Custom UHF Yagi antenna build

[darth boozer]

Log periodic antennas are generally preferred for wideband operation, while Yagis are essentially narrowband.

https://k3rrr.com/diy-log-periodic-dipole-600-700-mhz-tv-antenna-not-coat-hanger/

 

[Baluncore]

Log periodic antennas are generally preferred for wideband operation, ...

The Log-Periodic Dipole Array was originally analysed and described as being a series of 3 element Yagis. It turns out that, when you look closely, significant RF currents only flow in two adjacent elements, which makes the LPDA worse than a 3 element Yagi. The advantage the LPDA has, is in the monitoring or surveillance of entire bands, when an LPDA is mounted on a rotator at the top of a tower.

... while Yagis are essentially narrowband.

Three element Yagis are not narrowband. The bandwidth of a Yagi is dependent on the number of elements. The maximum gain is about 16 dB. Beyond about 24 elements, there is no gain advantage, just more wind buffeting and ice damage.

The LPDA is only used for Digital TV when the transmit site RF frequencies cannot be first identified, such as when the antenna is sold nation-wide by a chain store.
For DTV, if you know the RF frequencies, a short four element Yagi will outperform an LPDA.

 

[Baluncore]

Attached is a prototype 6 element YO file, RM_6.YAG as text, with element dimensions. More than 6 elements starts to limit bandwidth or gain.
The performance of the prototype is here.

 

[sophiecentaur]

For DTV, if you know the RF frequencies, a short four element Yagi will outperform an LPDA.

It depends a lot on your requirements and one should always start from there. Looking up at all the antennae on neighbouring houses can be helpful in choice of what to buy or make.

I live a long way from a main transmitter and the multiplexes sit over a very broad range of frequencies and powers. So I'm starved of received signal strength here and there are no multi-path generators nearby. That was the major factor in my choice of a twin yagi array. Someone near the transmitter with an office block nearby would have totally different requirements.

Designing and building is a long job and the only basis for accurately judging performance is A/B comparison against the last one you put up. But construction can be fun and you end up with a cupboard full of bits that are suitable for marks 4,5 and 6 models. 3D printing is a godsend.

At least antenna building is a daytime sport so you still have viewing time in the evening.

 

[Guineafowl]

I’ve made a first attempt with a designer:

This is Yagi calculator by VK5DJ, v.2.6.20.
3.15mm is the diameter of my welding rods, which I have plenty of. Coax not selected as it doesn’t have RG-6.

And the performance - not quite as good as Bauncore’s:

 

[Baluncore]

You have better impedance matching, but the clustering of the elements into closely coupled groups, suggests a deviation from reality. The clustering makes it more like a 4 element response. You might change the Tradeoffs, and turn on the optional resistive losses.

3.15mm is the diameter of my welding rods, which I have plenty of.

Are those bronze rods?

 

[Guineafowl]

You have better impedance matching, but the clustering of the elements into closely coupled groups, suggests a deviation from reality. The clustering makes it more like a 4 element response. You might change the Tradeoffs, and turn on the optional resistive losses.

Are those bronze rods?

Mild steel. Using thicker rods (I have lots of 12mm steel) and a folded strip dipole seems to improve things:

I can’t see how to turn on the resistive losses, but I may have a different version. Yours gives dBd where mine gives dBi, for example.

 

[Baluncore]

Mild steel. Using thicker rods (I have lots of 12mm steel) and a folded strip dipole seems to improve things:

Bare iron or steel will not work for resonant antenna elements at UHF. It has too much surface resistive loss. Galvanised wire, like a wire coathanger, might work for a prototype, but it will not last. You really need aluminium, copper, bronze or brass. As tube, rod, or flat strap.

Also, YO runs on DOS, so only permits 8 letter filenames. Then they become "number~1".

I can’t see how to turn on the resistive losses, but I may have a different version.

YO version 5.10 from the link I posted.

 

[Guineafowl]

I have some 15mm copper pipe lying around, and some 3/16”(4.7mm) cupronickel brake pipe for the folded dipole. Also a bagful of 15mm pipe clips.

In both of our YO calculations, the elements are around 100mm long, but the half-wavelength of 641.4 MHz is around 233mm. The yagi designer gives elements sized more like the latter figure, as well. Which is right?

 

[Baluncore]

Avoid expensive new materials when prototyping. I use scrap aluminium sheet, cut into strips with a guillotine. For this design, I would use 12 mm strips that will pop rivet onto the boom. The folded dipole is then easy to fold.

Which is right?

Double the YO element length to get the actual dipole length to cut.
It is mentioned in the documentation, and comes from the legacy file formats of previous versions. The NEC software takes advantage of symmetry about the boom.

 

[Averagesupernova]

Avoid expensive new materials when prototyping.

I'd agree here, but I'll take it a step farther. If a person likes tinkering with antenna building you might be wise to get some small diameter brass tubing. I've used some that is about the size of a brass welding rod. Where I come from that would be spec'd in inches, specifically .125 inches. Two different sizes will telescope, one into the other. Give one a slight pinch to grip the other. Lengths can then be played with and adjusted. I used the appropriate size copper wire at the base of the element that had a wavy bend or two so the brass tube would grip the wire. I did not use this scheme for a beam so I wasn't attaching anything part way up the element for a matching network such as a gamma or Tee match. Not sure what you'll be using at 600 plus MHz but the same idea could be applied. Various sizes of tubing that slide one over the other. The brass tubing I bought came from a hardware store here in the USA. I suspect such a thing can be found at a hobby store also.

 

[Baluncore]

Various sizes of tubing that slide one over the other. The brass tubing I bought came from a hardware store here in the USA.

Telescoping brass tube comes in 1/64" increments. But there is a problem. It must be soldered, or nickle plated, to make sure a good sliding contact is made at the external sleeve step. Without that conductivity guarantee, the oxide layer forms a bazooka-balun of the sleeve joint, or inserts a capacitor into the element.

 

[Guineafowl]

Avoid expensive new materials when prototyping. I use scrap aluminium sheet, cut into strips with a guillotine. For this design, I would use 12 mm strips that will pop rivet onto the boom. The folded dipole is then easy to fold.

Double the YO element length to get the actual dipole length to cut.
It is mentioned in the documentation, and comes from the legacy file formats of previous versions. The NEC software takes advantage of symmetry about the boom.

Thanks. It’s scrap copper pipe, old imperial size, so no great loss.

I did wonder if the values were halved in the YO software.

 

[Baluncore]

Thanks. It’s scrap copper pipe, old imperial size, so no great loss.

The scrap price of Al and Cu are similar per weight, but they have different densities. I would trade thicker Cu tube for an equal weight of Al sheet.
Old road signs are a good source of antenna element material when it is cut into strips, it is lighter and has lower windage.

Alternatively.
Use a pair of scissors to cut a helix from a used aluminium can. Mind your fingers. Unroll the helix to make a long strip. Bend the strip along the centre line to give it length-ways rigidity from the angle. Attach cut lengths of the strip to a wooden boom with rubber bands or staples.

Practice plating aluminium with solder through vegetable oil or turpentine, by scraping the oxidised surface with a blade until it is bright, then scratching with solder and a hot soldering iron. Once plated, you can solder it again without the oil.

 

[Averagesupernova]

Telescoping brass tube comes in 1/64" increments. But there is a problem. It must be soldered, or nickle plated, to make sure a good sliding contact is made at the external sleeve step. Without that conductivity guarantee, the oxide layer forms a bazooka-balun of the sleeve joint, or inserts a capacitor into the element.

The idea is not a permanent antenna with the sliding tubes. They're meant to be used to prototype antennas. Many different antennas can be experimented with this way using the same tubes. Once a person is happy with the results a more permanent construction technique is used. I've used this scheme with two vertical antennas in a time difference of arrival scheme. I adust the antennas to the exact frequency first, then go about tracking down the transmitter.

 

[Baluncore]

The idea is not a permanent antenna with the sliding tubes. They're meant to be used to prototype antennas.

Nickle plated brass is available, preassembled, and it costs less than telescoping brass tube.
Each six section unit, makes two or three adjustable directors for UHF.
5X Replacement 49cm 19.3" 6 Sections Telescopic Antenna
https://www.ebay.com.au/itm/394138218259?

 

[Guineafowl]

A quick update: I’ve built a prototype using the 15mm and 3/16” copper pipe that I have. I’ll come back with some details and VNA plots when I have some more time, and have built the balun as specified by the design software.
In the meantime, I’ve been reading through the ARRL chapters on the subject.

 

[Baluncore]

A quick update: I’ve built a prototype using the 15mm and 3/16” copper pipe that I have.

The software for Yagi design gives realistic and accurate dimensions for the array of passive elements and the far-field pattern.

Once the Yagi antenna array of dipoles has been designed, only the driven element length should be adjusted, to optimise the impedance matching. All other elements should remain in place and with the specified lengths, as any other strategy will rapidly lead to mayhem and confusion.

 

[Guineafowl]

The software for Yagi design gives realistic and accurate dimensions for the array of passive elements and the far-field pattern.

Once the Yagi antenna array of dipoles has been designed, only the driven element length should be adjusted, to optimise the impedance matching. All other elements should remain in place and with the specified lengths, as any other strategy will rapidly lead to mayhem and confusion.

Yes, it’s all built and screwed together using the surprisingly accurate import DRO on my milling machine. I’ve found some ally strip that might make a neater and more adjustable dipole.

 

[Guineafowl]

Right, some progress, but some confusion.

Standard copper pipe standoffs allow the elements to be raised into the plane of the dipole, and have a useful centre rib for alignment.

Balun cobbled together as per the design software.

I expect there will be some comments on this, and how the VNA is connected to the assembly. I can’t find guidance on how the $50\Omega$ SMA leads of the VNA should connect to the $75\Omega$ coax.

VNA plot, which looks reasonable but varies greatly when moving the balun around, and also depends on how the calibration is done.

 

[davenn]

. I can’t find guidance on how the 50Ω SMA leads of the VNA should connect to the 75Ω coax.

just use 50 Ohm coax, you shouldnt be using 75 Ohm anyway
you are using a 4 : 1 BALUN .... at most the feedpoint is only going to be around 200 Ohms,
not 300 Ohms
a folded dipoler on it's own has a 300 Ohm feedpoint, but once put into a yagi with other elements,
the feedpoint impedance is going to drop substantially

ohhh and of course, the 4:1 BAULN should be of 50 Ohm coax as well

 

[Guineafowl]

just use 50 Ohm coax, you shouldnt be using 75 Ohm anyway
you are using a 4 : 1 BALUN .... at most the feedpoint is only going to be around 200 Ohms,
not 300 Ohms
a folded dipoler on it's own has a 300 Ohm feedpoint, but once put into a yagi with other elements,
the feedpoint impedance is going to drop substantially

ohhh and of course, the 4:1 BAULN should be of 50 Ohm coax as well

The balun was designed by the software, based on that particular Yagi design and with RG6 specified as the cable. Maybe it’s a compromise, to get the impedances close rather than matching?

By using 50 ohm cable, am I not just pushing the problem down the line? The amplifier and TV socket are 75 ohm, so at some point I’ll need to make the transition.

I might be approaching this the wrong way, but my question would be: you have that crudely built antenna/balun in one hand, and a nanoVNA in the other. How do you connect them together in order to analyse and adjust the antenna/balun to a 75 ohm input impedance?

 

[Guineafowl]

To add to the above:
I’m going to have to admit that the ARRL book is a bit beyond me. It’s a vast collection of knowledge, written by lots of very clever, very technical people. It’s a reference book, not a teaching book.

For example, the first chapter, ‘Antenna Basics’, talks of SWR without first defining it. That sets the tone for the level of assumed knowledge.

In short, what I could do with, is the book you read before that one.

 

[Baluncore]

For example, the first chapter, ‘Antenna Basics’, talks of SWR without first defining it. That sets the tone for the level of assumed knowledge.

Only a few amateurs really understand SWR. The rest of them minimise it, because that is what it says to do in the operator's manual.
https://en.wikipedia.org/wiki/Standing_wave_ratio

 

[Guineafowl]

Only a few amateurs really understand SWR. The rest of them minimise it, because that is what it says to do in the operator's manual.
https://en.wikipedia.org/wiki/Standing_wave_ratio

Yes, it’s easy to enough to understand on a basic level, as the reading on the dial to minimise, but what I mean is, the initialism ‘SWR’ is not first expanded. In a chapter called ‘Basics’. Just one example, but that, and other things, are telling me I’ve skipped over the introductory course.

 

[Baluncore]

The ARRL Radio Amateur's Handbook, is probably preliminary reading for the ARRL Antenna Handbook. There is also the RSGB Radio Communications Handbook.

Amateurs will hear the term SWR while studying for their first licence exam.

It is hard to define an introductory course, because people can arrive from so many different directions. That is where Wikipedia now comes to the fore.

 

[Averagesupernova]

Maybe it's just me but I noticed 35 years ago the ARRL publications were a bit 'scattered' concerning skill level. In the same book they will describe something in very beginner's terms but the next chapter a different subject is way over the heads of many readers. Multiple books with overlapping subjects might get the newby educated if one sticks to ARRL publications.

 

[Guineafowl]

I take it I need a 75 - 50 ohm adapter, one end like this:

And the other a male SMA.

The antenna was put to use today - repairing a 433 MHz car key fob and testing the buttons. I rigged up the antenna to a simple TV signal meter (just LEDs from 50-80 dBuV) and fired the fob at it. Worked a treat.

 

[Averagesupernova]

I take it I need a 75 - 50 ohm adapter, one end like this:
View attachment 345558


And the other a male SMA.

The antenna was put to use today - repairing a 433 MHz car key fob and testing the buttons. I rigged up the antenna to a simple TV signal meter (just LEDs from 40-80 dBuV) and fired the fob at it. Worked a treat.

That is likely just a resistive pad to eliminate reflections on transmission lines. It is certainly not lossless or even close. My guess is 5.7 db of insertion loss.

 

[Guineafowl]

That is likely just a resistive pad to eliminate reflections on transmission lines. It is certainly not lossless or even close. My guess is 5.7 db of insertion loss.

As you can see, I’m a bit stuck. I’m not getting an answer to this question:

I might be approaching this the wrong way, but my question would be: you have that crudely built antenna/balun in one hand, and a nanoVNA in the other. How do you connect them together in order to analyse and adjust the antenna/balun to a 75 ohm input impedance?

Not sure if it’s the wrong question, or if there’s a tacit feeling that I should find out for myself. I would like to get the antenna up and running, but at the moment, the TV signal strength meter is not registering anything.

Another point made to me by a TV engineer on another forum, is that for digital TV, signal strength need only be above a certain threshold, after which point the bit error rate becomes the important factor. The BER is increased by reflections.