Question regarding antenna bandwidth for coax vs: open wire line

[K1JJ]

More:

[K9MB]

Tom,
That is a beautiful, compact lightweight mechanical design and the performance you
Got on the first cut was excellent and compatible with the results the W4RZL published in his August 1968 QST article. Page 47,48.

In that article, W4RZL added a 40 meter dipole and a 20 meter dipole with good results. If I did one, may just do the 40 meter pair addition.
 This design is very good mechanical design too, because it forms triangles for great stability and the sag from the top gives much more support without over stressing the wire.

W4RZL did stagger the lengths, but if you are getting good results when they are the same, fine by me. I do wonder if a lesser difference, but not the same might be wider and give a flat middle, rather than the W you mention.
Since you have to cut it down anyway, why not cut down the bottom and leave the top tue same and see what happens.  What do you have to lose and you might find that it is flat all the way to 3990 and down to 3650…
If it does give you a “W” you do not like, slowly trim a bit off the longer (top) member and record the pattern. I bet you will find a length that gives you a flat plot in the middle and even more width.
Another thing is the 43 material is not as good as the 31 material for 80 meters. The analysis article showed that the 43 bead stack peaked much higher above 10mHz and tje 31 beads were pretty good at 80 meters. More 31 beads is better, IMO- not 43 beads for 80 meters…Much higher choking impedance.
Also, the 43 bead setup will change the load impedance at 80 meters. Not a good idea to change more than one variable at a time…🤪
I will look up the article on the W2DU designs and post it again.
Great progress, Tom, but stay with it amd you will find perfection in the next iteration or the next.
73, Mike


Edit: Here is research in W2DU Bead baluns.
Note where the 43 material peaks- much higher than 80 meters.
31 still not as good as 73 beads, but better at low frequency than 43 beads.

https://owenduffy.net/balun/W2DU/index.htm

[K1JJ]

Hi Mike -

That's a good idea to trim some off the bottom legs only as an experiment.  Nothing to lose to try. My guess is it will help widen out the curve, but we will have to see.  I'll give it a try tomorrow and see if it widens out.  

Yes, I am using nine 43 material ferrites. These are the same ones used in the class E rigs on 160 and 80M.  I should end up with at least 500 ohms on 75M, so it may not be as high as using 31 material but should be enuff to do the job.   The only way I'd tear down that center insulator now is if it burned up....

Once I post some final curves, I'd like to see some other guys post their curves using whatever they have.  I'm wondering how much effect height above ground or other factors affect the swr curves.


BTW, here's a pic of the towers from about 3/4 mile away.   We are NE of the towers, towards Eu, my favorite direction.   Another simulation glitch: There are supposed to be four towers there. For some reason the 120'er is supposed to be on the left, but is not there. That is quite the mystery to me.   More proof the alien computer can't keep up...     The tower on the left is the 190'er holding the delta loops. The one on the right holds the 40M 3-stack.    The missing tower (on extreme left) is supposed to be holding one end of the fan dipoles.


T

[K9MB]

Hi Mike -

That's a good idea to trim some off the bottom legs only as an experiment.  Nothing to lose to try. My guess is it will help widen out the curve, but we will have to see.  I'll give it a try tomorrow and see if it widens out.  

Yes, I am using nine 43 material ferrites. These are the same ones used in the class E rigs on 160 and 80M.  I should end up with at least 500 ohms on 75M, so it may not be as high as using 31 material but should be enuff to do the job.   The only way I'd tear down that center insulator now is if it burned up....

Once I post some final curves, I'd like to see some other guys post their curves using whatever they have.  I'm wondering how much effect height above ground or other factors affect the swr curves.


BTW, here's a pic of the towers from about 3/4 mile away.   We are NE of the towers, towards Eu, my favorite direction.   Another simulation glitch: There are supposed to be four towers there. For some reason the 120'er is supposed to be on the left, but is not there. That is quite the mystery to me.   More proof the alien computer can't keep up...     The tower on the left is the 190'er holding the delta loops. The one on the right holds the 40M 3-stack.    The missing tower (on extreme left) is supposed to be holding one end of the fan dipoles.


T

Hi Tom,
Hey- I told you Imwas a fruitcake for details. If the beads do not explode, they will do a great job and if you have residual rF on the line, as I said, just add another choke 1/2 wave down the line (125-130 feet) and it wil scrub it up completely.
I am anxious to see the results of your trimming the bottom. I feel confident that you will see less than 1.6 swr out to 4.0mhz and less than 1.5 from 3650 to 3950 or better.
Yes- height (and length😉) matters!
73, MB

[K1JJ]

OK, will do.  Tnx Mike.

Today I'll lower the dipoles to trim the legs and also measure the coax from the feedpoint and find the 125' point.  This will be easily accessible to add some ferrite cores. I left some extra RG-11 coax in the run that I could even add some turns thru the ferrites.   That will put the balun question to bed.  

I'll also play around with the leg trimming and see what happens.

It's a circus to lower all the legs and feedline and avoid twisting and snagging on stuff.  I'd like to get it finished and add the last bit of waterproofing.  Still need to climb and finish the 40M stack feedline weatherproofing.   There hasn't been any rain for a while.... good for the work - bad for the lawn.

BTW, you said, "if you have residual RF on the line"   ----      what is the best test to determine if this is an issue?


T

[W1ITT]

Tom...  If there is RF on the outside of the line, it should show some periodicity.  If you have something to read line current...on the order of a pickup loop, a diode and a sensitive (50 uA) meter,... you can run it along the line and see if there are peaks and nulls.  Of course this exercise in nontrivial if the coax is high in the air.  But you don't have to run the whole line, just a quarter wave section of it, perhaps even less to see a trend or the absence thereof.   I think the famous MFJ may have sold a clamp on current meter that you could use, as long as there were no QC problems in it.  Mix 31 is the stuff to use to cool things off at the wavelength of interest.
73 de Norm W1ITT

[K1JJ]

FB Norm -

Your test technique makes sense. I should build one and do the test on all the antennas here while we're at it.

Mike you were very close with your swr predictions.  I carefully pruned the legs and tested the fan dipole at least 10 times looking for the best general swr curve.   I went too far and then stepped back to find a compromise that JUST  covers 3.5 to 4.0 MHz at less than  1.4 : 1  swr.    I ended up with two staggered 1:1 swr dips -  one at about 3775 and the other at 3870.   The curve is not as perfect as the first one I got yesterday, but it is broader and covers the whole band - and the dips are close to my favorite operating freqs. What's not to like?  If I get down to the AM operating areas like 3625 - 3750, etc., the swr is really quite good.  The DX window at 3790 is 1:1.   I cannot recall ever having a 75/80 antenna quite this good that was coax fed. Openwire and a matcher was the only way except for one of those load resistor antenna hack jobs.

I didn't add any new ferrites yet to the 125' voltage point as Mike suggested, but will figure something out now that I located the proper spot on the RG-11 feedline.

All in all, I would highly recommend this fan design.

It should start at about 5% shorter than the 468/f formula.  It will need to be pruned and tested after that.  Much will depend on the leg spacing and height above ground, as well as straightness and flatness.  This project took a lot of time and effort - more than I expected....

T

[K9MB]

FB Norm -

Your test technique makes sense. I should build one and do the test on all the antennas here while we're at it.

Mike you were very close with your swr predictions.  I carefully pruned the legs and tested the fan dipole at least 10 times looking for the best general swr curve.   I went too far and then stepped back to find a compromise that JUST  covers 3.5 to 4.0 MHz at less than  1.4 : 1  swr.    I ended up with two staggered 1:1 swr dips -  one at about 3775 and the other at 3870.   The curve is not as perfect as the first one I got yesterday, but it is broader and covers the whole band - and the dips are close to my favorite operating freqs. What's not to like?  If I get down to the AM operating areas like 3625 - 3750, etc., the swr is really quite good.  The DX window at 3790 is 1:1.   I cannot recall ever having a 75/80 antenna quite this good that was coax fed. Openwire and a matcher was the only way except for one of those load resistor antenna hack jobs.

I didn't add any new ferrites yet to the 125' voltage point as Mike suggested, but will figure something out now that I located the proper spot on the RG-11 feedline.

All in all, I would highly recommend this fan design.

It should start at about 5% shorter than the 468/f formula.  It will need to be pruned and tested after that.  Much will depend on the leg spacing and height above ground, as well as straightness and flatness.  This project took a lot of time and effort - more than I expected....

T

Tom,
I love the results you got and your patience in trimming paid off.
That might theoretically be termed a “W” curve, but when that middle hump is as wide and flat as I imagine it is, we can call this thing FLAT! 😉😎
Of course, your results will only be replicable if the heights are the same for the copy, but that just means a bit different length. Starting low and trimming makes a custom design easy, if a bit time consumimg.

Your idea that the lengths be shortened by 468/f -5% is very similar to the ABRA fan dipole recommendation of 468/f x 0.96
See formula here:
https://wireless-girl.com/Projects/Antennas/Broadband80mDipole.html

The fact that you have independently seen optimal results by this identical formula makes the process much more likely to be successfully copied when trimming for height and other factors are added in.
Excellent work! 😎

Now, back to suppression beads:
I have copied links to the same beads in 43 and 31 material and the images for tue curves on both and fir this suppression application, the 31 bead is clearly superior.
I know that you have a lot of packaging and time in the stack on top, but if you want a stack of beads at the 1/2 wave voltage minima for any residual common mode current, the 31 beads are best. Actually, however, a number of loops through 5- ft240-31 cores will be much better than that, giving several thousand ohms of impedance.
To answer your question about tracing down currents- like brushing and flossing daily, one need not get rotten teeth and gum disease to use preventive care..😉😂.
If you do wish to look, however- odd 1/4 wave points are obvious places to sensing field meter.
If you have any SDR equipment, the acceptable level of rf in the shack is so,ething less than zero. By the way, so keeping the devils from riding common mode paths on feedlines, rotor cables, etc is essential. Like gum disease, if you never get it, it might be doubtful if you ever would have without dental hygiene, but who wants to run that experiment either? 😱😉😂
How about publishing pics of your curves?  Outstanding results should be shown.
73, Mike
Edit: Note in the curves the Rs values. The Rs value is very important in wide band suppression and is related to loss impedance. Ot can never go capacitive like regular Inductive reactance and tjis gives much better suppression of common mode currents. MB
43 bead:
https://www.fair-rite.com/product_datasheet/PN2643102402.html

31 bead
https://www.fair-rite.com/product_datasheet/PN2631102002.html_

[KD6VXI]

If you have residual rf on the line past your common mode choke you need to find out if it's be ause it's being picked uo by the proximity of the antenna (usually because you aren't at a 90 degree angle from the feed point) or because your choke is letting it by.

If the choke is letting it by you can find problems like cracked ferrite, permeability change, etc.

This is why I got for 5k or better impedance on my chokes.  If you don't design for it, a string of beads can eventually fail at QRO.

W2DU and I had this conversation when I showed him my string of beads.  He warned me that any residual common mode could cause issues down the road.  I used a MFJ  current analyzer for open wire line to make sure my common mode was down to nill.


Also, and probably not a problem due to the remoteness of Tom's qth, but will be of interest to city dwellers, having a common mode choke removed from the feed point (as the first choke in the system) is a nono.  The reason is thus:  if you ha e noise being picked up by the shield.


The coaxial cable is actually 3 pieces of wire.  The center conductor, the inside of the shield and the outside of the shield.

Having your choke at the feed point will keep any noise ingress from among it at the outside of the coax shield from among it to the inside of the shield, where there is nothing we can do about it.

I run 2 chokes.  One at the feed point of the antenna and one at the point where the coax comes in the shack.

This keeps any noise shield currents from making it to the inside of the coax at the feed point and also keeps them from being carried in on the shield into the shack.


I know you run 'legal limit' am, Tom.  Keep in mind that you can have a substantial rf current on the outside of the coax.  Is it enough with a string of beads?  K9YC says it will take appx 50 beads to make sure your common mode choke doesn't experience issues.  Maybe he's wrong.

--Shane
WP2ASS / ex KD6VXI

[K9MB]

Jim Brown has earned his knowledge by thousands of hours of work and research, do the chance he is wrong is near zero, IMO.
I have studied his work and it is very extensive and scientific in his careful metrology.
All you say is true Shane, but even 50
Beads will not come close to a single FT240-31 wrapped with 13 turns if rg400 or better yet - two 12 wrap FT240-31 in series which
Jim’s data gives over 12k of choking impedance.
Tom wants to light those
Post 4th of July sparklers up there, though, I think, do I am supporting his determination to
Prove it one way or the other.😉😂😂😂

[K1JJ]

OK on the 50 beads. Got it covered.   Almost all of my antennas are supported in the center, so I have coax turns on ABS plastic forms - with ferrite cores as mentioned.  The only antenna I have not supported in the middle is this broadband fan dipole, thus the cores.  But as I mentioned, I plan to add an additional choke at the 125' 1/2 wave point.  I left about 25' of extra RG-11 in the run, so can easily add the choke to the system.  The turns will have to be large diameter as per specs, but I plan to wind a robust choke with cores before running any KW power levels.


How about I order one of these and wind as many turns thru it as I have room and coax?  One FT240-31 will do it at the 1/2 wave point?

https://www.amazon.com/Fair-Rite-Toroid-Core-FT240-31-Ferrite/dp/B01959LAFA


Question:  I have one of these Motorola field strength meters with a little antenna. How would I use a probe (clip lead) to sample some RF from the outer shield and make it a meaningful measurement?   I could put 20 watts thru the antenna and wind a few insulated turns around the shield... do as Norm described above?   This is one of those diode detector FS units.

T

[K9MB]

OK on the 50 beads. Got it covered.   Almost all of my antennas are supported in the center, so I have coax turns on ABS plastic forms - with ferrite cores as mentioned.  The only antenna I have not supported in the middle is this broadband fan dipole, thus the cores.  But as I mentioned, I plan to add an additional choke at the 125' 1/2 wave point.  I left about 25' of extra RG-11 in the run, so can easily add the choke to the system.  The turns will have to be large diameter as per specs, but I plan to wind a robust choke with cores before running any KW power levels.


How about I order one of these and wind as many turns thru it as I have room and coax?  One FT240-31 will do it at the 1/2 wave point?

https://www.amazon.com/Fair-Rite-Toroid-Core-FT240-31-Ferrite/dp/B01959LAFA


Question:  I have one of these Motorola field strength meters with a little antenna. How would I use a probe (clip lead) to sample some RF from the outer shield and make it a meaningful measurement?   I could put 20 watts thru the antenna and wind a few insulated turns around the shield... do as Norm described above?   This is one of those diode detector FS units.

T

I was kidding you a bit, Tom, but I was serious that your beads choke might work well enough.
Your idea about winding through several FT240-31 cores will provide a high impedance and locating it at a high current- low voltage point in any common mode currents should work great as long as you pay attention to the 8 inch radius spec for the rg11.

One possible difficulty not mentioned before I’d that the mu of a lot of Ferrite cores can reverse and drop sharply at a critical temperature.,
Fair-Rite publishes temperature curves to look at this.
If the choke has several thousand ohms of impedance, it will never get hot enough to matter, but I have heard of cases when bead chokes can experience thermal runaway and chronic overheating can degrade them as well.
It would be great if there was a thermometer that had a radio link.
I wonder if a wireless thermometer might work for that?
I have an indoor/outdoor thermometer with a 4xx MHz link to the jndide thermometer.
Not sure how much range it has though…
If it worked, the remote sensor could be taped to the bead bundle and temperature monitored to see if the temperature rises..,
Or you could pour the double 4-1000a amp in and see if it lights up or sparkles.
You have uniquely useful power to do the research…😉


It is possible that none if this gloom snd doom will materialize, however…

[W1ITT]

Tom..  That Micronta meter has most of the parts you'll need for a current probe except the pickup.  Get a ferrite split core in the plastic clamshell thing that allows you to clip it over a wire.  The hole will need to be big enough to pass the coax csble and a few turns of small wire.  On the stationary half of it, wind a few turns of smallish insulated wire.  The wire gets terminated with a resistor...maybe 68 ohms or so, then fed to a sensitive diode detector, then to a potentiometer, then to the meter.  Glue the stationary half of the split core to the top of the meter and it'll look similar to the ones that MFJ sells.  On his site, he offers two of them.  The "compact" ladies' parlor model even includes a schematic link in the description.  You'll have the only one with the esteemed Micronta name on it  after you do the modification.  Or you can spend more money and buy the "calibrated" unit and measure amps, assuming it'll pass the diameter of your coax.  Probably just knowing if there is a relatively large peak to null ratio is good enough for our purposes.
73 de Norm W1ITT

[K9MB]

Using a clamp on with a secondary winding is a great idea.
I am attaching a screen shot from KF7P site.
I would get the #61 material for this purpose. They are ideal for flux transfer.
They are over 1/2 inch so there is room for some wraps and still close and slide easily.

[Chuck...K1KW]

Well, this is an interesting thread.

Having used ferrites in antennas years ago for baluns and 4:1 ferrite 10KW transformers with KLM's  dual driven element antennas, I found out that this isn't  good idea. 

If you exceed the peak flux density the ferrite is capable of, the ferrite becomes very lossy and will heat up under normal operating design conditions thereafter.

Well, what would do this. LIGHTNING!  A nearby strike and definitely a direct hit will slam the ferrite and render it useless.

I've got useless ferrites in my museum of crapped out parts slammed by nearby lightning strikes from years ago.  Took me a few years to figure it out.  Plus an engineer from  FaiRite to verify this happens all the time! 

Bottom line, DO not use ferrites on any antennas.

A solenoid coaxial choke optimized for a parallel resonance on the band is the best way to go.  You can get 10K ohms and better this way for choke impedance.

-Chuck

[K1JJ]

A solenoid coaxial choke optimized for a parallel resonance on the band is the best way to go.  You can get 10K ohms and better this way for choke impedance.

-Chuck

Hola Chuck -

Thanks for the input.    Yes, I remember the warnings about lightning and ferrites over the years but it seems to be disregarded by the majority these days.  But it does concern me cuz bad ferrite means major antenna tower work to replace it.

I've finished the 75M  broadband fan dipole project (for now) and very happy with the results.  I'm now focusing on the two delta loops for 75M.  Maybe you can advise me on that.  I plan to remove the openwire feed system and feed them both with RG-11  and some kind of balun at the feedpoints. They will be a forced fed coaxial feedline array, standard ~ 0 > 90 degrees phased, unidirectional with 40' boom spacing at 190' apex height.

So for the baluns you're recommending, I would wind, say, 20 turns of RG-11 onto an 8" ABS plastic pipe, just like a big coil and then resonate it for 75M to bring the impedance up to a high level, like 10K?  IE, no ferrite material at all...

Please describe this a little more and how the parallel resonance is achieved and measured...


T

[K9MB]

This is a branch in the thread of this discussion and I must admit to being a bit confused and muddled by it.

The term “Balun” is in itself confusing, because we call several kinds of different items “baluns”.

Also:
I have to say, I have never thought of this aspect of lightning damage.
Interestingly, there is some literature on using Ferrite to suppress EMPs from static sparks to lightning strikes.
Check out the links below:


https://www.tvtechnology.com/opinions/power-surges-part-two

http://support.nautel.com/tips-n-tricks/shields-up/

https://disasterpreparer.com/product/emp-ferrites/


I googled antenna choke balun destruction by lightning and did not get any hits on Google.
That might not mean anything, however, if what Chuck says happens and the ferrites  may not appear to be physically damaged.
Just like human conditions lkke heart disease or cancer, changes can occur and we are-at least at first- unaware of,them.

 I would love more details on the part about increasing loss in the ferrite.
The only article I found was in the bottom link and that article says that even repeated heating of Ferrites above the Curie temperature had no long term effects and permeabilitu returned to normal  when temperatures returned to normal.
By the way, the Curie temperature is the place on the ferrite temperature curve where permeability falls off a a cliff

But back to the kind of balun we have been discussing in this thread- the common mode choke.
The entire purpose of a common mode choke is to stop any common mode currents from flowing on a transmission line.

Actually, the thing that males #31 material effective is the high Rs (loss reactance) in addition to the high inductive reactance it gives when a conductor,passes,through a toroid or bead. There is XL and Rs in any choke wound on suppression material like #31 and the higher the Rs, the more effective the choke will be.

There is a lot of literature on transmission line baluns and I believe that the whole point of these devices is to convert a differential signal in a coaxial cable to a balanced output.
I am not sure that a transmission balun is at all effective in suppressing common mode currents on the shield of the coaxial cable, however. I will research that question…

I can see how a transformer balun made from #61 ferrite or #2 powdered iron might lose that crucial low loss characteristic that makes is a great flux transfer medium, but a choke balun is at it’s best when it has the highest possible u” or loss reactance because a high loss material with a high inductance makes the best common mode choke balun.

My confusion, therefore is that I am not sure what Chuck means by the term “balun” in the context of his post

Maybe you can give more details on whatkind of baluns you have used, Chuck.

On to the coiled coax shunted byma capacitor and tuned to 80 meters:

I can see an  peril in forming a parallel tuned tank circuit from the coiled feedline to maximize the impedance.

The most obvious point is that the tuned circuit is inductive below the resonant point, but above that point, it becomes capacitive and the higher you look, the becomes first less useful and then useless and finally it actually encourages more common mode current on the shield.

Secondly, the Q of this Tank circuit that is formed. Obviously, the lower the Q, the wider the band where it will exhibit suppression of the common mode current.

It actually makes me wonder- if you were going to introduce a resonant circuit, you would not just use a quarter wave folded stub like we used back in the 60s before we had any cores to think about.
They exhibit a very high impedance to common mode current on the shield for some hundreds of KHz, I think.

I want to know more, but I am skeptical that this coil of coax parallel tuned is  better than the effective #31 choke baluns or the alternative folded 1/4 wave tuned stub which has been around a long time and seems to have a fairly wide response.

Please discuss. 73, Mike

Article on ferrites and powdered iron cores at high temperature. See pages 5-6

https://elnamagnetics.com/wp-content/uploads/library/Magnetics-Documents/Using_Magnetic_Cores_at_High_Temperatures.pdf


https://hamwaves.com/chokes/en/index.html

http://www.na0tc.org/lib/exe/fetch.php?media=technical:balun_p2_website.pdf

https://www.w8ji.com/common_mode_current.htm


https://www.qsl.net/wa3mej/Articles/Common%20Mode%20Choies/common%20mode%20chokes.pdf
Why reactive chokes are not a good idea- read page 2

[K9MB]

Here is more data and information that was published by G3TFQ.
I consider his work and the work of Jim Brown to be the best out thereon common mode choke design.

Note the graphs in the attached picture.
The thing to look for is bands that are green with a black line through them.
Red means that the choking impedance is too low to be effective.
Green means that the choking impedance is greater than 4k ohms for olive green and over 8k ohms for dark green.
The black tracer indicates that the impedance is primarily resistive- not reactive.
Those coiled up pieces of coax with air dialectric are useless at all frequencies below 7 mhz and they are reactive at all frequencies which limits their useful frequency range, so they will become capacitive above their self resonant frequency.

https://www.qsl.net/wa3mej/Articles/Common%20Mode%20Choies/common%20mode%20chokes.pdf

[W1ITT]

If one does choose to use ferrites in a common mode choke, and K!KW gives us reasons to rethink that, it is well to remember that ferrites are not closely specified as to their characteristics.  One man's Mix 31 may not be quite what the next fellow's is and, perhaps, may not always be the same from batch to batch.
Add to this the fact that there are a things coming out of Commie China that are indeed ferrite but of varying quality.
If one decides to make a choke from a cookbook design it's probably best to choose a design that shows a wide area of good performance over your frequencies of interest.  And of course there's no substitute for proper two port measurements.  As long as the system impedance is near 50 or 75 ohms, one of the little Chicom fifty dollar network analyzers  of the NanoVNA series will provide useful measurements.
73 de Norm W1ITT 

[K9MB]

If one does choose to use ferrites in a common mode choke, and K!KW gives us reasons to rethink that, it is well to remember that ferrites are not closely specified as to their characteristics.  One man's Mix 31 may not be quite what the next fellow's is and, perhaps, may not always be the same from batch to batch.
Add to this the fact that there are a things coming out of Commie China that are indeed ferrite but of varying quality.
If one decides to make a choke from a cookbook design it's probably best to choose a design that shows a wide area of good performance over your frequencies of interest.  And of course there's no substitute for proper two port measurements.  As long as the system impedance is near 50 or 75 ohms, one of the little Chicom fifty dollar network analyzers  of the NanoVNA series will provide useful measurements.
73 de Norm W1ITT  

I agree that ferrites are variable on characteristics and I asked Jim Brown about that in an email and Jim said that he  was aware of the variability, but that his cookbook allowed for that variability and if you follow it, you will achieve good results. Note! : it is best to read and study his papers before bothering him, because he has no patience for stupid questions he has already answered.)

Jim Browns results and recommendations have been echoed by G3TXQ, one of  whose articles I just posted a link for.

Given the variability, there is no evidence that a coil of coax produces sufficient choking impedance at 160 or 80 meters to be effective in eliminating common mode currents on coaxial cable lines.
Also, one of the major weaknesses is that they are reactive and therefore inductive up to self resonance and  capacitive above and when capacitive, it has been demonstrated that the problem of common mode current on the shield of the feedline.
The idea of shunting a capacitor across the coil of coax will provied a large impedance peak, but it becomes much less useful below and capacitive above resonance, so it seems like a doubtful improvement.

As for Chuck, I have the very greatest respect for his opinions and knowledge and have profited from his posts from the past. The fact that I took the time to ask questions about his assertions is in itself a token of respect and I am open to any scientific evidence that it is the best solution to this particular problem.
I do not consider myself an expert in this area, but I have read a lot of articles and found that Jim Brown has done some serious work that is solid and his conclusions and advice deserve to be heard and followed unless someone does anything close to his work and concludes otherwise. G3TXQ, W1HIS, Joe Risert and others have also made great contributions and I have great respect for them
If we begin discussing current baluns using transmission lines, my Jerry Sevick books get a lot of attention for the very same reason.

By the way, I was not aware that,Fair-Rite had their cores made in China, though it is not surprising and is certainly something that adds to my overall concern about our exporting expertise to their oppressive Communist government, but the relative merits of high Rs Ferrite chokes vs inductive chokes to suppress common mode currents on coaxial cable feedlines is a matter of empirical  scientific evidence, not political conviction.

[K1JJ]

Great discussion!  I'm learning a lot from your reports, Mike.    But my head is spinning as to how to proceed.  

Let me state a few real world assumptions to see where we can go from here...


1)  Not all antennas have serious common mode problems. When using perpendicular feedline runs and simply "lucking out" with the layout,  the antenna may perform acceptably with no balun at all.  So it becomes a matter of degree - how much of a common mode problem does the antenna have in the first place?


2)  What is considered a high enough impedance for a balun?   For a 50 ohm antenna, is X5  (250 ohms)  enough or is X10 (500 ohms)  required to bring the particular antenna down to an "acceptable" level?   Is -20 dB enough or is -30dB?   This all depends on how bad the case of common mode is in the first place.

3) When using the thinner coaxes to wind multiple turns on a large toroid, (say smaller than RG-213) will the thin coax stand up to extreme QRO?   (power like a corntester might run)

4)  If a long string of ferrite beads is slid over the coax and actually does the job, can a simple test, like a 1500 watt carrier every month or so show any lightning damage by watching the swr rise from a baseline measurement when first installed?  IE, I'm thinking that because the ferrite bead technique is so easy and effective, a simple strap test every so often to verify the beads are still OK is not a big price to pay.  After a big strike they may need replacing - just accept it and make it easily accessible if possible.  My fan dipole uses cores and is easily accessed by ropes and pulleys, so no problem there.

5) Is there a false negative test for ferrites?  IE, if they explode into tiny bits, the swr may not show a rise, but the ferrite is ineffective.  Is there some other damage mode that does not show up with swr?

6)  If the feedpoint balun is of small impedance, would an additional  "token" balun, like 15 air turns and a few in/out ferrite beads on the ground, at the 1/2 wavelength point be a good insurance to further snuff any residual problems?  The ferrites could be checked by the periodic power strap/swr test also.

I'm trying to finish up my 75M delta loops new coaxial feedline project and obviously want to continue using ferrite -  but with some conditions related to lightning and minimum impedance required to do the job on 75M..


T

[W1ITT]

Tom...
In answer to your question (2), both the VOA and the Beeb want to see balanced-to-unbalanced  (common mode) currents on the balanced lines that go back to the transmitter hall at better than 26 db down.  In measuring curtain arrays for them and others I have never seen one fail to meet spec.  I suppose that an Amateur installation that doesn't allow for acres of undisturbed symmetry wouldn't be as good. The big difficulty is in keeping common mode currents out of the measurement setup.  Measurement errors become evident when lead dress variations lead to non-repeatable measurements.
After you've incorporated the most practical choke solution, remember that common mode current is not lost, but radiates, somewhere, at some angle.  Recall Maxwell's famous deathbed statement: " RF gotta go somewhere, chaps."
73 de Norm W1ITT

[K9MB]

Tom,
I did not use any balun on 75 or 40 for decades. My max power was only 700 watts though and I only got a little burn on
My lip at times from touching a metal desk mic …😉
The bottom line is whether it is important to keep as much RF out if ghr dhsvk as possible or to just wing it.
With the old tube receivers, the RF had little effect, but I hear that modern wide band SDR receivers hooked to big led displays, keeping the shack “quiet” had benefits…

😂😂😂 Your question about using a small cooled choke in-line reminds me- perversely of the old joke about the kid who was cautioned about his behavior on a personal subject leading to blindness and he later admitted to continuing the doubtful practice only until he needed glasses…😉😂😂😂
Seriously, the use of reactive chokes has a significant risk of increasing common mode currents as well as the hope of reducing them, as intended. See the zG3TXQ article I just posted. I would skip it, myself and go with the ferrite core wrapped with rg400

If you look at G3TXQ graphs, 25 1020-31 beads will provide 500 ohms of choking at 80 meters snd the black tracer means that it is an effective 50ohms.
However, think about that bleeder resistor you put in your Hv supplies snd how you decide in it’s value and power rating.
10K across 3kV will cause a lot of heat, while a 1Meg high voltage resistor will be much cooler.
The parallel is that, like a dropping resistor  for a plate volt
Meter, you may only need 50 micro amps and ideally, the common mode current will be close to zero.
Putting a few beads on s feedline will be effective if you were not going to have much common mode current anyway and it will therefore be cool enough, but if the head voltage causing a common mode current us high, then that 500 ohms will be a hot mamma until the ceramic blows…😉
So the answer is- who knows?
Jim Brown will tell you to design for the worst case scenario.
Like your choice to support your new Fan with 3/8 Dacron, it will appear to be overkill to guys who use poly baling twine from the farm store, but you are ready for more scenarios and I like you’ll chances to get through a cold icy New England winter (are there any other kinds??😉😂

RG-400 on a FT240-31 or better- two hooked in series is the gold standard for performance.
RG400 is rated at 1900 volts and 3000 watts CW up to 100mhz, so at HF it can handle all you like.
Jim Brown also swears by awg12 Teflon insulated silver plated copper paired with tape and wound the same way as rg400, but I really like RG400 .
RG141 is similar but stiffer and it uses steel solid core with silver plating. RG400 easier to wind in my own experience, having tried both.
I put my FT240-31 chokes in a pvc box but Jim Brown advises against it. If vents are provided, it will cover his objections kn that though.
The small FT-240-31 balun scan also be dipped or otherwise sealed and mounted on a HDPE plate and it will be lighter than 25 beads on the RG11 and much more effective.
Review the pics I posted for ideas on that method by a friend of Jim’s from 6 Land.
73, MB


http://www.rosnol.com/upload/product/RG-400%EF%BC%8F%EF%BC%B5.pdf

[Chuck...K1KW]

Hi Folks,

My comments are based upon experience (some of it bad) and a bit of theory.

I believe what we are discussing is minimizing feedline radiation due to unbalanced feed from coax to a balanced dipole.  If you run the feedline away from the dipole perpendicularly for a good 1/2 wave or more to get it out of the inductive field, then you may not even need the "balun". Equal but opposite currents would be generated on the feedline on the outside of the shield but cancel out. I have done this in the past and minimal (milliamperes) outside feedline current was measured taken at several points along the feedline (important!) at 1 KW.  

Years ago I had stacked KLM yagis up for the high bands which used ferrite based baluns.  After a lightning strike VERY close, I noticed the F/B ratio on two of the antennas had dropped from the usual 25 dB or so to around 10 dB the day after the storm.  No visible damage anywhere but the SWR had also gone up a bit so I decided to replace the baluns.  That fixed the F/B problem.  I sent the baluns back to KLM and their analysis was that they had become very lossy.  I mentioned the lightning and they said they had seen this before.  My guess was that they had ceased being baluns and the feedlines now became part of the antenna destroying the pattern.  These were 5 KW baluns with extra ferrite too!   I never put more than about 2500 watts into them. This happened once again and then I went to air wound solenoid choke baluns on all the antennas, never had the problem again.  Anecdotal for sure but.....

On solenoid choke "balun" design, I use an HP 4815A vector impedance meter to optimize the # of turns on 6" PVC.  I have 2 of these on my 2 stacked 40M beams.  To optimize this design, I connected the meter to the shield on either side of the coaxial choke and adjusted the # of turns to get the parallel resonant frequency to 7200 KHz.  It took 12 turns of RG-213 to accomplish this.  At resonance the impedance was 30K ohms and purely resistive.  At 7100 it dropped to about 16K and was slightly inductive, perhaps 10 to 20 degrees if I remember correctly.  At 7300 it was slightly capacitive and about 15K ohms.  Not bad.  Clean text book pattern with no measurable feedline currents.

Having not completely adhered to my no outside ferrites rule yet, in 2010 when I put up the latest 4 stack of 10E log periodic antennas, I used two ferrite based power combiners outside mounted up the tower to reduce the feedline count  coming into the shack.  I used this combiner set up for several years with no issues until another nearby lightning strike.  The next day when I used the logs, I blew up one of the combiners after a short transmission which started out fine.  I remember watching the SWR creeping up thinking maybe some water got into something until it went to infinity.   The combiner had split into a number of pieces and looked like it was overheated since plastic enclosure it was in had started to melt.  I wound up running all four transmission lines into the shack where the ferrite combiners are.  It all gets disconnected when storms come through.  

This is why I will NOT use ferrites outside.  

Back when I designed high power SS amps professionally,  the higher permeability ferrites would sometimes get permanently magnetized if an output device failed and sent a high peak current through an output transformer.  This would skew the hysteresis curve of the ferrite and render it lossy.  The ferrites would still work but run very hot.  Avoid ferrites if possible.  We eventually changed over to powdered iron which is much less susceptible to this failure but it isn't available in a higher permeability unfortunately.

-Chuck

[K9MB]

Tom...
In answer to your question (2), both the VOA and the Beeb want to see balanced-to-unbalanced  (common mode) currents on the balanced lines that go back to the transmitter hall at better than 26 db down.  In measuring curtain arrays for them and others I have never seen one fail to meet spec.  I suppose that an Amateur installation that doesn't allow for acres of undisturbed symmetry wouldn't be as good. The big difficulty is in keeping common mode currents out of the measurement setup.  Measurement errors become evident when lead dress variations lead to non-repeatable measurements.
After you've incorporated the most practical choke solution, remember that common mode current is not lost, but radiates, somewhere, at some angle.  Recall Maxwell's famous deathbed statement: " RF gotta go somewhere, chaps."
73 de Norm W1ITT

That is very interesting, Norm.
No doubt, high power broadcast transmitters operate in a different world from Amateur service.
I would like to know more about the feedline systems you
mentioned being balanced lines. I had assumed that the feeds were differential in
Massive rigid gas filled heliax feedlines.

The -26dB spec would make the differential power at 1/400 of the differential power.
In an amateur transmitter, that would be 250mw  milliwatts per KW of Differential power delivered to s ham antenna- not much radiated, but may be adewuate amount to tolerate getting into the shack maybe.
So do these big rigs use large current baluns to convert balanced to Differential feed?

I can see what you mean if the balun has a -26 dB leakage into the common mode.
In the case if suppression chokes, however, the high RS value
Prevents the current to flow significantly and the only significant RF left is in the
Infrared
Spectrum and measured in heat dissipated, not Hf and for low power like Hams use, that
might be a good solution to
The common mode problem.
Please tell me
More about the baluns use on these mega transmitters though. I find it very fascinating.
Any pictures to share?