Open Wire Feedines

[VK2FAK]

HI all...


Yes point taken.......difference between about 1.6 or so and 30- 88...


John

[VK2OMD]

Yes point taken.......difference between about 1.6 or so and 30- 88...

Well, the other way round... but water is not an insignifcant dielectric. Of course water on your feed line is likely to be polluted by particulate matter and growth like lichens etc... and who knows what its permitivity and loss tangent are... but it is probably a significant dielectric, even if only a thin layer.

By now you have probably seen articles with people reporting a 10% reduction in vf for some simple scenarios.

So, lets look at the numbers for an example scenario...

A wavelength of lossless 300 ohm line transforms a 50R load to 50R, VSWR(50)=1.000 ... no surprises there!

A 1.1 wavelength of lossless 300 ohm line (due to rain) transforms a 50R load to 75+j209,, VSWR(50)=14 ... so it is significant! This is going to spoil a fixed matching system severely. Of course the effect is greater for longer line sections.

So, for all the people who had notice ATU settings change at night on their HF antenna fed with open wire line, especially ribbon or windowed ribbon, there is part of the explanation.

Owen

[GM3SEK]

I noted that he ignores the effect of rain on vf... so he is wrong in the inference he makes that the system is insensitive to rain. It is sensitive to rain because the electrical length of the line sections is altered due to rain effect on vf.

Antennas are usually affected by rain to some extent, open wire feed lines are affected, more so if they use ribbon lines... and in the latter case, vf is typically the most changed parameter. There is lots of Ham discussion about the effect of rain on Zo, but little about vf. SInce open wire line is commonly if not mostly used with standing waves, the vf change is very relevant.

Someone already said that if you want a weatherproof system, look to coax. I quite agree.

Not necessarily. Correctly designed with attention to detail, "open wire" line at 70cm can be practically immune to water. I know this sounds wildly unlikely, so let me explain...

As a beginning moonbouncer in the late 1980s I had endless trouble with water getting into the coax phasing system - my first 8-Yagi array had no less than 13 points for water entry. Like many others at that time (including VK3UM) I found the best solution was to stop trying to keep the water out, and do the exact opposite: convert to an open-wire feed system that was completely exposed to the weather!

The key point is to design the open-wire feed system to minimize the sensitivity to water. The main requirements are: 1. Use low-impedance lines to avoid high voltages. 2. Avoid creating any high-impedance points that would be sensitive to water collecting, which means operating the line at a low VSWR (best of all, at a VSWR of 1). 3. Use an absolute minimum of spacers, and make those of PTFE because it sheds water.

No commercially made line can meet these requirements. You have to make your own.

In the late 1980s, many 70cm EME operators had good results with open wire made from heavy 3mm enameled copper wire with a typical spacing around 12mm. Done properly, this can have very low losses and proved far superior to coax in EME Yagi arrays; and when used with care it was quite insensitive to rain. If properly designed, it can be sprayed with a garden hose with only minimal change in VSWR for the whole array. However, this type of line has mechanical problems because it's absolutely vital that the wires are kept straight, but the line is too heavy to support its own weight while it is being installed. (The best way to build it is on strips of wood, which are only removed after until after the line is safely suspended between the Yagi feedpoints. Not very nice, and totally useless for portable work.) Even after installation, the line is still quite liable to twist and the soldered joints can break. Over time, water tends to get underneath the enamel so the working life may be only 2-3 years. When the copper beneath the enamel starts to go black, it's all over.

Still, no complaints about the 3mm copper lines. That 8x21el array worked a lot of goood stuff on EME, and still holds the 70cm world record with ZL3AAD.

In the 1990s I replaced that array with 12 Yagis with full polarization rotation, and also upgraded the "open wire" to rigid 200-ohm lines made from 1/4in ali tubing. http://www.ifwtech.co.uk/g3sek/eme/pol2.htm - scroll down to Feeding 12 Yagis. The centre-to-centre spacing was calculated for the impedance required, and was about 25mm. Because this line was operating at a VSWR of 1, there was no impedance transformation so the length was determined only by the optimum stacking distance. This rigid line needed even fewer spacers than the 3mm copper, so the garden hose had no effect at all. The only weather that stopped this array was thick frost on all the elements... until property developers built houses in the field right next to it.

There is a certain "old is good" theme in ham radio, and a revival of older techniques is popular... popular but not necessarily better.

Exactly... but neither is it necessarily worse. The key point is that ALL techniques - both new and old - need to be applied with some thought. This was an example where open-wire line proved better than coax in a very unlikely application; but only because it was done in a particular way. Brainlessly parroting a slogan like "Open wire is best!" would have gone disastrously wrong.


73 from Ian GM3SEK

[VK2OMD]

...
Exactly... but neither is it necessarily worse. The key point is that ALL techniques - both new and old - need to be applied with some thought. This was an example where open-wire line proved better than coax in a very unlikely application; but only because it was done in a particular way. Brainlessly parroting a slogan like "Open wire is best!" would have gone disastrously wrong.


73 from Ian GM3SEK

You are up bright and early OM!

The tension that exists on 70cm is between keeping line spacing narrow to reduce unintended radiation and local noise pickup, and a line construction that is stable without providing the surface to get polluted and retain water. Ribbon solves the mechanical problem, but creates the sensitivity of vf to moisture.

Narrow line spacing (around 0.01 wl) means quite small conductors at 70cm, requiring more support, and of course lossier... but still better than LMR400.

Now the effects of wide spacing will not be revealed when one looks up the magic VE7BQH antenna tables... the underlying models don't include the feed line and cannot capture the pattern or noise effects of feedline contribution to radiation / pickup. These incomplete models seem to be accepted as the be-all and end-all by buyers and sellers.

I use home made open wire line with tensioned bare copper wires spaced 50mm and rigid spacers eveyr 500mm on a HF multiband antenna. The antenna system tuning is senstive to rain, but I think mainly due to proximity of vegetation rather than feed line issues as that sort of construction is, as you note, less subject to vf variation with moisture.

Owen

[VK2OMD]

...
There is a certain "old is good" theme in ham radio, and a revival of older techniques is popular... popular but not necessarily better....

Let us compare the loss of a home made 300 ohm open wire line suited to 432MHz (meaning low radiation) vs an inexpensive low loss coa.

The scenario is 3wl of copper open wire line with vf=0.9, and the same physical length of LMR400, both with 50 ohm antenna load. (This is a practical length of a phasing line section for a 4 x bay of long Yagis.)

There isn't much point evaluating conductor loss in a line that has serious radiation loss, so using the graph I posted earlier lets choose a line spacing of 0.16 wl, that's 11mm. For Zo approximately 300, conductor diameter is 1.5mm. Typical TV ribbon will be worse than this.

Using TWWLC, I calculate the length to be 1.874m and the loss under mismatch conditions specified to be 0.204dB, and using TLLC, the same length of LMR400 to be 0.163dB.

The coax in this scenario has slightly less loss.

The example Ian recommended of 3mm conductors spaced 12mm has lower Zo (220) which is not a problem in this particular scenario, and loss in 1.874m is 0.101dB... quite less than the same length of LMR400. Using this type of line, if you were obsessive about the equivalent noise temperature of the loss in the feed line that connects the individual Yagis to a single feed point where you can deploy an LNA, you might not start with 50 ohm antennas, you might start with 200 ohm antennas and so reduce the loss under mismatch to something more like 0.043dB or a 10K noise improvement over the LMR400.

A hidden loss of open wire line when used in multiples of half wavelenth is that there could be up to an extra half wave needed to make the distance, whereas the matched coax need be no longer than the physical route, greater relative impact on shorter lines.

Owen

[VK2FAK]

Hi all.....

Owen ....interesting comparison ....with using the coax. did you take into account the use of a whole heap of coax connectors..to join those cables....?

Ian ..thanks for your 2 cents worth..interesting about the alun phasing lines, I did see another similar to that but the separation looked wider... as my testing was going to be on small beams...the array fixture would be pretty rigid .
Ian .. the question I would be asking is, if you were building now, would you go right to coax or would you be thinking Open Line....?


John

[VK2OMD]

Owen ....interesting comparison ....with using the coax. did you take into account the use of a whole heap of coax connectors..to join those cables....?

There is a Ham myth that the loss in connectors is substantial. It is almost always uttered by 'experts' who have never measured it!

I have given you calculated values for different line sections alone for comparison purposes, given to three decimal places as the losses are quite small, but relevant. Of course in implementation, you will never achieve perfection, but in terms of the ultimate achievable the numbers show that the outcome doesn't depend simply on rules of thumb like "open wire lines are better".

Now lets think about your "whole heap of coax connectors". In a minimal implementation where the antenna has been provided with a single coax connector, and you use a combiner with coax connectors for all inputs and outputs, you add just two more coax connector pairs in the loss path by cabling up for an array of 4 Yagis compared to a single Yagi. Sure, there are 8 more connectors, there are 4 coax cables, but they are NOT in cascade in loss tems and emotive language like "whole heap of coax connectors" disadvantages you in finding the correct answer.

It would be naive to think that a connector pair is perfect, of course there is some imperfection that leads to less than perfect power transfer. It would be equally naive to think you can join open wire lines at a junction without the same issues. How do you bring multiple open wire lines to a junction without disturbing Zo by the proximity of the other lines at the junction? The question is about the magnitude of disruption.

Ian ..thanks for your 2 cents worth

I know you didn't intend it, but the comment devalues Ian's comment.

VE7BQH publishes tables of Yagi performance that are used by many hams as the definitive reference of the state of the art. If you were to implement a minimal feed system to deliver a single coax connector interface using LMR400 for the largest array in the 432MHz table, you would expect around 0.4dB of loss. Note that this feed point would be in space in the plane of the DEs... it really is the least length of coax in a H configuration to a common feed point. That loss represents an equivalent noise temperature of about 30K which is not included in the noise analysis, ie it is treated as lossless / noiseless. Common implementation where the LNA is fixed to the mast have higher loss than mentioned, though lower loss cable can help reduce that.

The issue you are grappling with is a real one, and very important on extreme performance systems and as inexpensive LNA noise figures nudge several tenths of a dB the tenths of a dB loss in front of the LNA becomes more and more significant.

As a learning exercise, you could build the antenna with TV ribbon and get it working. You will learn heaps from it, but it won't be extreme performance... you aren't seeking that. You can then evaluate the effect of wetting the antenna and feed system, noting that new cables tend not to wet like older cables do. Some Ham testers have used detergent (probably ionic detergent) to simulate aged cable, but it is doubtful if it is a good simulation.

Owen