Dipole antenna, BALUN and connection to Raspberry Pi

[PiCrust]

Hi,

I wonder if anyone can help.

I am trying to learn about about FM transmitting with the aid of my Raspberry pi.
I currenty have my Raspberry Pi set up and it modulates GPIO4 and produces
an FM signal which I can listen to around my house, using a 20mm long jumper
cable on GPIO4.

Set up by following this tutorial:
https://makezine.com/projects/make-38-cameras-and-av/raspberry-pirate-radio/

I now would like to understand the theory of how to attached a tuned Dipole Antenna
to the Raspberry Pi, and if you would require the use of a BALUN to balance the feed
if using a COAX CABLE.

I have sketch out in the image below how I think the set up would be. This is just
how I assume it would need to be connected and as I have no experience in this field
it may therefore be incorrect.

Also, due to the way the Raspberry Pi modulates the signal and produces harmonics
I will not be creating a working set up, but am more interested on how the theory
of how to get it to work so as to further my knowlege and get a fuller understanding
of how to transmit an FM signal.

A couple of questions that I have relating to how this would be set up are:

1.
Does the downward pointing leg of the Dipole need to be connected to Ground,
if so why does this need to be connected to ground?

If this is the case would it be ok to connect it to the Ground pin on the Raspberry Pi via the
braided outer wire of a Coax Cable.

2.
Would I require a BALUN to balance the feed from the Coax?
If so how which type would be best to use.

I have seen information on using a length of CoaxCable the same length as
1/2 wavelength to balance the feed, but this was shown on a 'Folded' Dipole
(see image below), and not on a standard 'non folded' diplole.

Would it still be possible to do this for a standard Dipole?

If so how would you connect the Coax Cable and the BALUN to the Dipole.

Would it be the same as in the image above where the middle core wire of
the CoaxCable is connected to the top Leg of the Dipole and then the BALUN
length of CoaxCable is soldered to this leg also, with its other end connected to
the bottom leg of the Dipole and all the three braided ends are connected together
(with none of the braid attaching to either the top or bottom leg).

If that is the case, how does it ground? as I assumed that the bottom leg would
require to be attached to a ground.

Any help / explanation would be appreciated.

 

[MikeMl]

Dipole with balun:

More symmetric antenna pattern, closer to free-space, ideal pattern, with deep nulls. Polarization matches plane of dipole. Little or no radiation from the feedline. Should be a 70Ω feedline, 1:1 balun. Balun can be coax or ferrite transformer or ferrite sleeve. Big nulls off the ends of the dipole, so sometimes this works against you. Polarization of receiving antenna must match transmitting antenna, sometimes works against you.

Dipole without balun:

Closer to isotropic pattern, with nulls filled in. Feedline radiates, becomes part of the antenna. Antenna radiates a cross polarized field, so cross-polarization loss is less likely. Less pronounced nulls. Can get strong fields at the transmitter end of the feedline, so that can create RF feedback in equipment (high power problem, ain't gonna happen at a couple of W).

Bottom line: I wouldn't bother with a balun, especially at your power levels. The power lost in a transformer balun (including a coax transformer ~10%) usually isn't worth the hassle... If you experience some coupled RF issues at the transmitter, make a coiled-coax RF feedline choke, or put several Ferrite beads on the feedline near the antenna. This prevents RF current flowing on the outside of the feedline...

I deal with this issue at power levels of up to 1500W, where having a balun to keep RF out of the shack makes a difference.

 

[JimB]

First of all, welcome to ETO.

Your queries:

Also, due to the way the Raspberry Pi modulates the signal and produces harmonics
I will not be creating a working set up,

I am glad to hear that, there is already far too much junk poluting the RF spectrum.
Your RF neighbours thank you.

1. Does the downward pointing leg of the Dipole need to be connected to Ground,
if so why does this need to be connected to ground?

No, it does not need to be connected to "ground".
We could spend a lot of time discussing what constitutes "ground".

2. Would I require a BALUN to balance the feed from the Coax?

No.
A BALUN is a BALanced to UNbalanced transformer.
A dipole antenna is inherently balanced, coax cable is unbalanced. In an ideal world you connect the two together using a BALUN, otherwise you get RF currents flowing on the outside surface of the coax, this can cause problems in some systems.
For trivial applications like this it is not necessary.

I have seen information on using a length of CoaxCable the same length as
1/2 wavelength to balance the feed, but this was shown on a 'Folded' Dipole,
and not on a standard 'non folded' diplole.
Would it still be possible to do this for a standard Dipole?

Yes.

If so how would you connect the Coax Cable and the BALUN to the Dipole.

Exactly like the picture, at the centre of the dipole. The only difference is that the dipole is two conductors, each being a quarter wavelength long, and the folded dipole is effectively a full wavelength loop.

If that is the case, how does it ground? as I assumed that the bottom leg would
require to be attached to a ground.

When a dipole is vertically polarised it is conventional to put the "earthy" end downwards. in reality it probably does not make much difference, although I have no hard evidence to prove this.

An antenna does not have to be earthed (or grounded to use USA terminology), but there needs to be two parts to an antenna so that a current can flow.
Consider the simple dipole, two parts, and RF current flows through space between the parts.
Consider a monopole antenna, there are two parts to that also, although is may not be obvious.
The non-obvious part is the chassis/body of the equipment it is connected to.

Does this help?

JimB

 

[MikeMl]

If you want a vertically polarized antenna, rather than a horizontally polarized dipole, then the usual way of making it is called a ground-plane vertical antenna, where the lower 1/4λ element is replaced with ~four radials, sometimes co-planar, sometimes bent down at a ~45degree angle.

Note that the down-bent g. p. exhibits a feed-point impedance closer to 50Ω, so is usually used with 50Ω coax, unlike a dipole, which is suitable for CATV coax.

 

[PiCrust]

Hi,

Thank you for all your informative replies, which have definitely helped.

Jim,
Just to clarify what you stated below about not 'earthing / grounding', the Dipole.

An antenna does not have to be earthed (or grounded
to use USA terminology), but there needs to be two parts to an antenna so that a current can flow.
Consider the simple dipole, two parts, and RF current flows through space between the parts.

How would you then connect the feed coming down the center wire of the Coax cable, from the
Raspberry Pi's GPIO4 to the Dipole? ( For the first part I guess you would need to solder the jumper
wire that is connected to the RasPi's GPIO4 pin to the centre wire of the Coax Cable), but then how should
you correctly connect the Coax Cable to the Dipole legs.

I have show three options below, can you confirm which one would be the correct way
to connect, if none of them are correct could you explain the correct way to do it and the
reasons behind it.


1.
Dipole with no connection to Rasberry Pi's ground pin
but braided outer wire of the Coax Cable connected to
bottom leg of Dipole. (as shown below)

2.
Dipole with no connection to the bottom leg of the Dipole.
So just the centre wire of the Coax Cable connects to the
top leg of the Dipole. (as shown below)

3.
Dipole with centre wire of the Coax Cable connecting to
both the top and bottom leg. (as shown below)

 

[MikeMl]

None of those. Only correct way is what you showed in your Post #1.

How do you generate RF at a GPIO pin 7 on the Pi?

 

[JimB]

It really offends my sensibilities to be talking about connecting an antenna to a microcontroller GPIO pin!

To connect a dipole antenna to a real radio transmitter:
At the antenna, one leg of the dipole will be connected to the centre core of the coax, and the other leg connects to the screening braid of the coax.
At the transmitter, the centre core of the coax will connect to the "live" antenna terminal, the coax braid will connect to the chassis/0volt line of the transmitter.

There are of course variations of this, depending on how the output stages and filtering of the transmitter are configured.

JimB

 

[JimB]

How do you generate RF at a GPIO pin 7 on the Pi?

Having skimmed the original article, you just vomit harmonically related crap all over the radio spectrum.

JimB

 

[MikeMl]

Having skimmed the original article, you just vomit harmonically related crap all over the radio spectrum.

JimB

In that case, I would be for putting a bandpass filter between the GPIO pin and the antenna. I'm glad I dont live next door to the Axxhole who advocates that circuit. (In case you cant figure it out, substitute an s for the x in the expression).

To the OP/TS: you want to hook that RFI generator to an antenna???

 

[PiCrust]

Hi Jim & MikeMI

Thanks again for your comments/info.


Jim,

I did not mean to offend you, I was merely wanting to
learn and try to understand how a Dipole works along with
it's connections.

I do not have access to an actual transmitter to see how that works.
So was trying to see how it would, or if it could work with a Raspberry pi,
as I can physically see it's connections in the flesh.

As I know the Raspberry Pi has the problem with harmonics
I would never physically try to connect it up to a Dipole,
[The below video shows the harmonic problems]

it was merely to try and get to grips with how the feed would
work with dipole to transmit. I have seen that others have
tried to address this problem by creating a Band Pass Filter
that would eliminate the harmonic's.
[The below video shows the Band Pass Filter]

I've watched the following youtube video again, which shows the basics
of how a Dipole works.

And that shows a '+' and a '-' coming in as a feed into the centre of the
Dipole. With that in mind and also what you have stated in you second to
last post, I assume that the '+' = live feed (centre wire of Coax cable)
with the '-' = chassis/0volt line (braid wires of Coax cable).

 

[MrAl]

Hello there,

First, the harmonics, as you might know, are related to the fundamental frequency and with a square wave the amplitude reduction is 1/n, where n is the harmonic number. So if you generate a 90kHz square wave at the 999th harmonic the amplitude is down by 1/999, which is close to 1/1000 or 0.001 of the amplitude of the square wave which is probably 5v. But that's only if the rising edge is as sharp as a 90MHz sine wave, which it probably isnt. That means more loss or maybe very little signal at all. So if you are going for a high harmonic then it would be best to use a high frequency device that can generate a nice clean square wave, as a buffer. I had some pretty good luck with TTL back in the early 1970's, and got a range of about 1/2 mile with a simple antenna, about two stories high.

A bit later than that time, i also built a CB antenna made from copper pipe. Copper pipe makes a really good antenna. With the copper pipe i was easily able to get several towns over where with the regular type CB ground plane antenna i was never able to do that, and that was even without a matching device and only 5 watts of transmitting power (the legal limit back then, not sure what it is now).
It was a dipole with about 8 feed of pipe on one side and 8 feet on the other, and a hefty fiberglass diamond shape in the center, with the pipes bolted to the fiberglass. The feed line came out at a right angle to the pipes, and i used heavy duty fishing line for guy wires. It survived many years like that.

With a dipole you feed the center and that's it. The ground of the coax (the shield) connects to the lower pipe, the center conductor to the upper pipe. If you cut to length it will be slightly inefficient as there are fudge factors for the diameter of the antenna arm too, but im sure you can find this info on the web.
As i said, copper pipe is good because you can get it at any hardware store, and it is a good conductor and holds up to the weather too.

 

[JimB]

Jim,
I did not mean to offend you, I was merely wanting to
learn and try to understand how a Dipole works along with
it's connections.

Do not worry, it is not you who offends me, it is the concept of using a digital I/O port of a microcontroller to generate RF, as is well demonstrated by the the first video which you linked. Good information.

However, the second video where the guy builds a "bandpass filter", that was nothing short of a joke.
There was not much filtering going on there.

As for Mr Antenna Theory in the third video, there were one or two oddities in the poorly presented piece.
One thing that sticks in my mind was where he was talking about the "directivity", from the numbers he was quoting I think he really means "Gain relative to an isotropic radiator".
Directivity is something different altogether.

So all in all, I am not impressed.

Having said (ranted) that, if you want to know about radio, just ask.

JimB

 

[MrAl]

Hi there Jim,

That's one thing i wanted to get a little more into again as i havent done anything in radio for years now, and i really do mean 'years' ha ha.

One of the problems with the receiver i remember was keeping on the right frequency using simpler components. Maybe they do it differently these days.

But i also forgot to mention the bandpass filter. These are too easy in theory, but in practice we'd want to keep it tuned well too so that we dont loose effective radiating power. I suppose air core coils should be good enough, but maybe you know better ways.

So for this thread anyway, i would think that generating a given frequency much lower than desired, then using a bandpass filter tuned to a harmonic N times higher than the generated frequency, where N is an odd number, should work ok. Generating the highest possible frequency is probably a good idea too to keep the natural reduction in power as minimal as possible, and using high speed TTL would give sharp rise and fall times. Something like in the F family should work good, even a simple nand gate or inverter.
I cant remember what i used in my transmitter back in the 1970's. Probably TTL that was available back then.

 

[JimB]

Hi there MrAl,

In days gone by it was usual to have an oscillator running at a relatively low frequency and then multiply up to the desired transmit frequency.
If we consider a transmitter giving 144MHz output, the primary crystal controlled oscillator may be on 8 or even 4MHz with a total of x18 or x36 multiplication, usually with basic filtering at each multiplier stage and then a good low pass filter to remove the harmonics from the output after the power amplifier.

Nowadays a synthesizer would be used to control an oscillator running at 144MHz and then stages of amplification to give the required output power, and then a good LPF to remove the harmonics.

For VHF use, air cored coils are still a good choice if you are running more than a few milliwatts.

At HF (below 30MHz) it is more usual to use toroidal iron core inductors, big chunky ones when the power gets above 10 watts or so.

The other thing about synthesisers is that where you want something which is tunable across a range of frequencies, you get excellent stability because the thing is referenced to a crystal, and you dont have the mechanical and electrical headaches of designing a free-running LC oscillator with high stability.

JimB