Jump to content
Sign in to follow this  
Master_Scythe

Smallest, most direction antenna

Recommended Posts

So I'm gonna nerd out here for a second, with literally 0 knowledge on this topic.
I don't want to reveal exactly what I need this for, as it's a bit of a business idea.

What I'm trying to figure out, is what sort of antenna I can build for an FM signal (80mhz~108mhz) for maximum range\gain, and minimum spread.
I need this device to be compact (which I know will be hard with such a low frequency), so while I know 1/4 length antennas are good, can I continue to divide? 1/8th antenna?

By accurate, I'm talking something like a 10 degree beam.
Considering how accurate I'd like, and how the beam should cone-shape bigger, the further I am away (yes?) even a tighter beam would probably be fine.

 

Does anyone have some reading or quick advice for me? Directional small high gain antenna, is the OPPOSITE of what most people want in the FM world, so finding 'quick' resources, without studying a lot of antenna theory online is tricky.

Share this post


Link to post
Share on other sites

 

I wish I had better advice for you but antenna theory is hard - I have a telecommunications degree and we touched on it literally for one exam question in one topic on a subject about digital transmission.

Antenna theory is a fairly well-established field so few people study it any more, as the people who know about it are enough to supply the industry with expertise.  If you're setting up a transceiver system you usually just assemble it from off-the-shelf parts from dedicated vendors.

Having said that, you can keep multiplying or dividing your antenna length as much as you want - I just don't know how, or how quickly, that affects the transmission/reception.

Speaking of that, do you need the antenna to transmit, receive or both?

 

I'd try to get better advice from someone like these guys:  https://www.solidsignal.com/fm-directional-antenna-th.asp

Share this post


Link to post
Share on other sites

Been a while since I've had to do antenna design, forgotten a lot of the nasty fiddly bits. Microwave is a damned sight harder and I've tried to neuralise myself on THAT shit...

Anyway - directional antenna would usually be a Yagi. You could also have a look at a dish, unless you really want to go nuts and look at a phased array, but they will really drop you into rocket science/Propeller Head Mode.

And only a 10 degree beam? You'll need a bloody long Yagi to get that, long as in number of passive elements. Parabolic dish would need some serious fine-tuning/adjustment to get that as well.

There's also antenna mounting "stability" with such a narrow beam to consider as well - once you get a low-level signal the slightest shimmer/shake of the antenna and it'll be pointing off target.

Then there's the good old mucking about with the coax cabling and appropriate connectors as well. Thinner the diameter of the coax, more lossy it can be at your desired freq. Not by much, but when you start talking maximum gain vs minimal spread then that can rear it's ugly head into the equation.

Aaaaaand a few Obligatory Dumb questions:

 - as most of your frequency range there is across the FM Free-to-Air audio band, presumably you're only going to be receiving and not transmitting, right? Right?

 -  what are your engineering/metalworking skills like to do DIY?

 - indoor or outdoor operating environments?

 - what's your definition of "compact", because you'd better have a look at commercial antennas and compare them to your "compact" requirements. Last time I looked at a Yagi with that sort of gain it had something like 15 passive elements on it, and on the 470-odd MHz band it was near 2 metres long. Not exaclty pocket-sized nor... 'compact"...

It was also chronically illegal for anything remotely resembling Free-to-Air transmission. Didn't stop the Customer though, until he got sprung by the authorities over here for splattering and drowning out his local phone tower...

Reckon you should be sidling up to Uncle Google and Cousin youTube for the basics if you're thinking of doing a home-rolled one.

Else Uncle Google et al to see which manufacturer has got something off-the-shelf that'll might fit into your "business" requirements there.

Edited by merlin13
  • Like 1

Share this post


Link to post
Share on other sites
11 hours ago, merlin13 said:

- as most of your frequency range there is across the FM Free-to-Air audio band, presumably you're only going to be receiving and not transmitting, right? Right?

 - what are your engineering/metalworking skills like to do DIY?

 - indoor or outdoor operating environments?

In Order;

Yes, but I've made this a little easier. I'm now going to use 433mhz (as found in car key fobs), so the antenna will be MUCH smaller.

Metalworking skills are great.

The antenna won't be fixed, so outdoor use, but no need for element exposure.

 

Quote

And only a 10 degree beam? You'll need a bloody long Yagi to get that, long as in number of passive elements. Parabolic dish would need some serious fine-tuning/adjustment to get that as well.

There's also antenna mounting "stability" with such a narrow beam to consider as well - once you get a low-level signal the slightest shimmer/shake of the antenna and it'll be pointing off target.

It'll be handheld, I want to use it for short range locating.
I want the user to be able to use it at 300m and get a 'general direction' but as they get closer, need to be more and more accurate.

 

11 hours ago, merlin13 said:

what's your definition of "compact", because you'd better have a look at commercial antennas and compare them to your "compact" requirements. Last time I looked at a Yagi with that sort of gain it had something like 15 passive elements on it, and on the 470-odd MHz band it was near 2 metres long. Not exaclty pocket-sized nor... 'compact"...

Yeah, to be fair I'm going to be better off 'blocking' the signal to gain directional reception. Gain is nice, but 'accuracy' is better.
What I'm essentially trying to build is the same technology as an animal tracker, but smaller, and only designed to operate over 300~500m, rather than kilometers.

It's REALLY hard to research this, because people talk about 'ideal antennas' in every situation, and I don't have ANY antenna theory background (outside of matching half wave dipoles to frequencies and impedance).

 

Really, I'm looking for the most pocket-able antenna (say... below 20cm...ish length?) with the most gain possible, and the most narrow beam width, and I'll attain that however I need to.
It doesn't need to be correct it just needs to work.

If the chinese sweat shop can make a coin-sized FM radio work, I can make this work....

Im going to look into how small i can make a Parabolic dish\Grid antenna.... I bet it's too big....

So far, my closest bet seems to be a Horizontal Dipole, that has been bent, and then encased in a VERY narrow  opening reflector.
https://www.qsl.net/kk4obi/Center-fed L-dipoles Vertical.html

I'm only trying to receive a 'buzz' from a transmitter, the signal doesn't need to be clean.

 

Perhaps a 90 degree bent, center fed dipole is my answer?

Apparently if I go narrower than 90 degree wierd things happen... but I'm yet to figure out what that is 😛

Lateral%20Equal%20Arm%20L-dipole%203D.pn

 

 

 

EDIT:

 

I see what they mean by 'below 45* the tips become part of the dipole"

3D%2045%20ang%20up-down%20pattern%20comp

OK, so a 90 degree dipole might have to do, and might be the best I can achieve....

 

 

the quest continues.

I wonder how I'm best to shield the dipole to make sure only a 'beam' gets into the receiver antenna...

Edited by Master_Scythe

Share this post


Link to post
Share on other sites
40 minutes ago, SquallStrife said:

How about a helical antenna?

https://www.tutorialspoint.com/antenna_theory/antenna_theory_helical.htm

I actually built a pair of these in the early 00's to set up a long-ish range (3km) wifi link. 

I did consider this, but you lose 3db for circular polarised vs linear polarised.

And also, because due to the coil spacing on a low mhz, it ends up being bloody long.

Share this post


Link to post
Share on other sites

 

Man, where do you guys get your antenna knowledge from?  I learned almost nothing about this stuff AT UNI.

Unis generally tailor their courses specifically for what industry requires, but I was and am still a little annoyed at missing this stuff.

 

Share this post


Link to post
Share on other sites
14 hours ago, Master_Scythe said:

I did consider this, but you lose 3db for circular polarised vs linear polarised.

And also, because due to the coil spacing on a low mhz, it ends up being bloody long.

The narrow beam width should make up for that and then some.

It'd also mean your quad could land in any orientation and still receive a signal, right?

You're right that the bulk would be significant though, I just didn't realise HOW bulky it would be.

At 433MHz, λ=~70cm, so a meagre 12-turn antenna would be over 2m long. Makes a lot more sense for microwave.

Share this post


Link to post
Share on other sites
9 hours ago, Kothos said:

Man, where do you guys get your antenna knowledge from?  I learned almost nothing about this stuff AT UNI.

I've spent about 3 hours on google..... 😛

  • Haha 1

Share this post


Link to post
Share on other sites
On 9/21/2018 at 9:59 AM, Master_Scythe said:

It'll be handheld, I want to use it for short range locating.
I want the user to be able to use it at 300m and get a 'general direction' but as they get closer, need to be more and more accurate.

Yeah, to be fair I'm going to be better off 'blocking' the signal to gain directional reception. Gain is nice, but 'accuracy' is better.
What I'm essentially trying to build is the same technology as an animal tracker, but smaller, and only designed to operate over 300~500m, rather than kilometers.

It's REALLY hard to research this, because people talk about 'ideal antennas' in every situation, and I don't have ANY antenna theory background (outside of matching half wave dipoles to frequencies and impedance).

...

Really, I'm looking for the most pocket-able antenna (say... below 20cm...ish length?) with the most gain possible, and the most narrow beam width, and I'll attain that however I need to.
It doesn't need to be correct it just needs to work.

Sounds like the requirement is pretty close to a ARDF/radiosport configuration, on 70cm/UHF.

Don't be thinking that a narrow beam is always what you want.

Narrow Beam antennas are rarely a nice clean pattern. Expect several side lobes, and some of them are quite distinctive - so while they might be 10dB below the main lobe, they are still 10+db above the nulls on either side of them. Assuming you are manually sweeping the antenna around, you might get a ...oo...oOo...oOOo...oOo...oo... sort of result as you sweep past.   Very easy to end up chasing a lobe off the boresight even worse if you are in a location with lots of reflections.

A lot of people use a relatively low gain, one wider smooth lobe at the front, and enough front -back ratio to be able to tell the difference. Then you sweep the single lobe and listen/watch for the peak gradual rise then fall, and you can get the peak within +/- 10 degrees, which is good. An alternative is a loop, that has a very narrow null, and then you steer for the sileence.

In terms of 'pocketable' does it need to be that size in use, or in transport. There are several designs for collapsible antennas - and cutting up a cheap tape measure for metal elements that roll up - and flex/bend when you bang into things is a classic design for ARDF antennas.

Also be aware with dishes - you end up with a beam in azimuth and elevation - doubles the complexity, because you can't hunt in one dimension at a time, you have to always hunt polar-2D.

Share this post


Link to post
Share on other sites
45 minutes ago, stadl said:

Sounds like the requirement is pretty close to a ARDF/radiosport configuration, on 70cm/UHF.

Don't be thinking that a narrow beam is always what you want.

Narrow Beam antennas are rarely a nice clean pattern. Expect several side lobes, and some of them are quite distinctive - so while they might be 10dB below the main lobe, they are still 10+db above the nulls on either side of them. Assuming you are manually sweeping the antenna around, you might get a ...oo...oOo...oOOo...oOo...oo... sort of result as you sweep past.   Very easy to end up chasing a lobe off the boresight even worse if you are in a location with lots of reflections. 

Thank you makes a lot of sense.

 

On this though, isn't that why a 45* or less L dipole would be the best 'small' design?

The lobe looks VERY clean, just need to block or reflect the back one?

Share this post


Link to post
Share on other sites

The L dipole won't really give you enough directional to be useful for DFing. They are more of interest when you want to reduce height/ground real-estate in n a HF antenna when the lengths are 10s of metres. You get a longer resonant length in a shorter space - lose some omni, but at least gain a little by pointing the small gain in a useful direction, and the weak end where there's nobody you want to communicate with.

The shielding/aperture concept not so easy in reality - unless you are building a dish.

But in the case of a simple 2 element beam (e.g. yagi) antenna you have the reflector at the back which does give some front-back, while being about 35 wide, 20cm long. A 3 element beam, would be adding one director element in front of the driven element, and up the gain/push the pattern a bit more to the front, and double the length - with the right construction either option can be very portable.

Having a significant back lobe is not the end of the world. If you stand in one spot and spin 360, you'll get two directions, move a few times , and repeat, and quickly you find the geometry only works for one of them, as one way direction your Line of Bearings will cross, while the otherway they won't 🙂

 

Share this post


Link to post
Share on other sites
On 9/21/2018 at 10:29 AM, Master_Scythe said:

Yeah, to be fair I'm going to be better off 'blocking' the signal to gain directional reception. Gain is nice, but 'accuracy' is better.

...

I'm only trying to receive a 'buzz' from a transmitter, the signal doesn't need to be clean.

 

hmm.  i am curious how well blocking the signal from other directions would work.  like, for example, putting the receiving antenna behind a hole in aluminium.  but i imagine reflections would be an issue.  maybe a narrow tunnel through a series of thin aluminium wafers ???

 

as for the signal — i am assuming this is omni — but will you be trying to detect the presence of the carrier only?  if there is no 'signal' per se, periodic on-off keying might still be a good idea.  unless youre in the middle of nowhere, there could be a lot of 433mhz noise about. 

Edited by @~thehung

Share this post


Link to post
Share on other sites

How about beam forming techniques like some routers use?

Multiple cheap-arsed directional antennas with signal metering.  With a bit of maths you can work out where the object signal is coming from.

But whether multiple little ones with mediocre accuracy end up as being better than one or two big ones with OK to good accuracy?

  • Like 1

Share this post


Link to post
Share on other sites

Actually I shouldn't credit it to routers.  More like early heatseeking missiles and even then it was probably an old idea.
4 sensors representing 90 degrees of a circle each, which then control vanes which dictate the direction it flies based on where the strongest heat source is being sensed.

Share this post


Link to post
Share on other sites
10 hours ago, @~thehung said:

hmm.  i am curious how well blocking the signal from other directions would work.  like, for example, putting the receiving antenna behind a hole in aluminium.  but i imagine reflections would be an issue.  maybe a narrow tunnel through a series of thin aluminium wafers ???

Not practical for the sizing he wants unfortunately.

This sort of concept works like a scaled up version of a pinhole camera, or a spotlight with baffles. But consider the scale difference between the wavelengths  - Light is ~700 nm, while UHF (433MHz) is ~700mm. UHF is on a scale 1,000,000 times larger,. You can build a box/baffle, but it will be km across, and directional on a solar system scale.

 

Share this post


Link to post
Share on other sites

...Man, where do you guys get your antenna knowledge from?  I learned almost nothing about this stuff AT UNI... Unis generally tailor their courses specifically for what industry requires, but I was and am still a little annoyed at missing this stuff.

...I've spent about 3 hours on google..... 😛

Pffft - Uni. For this Little Black duck, couple of semesters doing the old E&C Cert whilst dicking about with military comms and other nice stuff for Dept of Defence back in the early '80s. Few yeras on 'n off doing TV and various radio repairs/installs for various frequencies and Tx/Rx types, then then over a decade doing industrial radio control and monitoring this century...

Anyway, to get around using anything with sidelobes you'd need to think about playing with a meter or some sort of signal level display, especially if you want to do any type of "Distance To Target" setup based on detected RF levels.

And of course RF is pretty insidious on reflecting off "structure" to boot. So for serious RF location you might need to think triangulation when you're looking at tracking down a point source - take a directional fix, wander around some more, take another fix.

Then the higher the freq the worse the ambient power absorption, especially combined with what I'd assume to be a pretty low-power RF source to start with. Then stick a lump of "something" between you and the target and your detectable RF levels can/will plummet as well.

Don't get me wrong, it's do-able. But you will have the hurdle(s) of Canna Change The Laws Of Physics to work against. Just going to be somewhat... fiddly...

So I'll ask the obligatory Dumb Question - just WTF are you aiming (heh) to do with this?

Share this post


Link to post
Share on other sites
11 hours ago, stadl said:

Not practical for the sizing he wants unfortunately.

This sort of concept works like a scaled up version of a pinhole camera, or a spotlight with baffles. But consider the scale difference between the wavelengths  - Light is ~700 nm, while UHF (433MHz) is ~700mm. UHF is on a scale 1,000,000 times larger,. You can build a box/baffle, but it will be km across, and directional on a solar system scale.

i guess my mental picture of how materials block RF is seriously flawed.  could you elaborate though?

just to be clear, i was envisioning many wafers stacked perpendicular to the tunnel direction, with a hole in each wafer (and all else encased).  the idea being that the only path in is through all the holes, meaning that most off-axis paths would be presented with a flat piece of aluminium and reflected/absorbed between layers to nowhere.  lets say this thing is 1m long. 

i was concerned about reflections at the rim of each hole, hence the thinness of the wafers and many layers.  but are you are saying 70cm is a big enough wavelength to enter from, say, 90º (or 180º?) and diffract everywhere within the enclosure like crazy?

 

Edited by @~thehung

Share this post


Link to post
Share on other sites
12 hours ago, @~thehung said:

i guess my mental picture of how materials block RF is seriously flawed.  could you elaborate though?

just to be clear, i was envisioning many wafers stacked perpendicular to the tunnel direction, with a hole in each wafer (and all else encased).  the idea being that the only path in is through all the holes, meaning that most off-axis paths would be presented with a flat piece of aluminium and reflected/absorbed between layers to nowhere.  lets say this thing is 1m long. 

i was concerned about reflections at the rim of each hole, hence the thinness of the wafers and many layers.  but are you are saying 70cm is a big enough wavelength to enter from, say, 90º (or 180º?) and diffract everywhere within the enclosure like crazy?

 

Yes it sort of works that way, but remember the scale, and the need to fully enclose the receive antenna except for the aperture - otherwise, wile you will have a small peak surrounded by a big null for maybe 30 degrees either side, the back side, that sheet will act as a huge reflector and give the antenna pattern a fat arse for 180 degrees at the back

In terms of scale for the aperture - The cutoff frequency for a rectangular aperture is where the wavelength is less than double the longest side. So to pass a 70cm signal, the aperture needs to be no smaller than 35cm That gives you a view of the scale you are building - a pinhole camera style setup with a 35cm+ pinhole.

Then you have diffraction and reflections, as you have realised - several nested boxes - or at least stacked aperture plates could prevent this with this - but then you're probably just better using a tube as the aperture - but size will need to be probably a wavelength long. You usually then need conductive stuff far enough away from the antenna that it does not form part of the antenna - several wavelengths would be nice.

I'll confess I'm reaching the limits of my current knowledge and understanding, on this topic. I understand the basics, but have forgotten the details - it's been 25 years since I studied, and antenna design and RF black magic haven't been much of my dabbling in amateur radio since.

Share this post


Link to post
Share on other sites
22 hours ago, stadl said:

that sheet will act as a huge reflector and give the antenna pattern a fat arse for 180 degrees at the back

Couldn't you follow the antenna's shape with the block and reflect the rear lobe out the same way?

I mean, this is a RECEIVING antenna, so its probably easier to block in incoming signal than an outgoing on right?

 

So, say, my antenna is an L shape, can't I put a bigger L (by only mm's) off the rear of the first (not touching, of course).  So    |_ L  Though do I need to actually pull the 'shield' to ground? Or is it just nor worth worrying about the rear lobe of an L dipole?

Share this post


Link to post
Share on other sites

A single element won't act as a reflector unless it is placed at the correct spacing. A few mm away won't act as a reflector, it will just act as a large capacitive element and mess up the antenna. Placed at the right distance from the main element it will have the correct parasitic resonance and cancel out signal in a particular area, and reinforce it in other directions - that is the fundamental design of a 2 element beam/Yagi antenna.

And yes, the antenna elements are not required to be straight - so you could have L shaped elements - but you would need to model it to determine the variations to length and spacing, as it won't be exactly the same. If you want to take that to extreme, look at the Hexbeam/Spiderbeam antennas that (on larger wavelengths typically 10-40m)  bend the elements one or more times in Ls, Ws and Cs to allow longer elements to fit in a smaller space - but they are not simple.

 

 

 

Share this post


Link to post
Share on other sites
1 hour ago, stadl said:

And yes, the antenna elements are not required to be straight - so you could have L shaped elements - but you would need to model it to determine the variations to length and spacing, as it won't be exactly the same.

image.png.1dff1df303c98cc8dab22862adcb0e2a.png

Isn't that what this model i posted earlier already covers?

 

So long as it's less than a 45 degree L, you get a nice direction lobe out the center?

Share this post


Link to post
Share on other sites
2 hours ago, Master_Scythe said:

Isn't that what this model i posted earlier already covers?

So long as it's less than a 45 degree L, you get a nice direction lobe out the center?

That's a 90 degree model - has negligible applicability to a 45 degree model. 6dB is not bad, but it's also identical front and back, which is not ideal for DF

You could DF with it - but I still don't know why you are obsessed with an L when a Yagi would probably be better. Less gain, and some front-back ratio is more useful.

 

 

 

 

Share this post


Link to post
Share on other sites
2 minutes ago, stadl said:

That's a 90 degree model - has negligible applicability to a 45 degree model. 6dB is not bad, but it's also identical front and back, which is not ideal for DF

You could DF with it - but I still don't know why you are obsessed with an L when a Yagi would probably be better. Less gain, and some front-back ratio is more useful.

 

Mainly due to simplicity and size.

How small and simple could I manage a Yagi build?
You seem to know what you're on about 😉

Share this post


Link to post
Share on other sites

Longest element will be half a wavelength - so approx 35cm, and element spacing will be approx 20cm. So a 2 element is 35x20ish, and a 3 element is 35x40ish - + extera boom length of you want a handle 🙂

I have seen some nice designs using a rudimentary pistol grip under the boom, which reduces the boom grip and can be prefereable for some people holding, while others just extend the boom 10-20cm behind the reflector to grip it that way. - Lots of personal choice on construction. Some of the simplest designs use tape measures (cheap flat spring steel) and a bit of broom handle or PVC pipe/conduit

There are a several designs around - googling for '70cm 2 element beam' or '433Mhz 3 element yagi" or even "UHF direction finding antenna" or combinations of whether you want to work on 'beam vs yagi' frequency vs wavelength etc.

http://www.csgnetwork.com/antennae3ycalc.html

While the theory can calculate element lengths and spacing, the element material and thickness , what the boom is made of (conductive or non-conductive) etc, all vary the system, and lots of people have through trial and error created common designs that work, and some are sold as kits, but many are published with design notes and pictures for you to build at home.

The good thing is that the Amateur radio 70cm band covers 430-440MHz (with extensions to down to 420 or up to 450 in some parts of the world). The 433MHz ISM band commonly used by garage door openers and other low power devices is in that band that all have to work or not work on their own merits, in a generally unlicenced maship. But , the band overlap, means that an amateur radio design for 70cm beam should work for 433 ISM band devices without modification 🙂

With most single band designs you can elect to delete director elements from the front if you need to modify the design. It WILL reduce the gain/widen the lobe, and it might impact other characteristics, but it should still be resonant - just don't push it too far. Chopping a 4 element design to 3 is probably manageable, chopping a 13 element design back to 3 is asking for problems. But it's worth a try if you find a design. Also be wary deleting elements for dual-band antennas (e.g. 2m/70cm) , as you may not know which elements are shared between bands, or have some complex interaction.

https://en.wikipedia.org/wiki/Yagi–Uda_antenna

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×