Feed line choke windings - coax or wire?

[JHNC]

I am planning to construct a feed line choke for a 9:1 end fed wire antenna. Research reveals that using this type antenna, the coax outer braid is used by the antenna as the counterpoise.  I am getting RF in the shack despite having a separate wire counterpoise attached the toroid. I have 2 questions on construction:

1. I have seen examples of choke toroids wrapped using both coax [RG-58 in this case] and toroids wound with stranded wire, as is the toroid in the 9:1 transformer. www.earchi.org/92011endfedfiles/Endfed6_40.pdf I am wondering, but cannot find any information, as to whether a choke toroid should be wrapped using coax or 2 lengths of insulated wire. It seems to me that using coax for the choke toroid is the best method as I am trying to keep as much RF off of the outer braid as possible. Using insulated wire seems to be best for the 9:1 toroid as we are trying to transfer as much power to the radiating wire as possible. Am I correct here?

2. Using either coax or insulated wire, I have seen some 9:1 toroids and choke toroids wound with a "crossover" at the 1/2 way point, that is; for a 12 turn toroid, 6 turns are wound, then the coax or wire is passed thru the center of the toroid and the last 6 turns wound in the opposite direction. I believe the reasoning had to do with reducing capacitance issues [?].
Any thoughts or guidance on this?

[A.Z.]

Start by carefully reading the following

http://www.karinya.net/g3txq/chokes/

https://www.dj0ip.de/rf-cmc-chokes/different-kinds-of-chokes/

then decide, keep in mind that, to tame CMC on all HF bands you will probably need more than one choke

My usual "all rounder" is a choke wound around an FT240-43 using 17 or 18 turns of RG-174 with windings "Joe Reisert" style (that is a cross turn)

Notice that, with an endfed you will NEED a counterpoise AND a choke so, please, add a counterpoise to that antenna, it may be just a run of insulated wire dropping down to ground and laid down there, but you'll need it to both allow the choke to work and allow the endfed to radiate properly

Also, if possible, keep the antenna (the feedpoint too) as high and "in the clear" as possible

Oh, and since we are at it, here are some additional notes which may be of help

https://www.eevblog.com/forum/rf-microwave/putting-up-a-random-wire-antenna/

HTH

[Wallace Gasiewicz]

Are you saying that you are using a 9:1 transformer with your end fed antenna?
If so, what is your standing wave? Can you measure the impedances of your antenna at the frequencies used?
A "choke" is typically used just to reduce "common mode" RF that is traveling down the shield of coax and this type of choke can just  several wraps of the feed line coax around a plastic pipe, commonly called an "ugly balun".  Usually 3-4 inch PVC pipe. You can also use a toroid. Since the feedline coax shield is used as the "counterpoise" the feedline choke is usually near the transmitter, but this means the rest of the coax is radiating and is in the shack itself.
So part of the antenna is actually in the shack.

[A.Z.]

Are you saying that you are using a 9:1 transformer with your end fed antenna?
If so, what is your standing wave? Can you measure the impedances of your antenna at the frequencies used?
A "choke" is typically used just to reduce "common mode" RF that is traveling down the shield of coax and this type of choke can just  several wraps of the feed line coax around a plastic pipe, commonly called an "ugly balun".  Usually 3-4 inch PVC pipe. You can also use a toroid. Since the feedline coax shield is used as the "counterpoise" the feedline choke is usually near the transmitter, but this means the rest of the coax is radiating and is in the shack itself.
So part of the antenna is actually in the shack.

well, if it isn't an EFHW which needs a 49:1 UnUn but a well "calculated" endfed, there's nothing strange in using a 9:1, no "magic" just a matter of using a wire length which in NOT resonant (neither 1/4 nor 1/2 or multiples), that way, with a proper antenna wire length, the impedance (on ham bands)  could be handled by the 9:1 UnUn and transformed to something which is "near" to the characteristic impedance of the feedline, at that point, an antenna matching unit (an L match works fine in own experience) will make the transmitter happy, then... ok, in RX an endfed isn't exactly the best, but... pair it with an RX ONLY loop and the results will be surprising

[JHNC]

Wallace Gasiewicz: When I refer to the 9:1 toroid, I am referring to the antenna's internal toroid. Please refer to the link for details of construction.A.Z.: I will review your links and will be back if I have any further questions
Thanks to you both for the replies.

[A.Z.]

Wallace Gasiewicz: When I refer to the 9:1 toroid, I am referring to the antenna's internal toroid. Please refer to the link for details of construction.A.Z.: I will review your links and will be back if I have any further questions
Thanks to you both for the replies.

you'll find me here, feel free to ask !

[Wallace Gasiewicz]

I use a homebrew 9:1 with a G5RV. The 9:1 is homebrew and works just fine.  I got the plan for making it from someone in a Kiwi Ham Radio group, since I wanted one that would handle more power so I wound a big one myself on a toroid.

[A.Z.]

I use a homebrew 9:1 with a G5RV. The 9:1 is homebrew and works just fine.  I got the plan for making it from someone in a Kiwi Ham Radio group, since I wanted one that would handle more power so I wound a big one myself on a toroid.

a bit of caution here, a g5rv is a balanced antenna (and should be fed using ladder line, so I can't see where you are using that 9:1) so it will eventally need a BalUn, an endfed is an UNBALANCED antenna, so you will need an UnUn to feed it

[HB9EVI]

I'd advise to keep it simple, so: if your feedpoint impedance is 50 , use a CMC wound with coax, RG-174, RG-316 whatever, W1JR-style is doing just fine and reduces capacitive coupling

if your feedpoint impedance is any other than 50 , wind a balun for undefined impedance how DG0SA (sk) describes in his worksheets https://www.dg0sa.de/balun1zu1undefklein.pdf

trying to avoid the unholy 1:9-magic-balun-whatsoever nonsense discussion, I'd recommend to analyse the impedance of an antenna thoroughly - in almost all cases any random wire antenna will be way off Z=450+j0 - so the 1:9 thing is just another dummy load in its own way

[A.Z.]

I'd advise to keep it simple, so: if your feedpoint impedance is 50 , use a CMC wound with coax, RG-174, RG-316 whatever, W1JR-style is doing just fine and reduces capacitive coupling

if your feedpoint impedance is any other than 50 , wind a balun for undefined impedance how DG0SA (sk) describes in his worksheets https://www.dg0sa.de/balun1zu1undefklein.pdf

trying to avoid the unholy 1:9-magic-balun-whatsoever nonsense discussion, I'd recommend to analyse the impedance of an antenna thoroughly - in almost all cases any random wire antenna will be way off Z=450+j0 - so the 1:9 thing is just another dummy load in its own way

sorry, but I disagree, if we are feeding a "random" (non resonant end fed) wire antenna with coax, we should try our best to  keep the antenna feedpoint inpedance (on all/most bands of interest) to a value we can transform to something near the feedline impedance, that's why, for the "random" we choose some given lengths and why we use a 9:1 transformer (unun)

That said, please, would you be so kind to tell me how you'd go to feed a "random" w/o using a remote antenna tuner ?

[A.Z.]

Are you saying that you are using a 9:1 transformer with your end fed antenna?

Yes, it's a NON resonant end-fed, see here for further infos

https://www.hamuniverse.com/randomwireantennalengths.html

https://udel.edu/~mm/ham/randomWire/

this allows to have a feedpoint impedance which isn't too low or too high on the ham bands so the 9:1 is able to bring the impedance down to a value "near" the one of the coax feeder

If so, what is your standing wave?

here's the SWR

Code: [Select]


 1.850 - 2.00:1
 3.780 - 1.60:1
 7.150 - 1.65:1
10.130 - 2.10:1
14.200 - 1.35:1
18.130 - 2.10:1
24.900 - 3.00:1
28.450 - 3.50:1


at that point an ATU in the shack has no problem in finding a match and making the TX happy

A "choke" is typically used just to reduce "common mode" RF that is traveling down the shield of coax and this type of choke can just  several wraps of the feed line coax around a plastic pipe, commonly called an "ugly balun".

The "ugly balun" has a very narrow bandwidth, I prefer chokes wound on toroids, since they offer much better performance over the HF bands, a good idea is to place a choke right after the UnUn and a second one right before the coax enters the shack/building, then if desired additional chokes may be added along the feedline

Since the feedline coax shield is used as the "counterpoise" the feedline choke is usually near the transmitter, but this means the rest of the coax is radiating and is in the shack itself.
So part of the antenna is actually in the shack.

And this isn't the case if you use a counterpoise system and chokes along the coax, so since installing a counterpoise isn't so difficult, I prefer using it and avoiding to use the coax as part of the antenna, also since it causes a number of issues in both TX and RX

[HB9EVI]

depends how I'd use the random wire.

if it shall serve for a fixed frequency/band only, I'd maybe consider using a matching transmission line as impedance transformation network - if I wanted to avoid at all costs an L/C-network for impedance matching. And certainly I'd go for a coax CMC befor the transformation network.

a random wire for multiband use is a different topic, where I don't see much more options than a set of switchable L/C.
If I'd go down that road (and I actually did), I'd use a CMC based on coax before the L/C-network, so on the 50ohms side, since it's less critical than doing the balancing with a CMC for undefined impedance after the L/C-network - but it's not impossible to do it that way too.

this can be done with simple automatic asymmetric tuners too, where you just have to watch out that the tuner is entirely off GND potential; to my practical experience so far that works ufb

and I don't longer discuss the 1:9-balun topic; it already has been said all to that on many places over many years; it's just not possible to kill this topic once and forever; it comes back all the time like hemorhoids or the common cold - and is about as useful as the latter two things.

[A.Z.]

depends how I'd use the random wire.

if it shall serve for a fixed frequency/band only, I'd maybe consider using a matching transmission line as impedance transformation network - if I wanted to avoid at all costs an L/C-network for impedance matching. And certainly I'd go for a coax CMC befor the transformation network.

a random wire for multiband use is a different topic, where I don't see much more options than a set of switchable L/C.
If I'd go down that road (and I actually did), I'd use a CMC based on coax before the L/C-network, so on the 50ohms side, since it's less critical than doing the balancing with a CMC for undefined impedance after the L/C-network - but it's not impossible to do it that way too.

this can be done with simple automatic asymmetric tuners too, where you just have to watch out that the tuner is entirely off GND potential; to my practical experience so far that works ufb

and I don't longer discuss the 1:9-balun topic; it already has been said all to that on many places over many years; it's just not possible to kill this topic once and forever; it comes back all the time like hemorhoids or the common cold - and is about as useful as the latter two things.

Well, that's up to you; in my case the point was installing an antenna to cover as much HF bands as possible and with a length which was below 30m, going vertical wasn't an option, either, so at the end I went for the random and it worked, then ok, at another location I've different antennas, but for that particular one, the random works quite well for me. As for the 9:1 I don't think it's "magic" or whatever, it's just the right tool for the job (in THIS case), sure, installing a remote ATU at the antenna feedpoint would be better, but I didn't want to sheet out the money for it nor to collect the bits and parts and build it, so the use of the UnUn in this case was the cheapest and easiest way and, all in all, the results aren't all that bad; YMMV by the way

[HB9EVI]

well, my inverted L is random length too, I used the maximal available space I could get; maybe there'd be a frequency where 450+j0 could be found, but this is not the case on the band I built it for (2200m/630m/160m/80m).

From my experience in discussions about antennas, people tend to overestimate their antennas, in length, in height, in gain, in the radiation pattern; so people want to work on 80m or 160m and just don't realize that the feedpoint impedance for any of those random wires normal ham operators can build (so without being owner of hectares of land around the house) is not only far away from 450+j0, no it's commonly maybe just around 5-j1500 - but they still believe that the 1:9 is their way to go for it.

but I guess we drifted off the original topic... the question was, how to wind the CMC; and from my experience it's often easier and more convenient to do symmetry on the 50ohms side, even with asymmetric tuners - in the most convenient case you have an automatic tuner which can be fed by a bias-tee, so DC+RF over the 50ohms CMC and you're ready to go, without further CMC for the DC supply - but even this is no big thing.

[JHNC]

Good discussion, folks. I have read all of the replies and links. I still am seeking an answer to my first question. I will rephrase it:

For A CMC, we wind the choke's toroid with coax as we are trying to impede the flow of current on the outer braid, but not the center conductor and therefore; keep excessive RF out of the shack. Am I correct?

And, for an antenna transformer, we wind the toroid with 2 or more separate wires as we are trying to induce maximum current in the toroid and therefore; transfer maximum current into the radiating element. Am I correct on this also?

[HB9EVI]

hi there

well 'common mode choke' already tells what it's doing - suppressing common mode signals, that's about it, never mind from what you wind the CMC, coax or single wires. when using single wires, you have to watch out for the correct winding sense, but otherwise it does the same.

to use coax for a CMC is simply due to stay within the 50ohms system impedance; you can reach almost 50ohms by using two twin-lead wires; they have about 100ohms, so 2 in parallel give you again about 50ohms; but this way the wires take more space on the core than with let say a single RG-174, so you maybe reach lower common mode rejection especially on the lower end of the frequency spectrum in question

if we talk about things like a 1:4 or a 1:9 balun, the term 'balun' is often misleading. A 1:4 balun is 'transforming' impedance from 50 to 200ohms, it is a transformer, so it isn't a 'balanced' signal at the end, and it doesn't block common mode signals; so to be 100% correct: in an impedance transforming network, you still need a CMC to block common mode signals and reach full symmetry

[Wallace Gasiewicz]

For A CMC, we wind the choke's toroid with coax as we are trying to impede the flow of current on the outer braid, but not the center conductor and therefore; keep excessive RF out of the shack. Am I correct?

You are correct. However you can make things more complicated and perhaps better by separating the wired in the "choke toriod." I have not done this. The Choke just is there to prevent RF from traveling down the shield of the coax back into the transmitter. I would think I would put the choke at the wall of the shack. I dont think that the snap on ferrites are good enough, I think you need a really big toroid that you can wind the coax around. (Hamfest time)
The actual "load" (antenna) is then the long wire and the shield of the coax before the choke. A Counterpoise makes the other half of the antenna (the part that is not the long wire) different  and changes the impedance of the load, probably by a lot.

The transformer (the unun), transforms the signal into the antenna impedance. This impedance will be different for different freq. The idea is to get a transformer to get some sort of impedance that is workable on most bands. Length of wire is also important here.
I have an antenna tuner that I believe is meant for this sort of thing, it is a Sunair contraption that is automatic and meant for use with a specific tranceiver, the Sunair RT 9000. It is an enclosed watertight big box that is meant to go at the base of an antenna. There is memory involved and you can scan individual freq on several bands fairly rapidly. I believe that it was designed for shipboard use on steel ships where the Ground is the ship. Unfortunately for me the control connection is maybe 23 wire cable and I only have a 6 foot cable, not enough to put it outside.
I think the antenna they were using was not necessarily a long wire but some sort of multi freq antenna maybe something like a really big cone or eggbeater, or just multiple wires.

Here is an involved article illustrating the long wire and an Auto Tuner.

https://vk6ysf.com/longwire_antenna.htm

Whether you have a counterpoise or not, if there is significant mismatch on the line between your tuner and your antenna, there will be RF feedback. I believe the idea here is to get the impedance in some reasonable range and use the choke to keep RF feedback to a minimum. I think the 9:1 is popular for this use.
Here is an advertisement that shows the basic idea of a long wire nicely:

https://palomar-engineers.com/tech-support/tech-topics/best-hf-end-fed-antenna

The counterpoise idea has a lot of different forms. Sometimes another long wire is used.

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

Note that a Balun (balanced to unbalanced) is used here not the Unun used with a counterpoise.

Interesting topic you introduced, Thanks.

Wally KC9INK

[HB9EVI]

maybe also easier to understand:

a CMC forces the currents to be equal/symmetric (differential mode) on both the center conductor and the shield, the sum of both currents has to be 0, so every other current (common mode) like sheath current will be blocked

[A.Z.]

depends how I'd use the random wire.

if it shall serve for a fixed frequency/band only, I'd maybe consider using a matching transmission line as impedance transformation network - if I wanted to avoid at all costs an L/C-network for impedance matching. And certainly I'd go for a coax CMC befor the transformation network.

Using a "random" (no so random) for a single band is a bit of nonsense, since such an antenna is usually put up if/when one doesn't have room for another antenna, be it an EFHW or another one, but this doesn't mean that one shouldn't take care of installing the "random" in a proper way

As for CM chokes, my usual "go to" reference are the following

https://www.dj0ip.de/common-mode-chaos/

https://www.dj0ip.de/rf-cmc-chokes/different-kinds-of-chokes/

http://www.karinya.net/g3txq/chokes/

in particular, the third link will clearly explain why, using a coax wound (in air) choke (aka "ugly balun") isn't a good idea if your antenna is covering more than a single band

[HB9EVI]

well, from the point of possible antenna gain and/or radiation pattern it can be useful not to use a resonant wire length for a single band

and from the way I understood the original question we are talking about a CMC on a toroid; certainly it matters as well what material you're using. I'm not going to use a Fairrite material -61 for 80m or 160m, likewise I would not use material -77 for 10m. there's no 'one size fits all'. 

[JHNC]

As far as which mix toroid to use for CMCs: from what I have discerned from the linked articles in this post and a couple of other sources:

For 80m thru 10m, use type FT-240-43           [RFI Suppression 5 - 500 MHz];
 
For 160m, 80m & 40m use type FT-240-31     [RFI Suppression 1 - 500 MHz].

As far as power handling goes;  [per W1JR] RG-58 will work for a few hundred watts.
For higher power, using thin Teflon coax, such as RG-142 should be good for 4KW @ 10MHz ;
2KW @ 30MHz  *** estimated *** .

[A.Z.]

As far as which mix toroid to use for CMCs: from what I have discerned from the linked articles in this post and a couple of other sources:

For 80m thru 10m, use type FT-240-43           [RFI Suppression 5 - 500 MHz];
 
For 160m, 80m & 40m use type FT-240-31     [RFI Suppression 1 - 500 MHz].

As far as power handling goes;  [per W1JR] RG-58 will work for a few hundred watts.
For higher power, using thin Teflon coax, such as RG-142 should be good for 4KW @ 10MHz ;
2KW @ 30MHz  *** estimated *** .

correct, willing to go up with power just stack a pair of FT240 toroids one over the other and wind the choke; in my own experience 17 or 18 turns of RG142 around a #43 toroid will serve you pretty well on all HF bands down to 80, if you want to go down to 160 add a second choke wound on #31; if you want to go for a "belt and suspender" approach, place the chokes in three points, one at the antenna end of the coax, another at half length and a third one at the point where the coax enters the building/shack, at that point you may also want to add a gas discharging device

HTH