.  I also found more CMC on the feedline on odd-harmonic bands than on even-harmonic bands.

In 2015 I repeated the test using 80m antennas.  The antenna was at the same physical height (but only half the electrical height).
All of the baluns that had worked quite well in the 40m dipole test failed miserably on 80m.
There was no way to master this with any kind of transformer balun, except with the addition of a good Guanella choke.

At the end of the test, I concluded that the electrical height above ground played a major role in the amount of CMC on the feedline.
I also concluded that with antennas close to the ground, as most of us city dwellers have, more CMI is required and thus chose #43 for my preferred balun and choke mix.  If I were only building 20/15/10, of course I would choose another mix.

For baluns at the feedpoint of a low band dipole, I use #43; if I need 200 Ohms, then I place a 4:1 Ruthroff on #61 core between the 1:1 Guanella and the antenna.  Steve has me curious to try #52.  I did try it in a 4:1 dual-core Guanella and it failed miserably in an 80m OCFD.  I have yet to build a 1:1 choke or balun with it.

I know you are no fan of the OCFD antenna, Jim.  Neither is Dean Straw and I respect both of you guys quite a lot.
But I do know how to measure antennas, how to use them in contests and analyse my results.
I can show you my field measurements on real antennas and show you just EXACTLY what affect common mode current has on the 'shape' of the SWR curve as well as its ultimate level of SWR.  It can cause significant skewing of the curve and even double-dip in SWR.

One of my OCFD antennas has been under review by a local German magazine for about 3 months now.  It uses a hybrid balun that I designed.  
The antenna was used at full legal power in CQWW SSB and CW contests.  The test review will be printed in the February or March issue of the magazine.  
I am very pleased with the tester's ongoing feedback.
BTW, he also has a PHD in EE and is chief technical editor of the mag.

So yes, there is theory, but the rubber meets the road ON-THE-AIR, not in the lab and not in the text book.

Bob asked a simple question.
I answered it.
Steve answered it too.

Jim, tnx for the data-point on cross-winding producing less capacitance than continuous winding.
I failed to pick up on that point, but have always used cross-winding anyway, just for its mechanical benefit.

Happy New Year!

Rick, DJ0IP
(Nr. Frankfurt, Germany)

-----Original Message-----
From: TenTec [mailto:tentec-bounces at contesting.com] On Behalf Of Jim Brown
Sent: 28 December 2017 19:11
To: tentec at contesting.com
Subject: Re: [TenTec] 238 "Tuner" Balun

On 12/28/2017 2:46 AM, Rick at dj0ip.de wrote:
> The problem that we are fighting is common mode current

One fact that is being missed in this discussion is that the transmission line as a common mode element of the antenna system, and, like any antenna, current varies along the length of every element based on the boundary conditions. The most common "boundary conditions" are the open circuits at the ends of conductors that force the current to near zero (limited by capacitance at the ends to free space and the antenna's surroundings). When we add a common mode choke with a high choking Z, we force current to near zero, creating another boundary condition AT THE POINT WHERE THE CHOKE IS PLACED. A choke at the antenna feedpoint forces near zero current there, but the feedline may still connected (as  common mode element) via the tuner to ground. A choke at the tuner can another near-zero condition.

With a choke at the tuner and no choke at the feedpoint, the feedline is still part of the antenna, and is vulnerable to picking up noise on receive. This can be clearly seen (and studied) by a simple NEC model of the antenna that includes the feedline and the choke(s) using the Load function in NEC. My tutorial on chokes shows how to determine the parallel RLC equivalent circuit of a choke based on measurements of its impedance vs frequency. NEC computes currents vs length for every element of the antenna, and can display them graphically and as  table.

> In all of the tests that I have run on dipoles and OCFD antennas, the lower bands, being electrically closer to the ground, have significantly more problem with CMC than the higher bands.
> Therefore I chose to run with #43 when 80m is required.

#31 is several dB better on 80M and a lot better on 160M.

> I agree the point of using bifilar.

If the device in question is operating as a transformer (that is, two windings coupled magnetically but with no DC connection), from a common mode point of view, that's a BAD idea, because it increases the capacitance between turns, and reducing the common mode impedance. Without that capacitance, they have the same effect as a good common mode choke. Transformers are fundamentally different from common mode chokes in that the core carries the entire signal. For receiving only, loss in the core may not matter, but for transmitting, loss can burn a lot of transmit power. At low power levels, this simply makes the signal weaker. At high power levels, losses in the core are likely to cause it to self-destruct.

It should be obvious that cores having low loss at the frequency of interest should be used for transformers. Likewise, the core should provide a magnetic path (dimensions) capable of handling the flux density associated with the power level. Fair-Rite #61 is quite efficient below 10 MHz, with losses beginning to increase above that frequency.  Fair-Rite #67 handles more power to higher frequencies. 
N6RK, a pretty smart EE retired from HP's instrumentation group, uses
#67 for high power transformers in his antenna system.

The only advantage of a bifilar-wound transformer is slightly greater mutual impedance between windings, which increases efficiency and power handling.

Separating the windings (that is, placing them on opposite sides of the
core) reduces (greatly) the capacitance between windings, increasing its common mode impedance.

73, Jim K9YC
TenTec mailing list
TenTec at contesting.com

More information about the TenTec mailing list