Dipole Antenna Length

[Fried Chicken]

I am getting mixed reports on the optimum length of a dipole antenna.

My understanding is the two dipoles consist of 1. the ground/shielding, 2. the other/positive.
I've also heard of a rule, 468/frequency in MHz to give an approximation of the length of the wavelength in feet.  I've heard you further divide that by two to get each element length: each element is then 1/2 the wavelength.

Now I found a resource suggesting the length of the total dipole should be half the wavelength, each element a quarter of the wavelength:



For my reference frequency let's do 117MHz, so 468/117MHz gives a wavelength of 4 feet.

If I follow the Antenna Dipole calculator in SDRAngel I get the following:


So each element is 2 feet, or half the wavelength.  If I follow the youtube video's recommendation, each element of the dipole should be 1 foot, or 1/4 the wavelength.

I'm at a loss and am on the verge of breaking out the university physics textbook.  What works best here?

[A.Z.]

in meters it's easier

Lambda=300/MHz

and then apply a correction factor to deal with the conductor speed and the end effect, assuming copper you will then have

Lambda=Lambda*0.965

at this point, for a dipole your total length will be

TotLen=Lambda/2

and each arm length will then be

ArmLen=TotLen/2

all the above in ideal conditions, you will usually cut the arms longer and then, trim them equally to resonance

[RFDx]

Stick to the calculator in SDRangel.
Length (not wavelength) of a lambda/2 dipole antenna in feet is ~468/frequency, with frequency in MHz. For 117MHz both arms of the dipole would have 2 feet, 4 feet in total for the whole antenna.

[A.Z.]

is it better learning the calcs or using a calc w/o knowing how it works ?

Oh and then Lambda=Wavelength in case you weren''t aware; what's next the dipole gain ? Well, at 0.6 Lambda height it's more than 7 dBi, now maybe you'll need to search for dBi

[radiolistener]

I am getting mixed reports on the optimum length of a dipole antenna.

Half wavelength dipole resonant frequency is very close to a half wavelength. It should be trivial from antenna name. 
It is not exactly half wavelength but very close.

Actual frequency may vary depends on the wire thickness, presence of isolation material around wire (include air humidity, etc) and height above ground.

You can precisely tune your dipole with antenna analyzer by cutting wires step-by-step.

Also note that antenna feeder may be a part of your antenna, especially if you're don't use balun to match half wavelength dipole with coax cable and don't use RF-chokes to cut-off common mode currents. In that case coax cable length and geometry may significantly affect your antenna resonant frequency.

[Fried Chicken]

I should add what I'm doing.  I bought this kit and am exploring the frequency spectrum.  I've finally managed to reliably pick up the airport ATIS which was faint until I messed around with the antenna.  I found full wavelength total length worked better than half wavelength total length.

What I'm wondering is what the logic/physics here is.  That video that I posted is throwing of my conception completely

[DimitriP]

I found full wavelength total length worked better than half wavelength total length.

Oh...."rabbit ears".
You should have opened with that

Forget calculating and "adjust antenna for best reception".

[Fried Chicken]

I found full wavelength total length worked better than half wavelength total length.

Oh...."rabbit ears".
You should have opened with that

Forget calculating and "adjust antenna for best reception".

That's what I did, but i'm curious about the physics and how come I'm getting different recommendations

[radiolistener]

I found full wavelength total length worked better than half wavelength total length.

What I'm wondering is what the logic/physics here is.

The logic/physics here is that full wavelength dipole has bad VSWR, so its output has high mismatch with feeder input and as result it gives much more weak signal to the receiver.

Much more weak signal "worked better" because your receiver performance is bad and it is overloaded with signal from a good antenna.

The same issue for any cheap Chinese radios like Baofeng/Quansheng/etc. If you connect it to a good external antenna it can't work in the modern city environment, because it is flooded with high signal levels. This is because such cheap radios have low dynamic range and unable to see weak signals in presence of a strong signals on different frequency.


In other words, this is not dipole works better when it has length equal to a full wavelength. This is your low quality radio works better when antenna gives more weak signals due to bad antenna performance. It happens because your radio cannot work with a more strong signals from half wavelength dipole.

[radiolistener]

But on the other side, the more large antenna has better efficiency, so if you use proper antenna matching with feeder, you can get better antenna performance. But it also needs to include to analysis of losses in the more long wire and losses in matching circuit.

Here is a chart for dipole radiation resistance depends on it's length:


As you can see, dipole with length 1 lambda has radiation resistance about 200 Ω. So, if you attach it to the 50 Ω coax cable through 1:4 impedance ratio transformer it may give you better performance due to higher efficiency of more large antenna.

More longer wire resistance can be compensated with using more thick wire. So, the efficiency of such antenna will depends on your matching transformer efficiency.

[Fried Chicken]

Wow.  This is not in any way what I was expecting.  I thought it would have something to do with the ground of the AC signal being negative and the other positive causing a higher voltage across the antenna --> ground and therefore the full wavelength would be stronger.  This was my gut assumption, but would mean the youtube video so highly praised is incorrect.

I've looked up the input impedance of my radio:

What is the RTL-SDR input impedance?

Since these dongles are intended for TV, most dongles will have an input impedance of approximately 75 Ohms, although it is unlikely to be exactly 75 Ohms over the entire frequency range.

Remember that the mismatch loss when using 50 Ohm cabling on a 75 Ohm input will be very minimal at less than 0.177 dB.

The 75 Ohm impedance for the R820T can be checked on the datasheet which can be downloaded here.

However, newer dongles that come with SMA connectors will be 50 Ohms.

At this point I feel I need to either break out my university physics textbook, or accept things are the way they are.  I'm not sure what dipole radiation resistance is... I'm also not sure what's meant by feeder input.

[radiolistener]

I'm not sure what dipole radiation resistance is...

https://en.wikipedia.org/wiki/Radiation_resistance
Radiation resistance is that part of an antenna's feedpoint electrical resistance caused by the emission of radio waves from the antenna.

In other words, this is a part of antenna input resistance which leads to energy losses for EM radiation.
The antenna also has a part of resistance due to unwanted energy loss due to heating.

For receiver antenna you can view it as a signal source resistance.

I'm also not sure what's meant by feeder input.

Feeder input is a transmission line input end point which is used to put RF signal from source.
For receiver antenna this is transmission line endpoint which is used to connect to antenna.
If you're using coax cable as transmission line, it will be coax cable end which is connected to antenna.

[radiolistener]

I've looked up the input impedance of my radio:

First and the main goal - you're needs to make sure that the antenna impedance is good matched with your feeder (coax cable) impedance.

For receiver, input port impedance has some effect, but if there is some minor impedance mismatch, this is not so critical like for transmitter.

Regarding to RTLSDR, its impedance depends on the frequency.
For example my RTLSDRv3 input port has Z = 63.1+j14.6 Ω at 117 MHz, so there is no needs to worry if you're use it with 50 Ω or 75 Ω coax cable.

You're needs to take main care about good matching between the cable and antenna impedance.

If you want to match your cable with full wavelength dipole, it's better to use 50 Ω coax cable, because it will be more easy to match with 200 Ω antenna impedance by using 1:4 impedance ratio transformer (1:2 turns ratio).

For comparison, with half-wavelength dipole it's better to use 75 Ω coax cable, because half wavelength dipole impedance is about 73 Ω, so you can use 1:1 impedance match transformer as balun.

PS: adding several RF chokes on coax cable near antenna and near receiver can help to reduce noise interference from home equipment.

[A.Z.]

Now I found a resource suggesting the length of the total dipole should be half the wavelength, each element a quarter of the wavelength

exactly, for a resonant dipole, offering an impedance around 75 Ohm that's the size

For my reference frequency let's do 117MHz, so 468/117MHz gives a wavelength of 4 feet.

Let me do that in meters (I use metric units so it's easier for me), let's see...

Lambda is C/F, we assume C as 300 and F as expressed in MHz, so we have 300/117 which gives us 2.564 (meters), this is the full wavelenght (Lambda), now we calculate the WHOLE dipole spread (both arms) as Lambda/2 that is 2.564/2 and we get 1.282 which is our 1/2 wave dipole overall spread, at this point each arm of the dipole will have a length of a quarter wave, so 1282/2 will give us 0.64 which tells us that each arm of our dipole, for a resonance at 177MHz should have a length of 64cm or 2.09ft

So each element is 2 feet, or half the wavelength. 

see above... it isn't "half" but one quarter.

Note: in the above calculations I didn't consider the "end effect" shortening to keep things simple, but willing to consider it you may read this https://www.kb6nu.com/dipoles-shorter-half-wavelength/ notice that I usually stick with a reduction factor of about 0.965 for 2.5mm diameter copper wires, in this case this would mean that the arms length would become 0.64*0.965 that is 0.618 (61.8cm or 2.02ft)

[A.Z.]

I should add what I'm doing.  I bought this kit and am exploring the frequency spectrum.

That kit may be ok for V/UHF and then, only if you go listening outside in some place far away from the urban "electrosmog cloud", willing to get serious you'd better consider putting up a better antenna and place it outside, be it on a balcony or a window if not over the roof, connect it with a run of RG6 (TV/Sat) coax cable and start listening

[Fried Chicken]

I should add what I'm doing.  I bought this kit and am exploring the frequency spectrum.

That kit may be ok for V/UHF and then, only if you go listening outside in some place far away from the urban "electrosmog cloud", willing to get serious you'd better consider putting up a better antenna and place it outside, be it on a balcony or a window if not over the roof, connect it with a run of RG6 (TV/Sat) coax cable and start listening

In fact I'm about to buy some pizza pans and build a pizza pan antenna and stick it in the attic via the TV coax (I already bought the adapters).

Lots of examples:
https://turbofuture.com/industrial/How-to-Build-Your-Own-Planar-Disk-Antenna

https://thenack.com/2016/11/planar-disk-antenna/

http://www.wa5vjb.com/references/PlanarDiskAntennas.pdf

https://wa5vjb.com/references/Pizza_pan_vertical_2016%20WA5VJB.pdf

[A.Z.]

Well, if you are only interested in SHF signals up to the GHz, that one may be a start, but then don't expect it to give you some decent reception down at UHF/VHF or HF, they just won't perform; if you're going to install an antenna up the roof, my suggestion is to start with a discone, most of them will offer acceptable performance from 25MHz up to about 2GHz

Willing to get down to HF/VHF with decent performance, here's my suggestion

Pick this small 9:1 BalUn https://www.amazon.com/Balun-One-Nine-v2-Barebones/dp/B07XJRTF94 and put it inside a plastic w[e]at[h]erproof junction box, then use some 16AWG insulated wire and lay it out over the roof (or inside attic if you have a wooden roof) forming an horizontal loop and connect the ends of the wire to the balun, cut the loop wire wire at the opposite side from the Balun and insert a resistor with a value around 500...600 Ohm, the latter will flatten the impedance curve allowing to obtain a good match over a wide bandwidth, such a terminated  loop will also offer the advantage of being less "noisy" than other types of antennas, yet it will be able to set the SNR and allow you to pick up signals from MW up to around 50MHz at least

[ahbushnell]

Well, at 0.6 Lambda height it's more than 7 dBi, now maybe you'll need to search for dBi

A dipole is not 7 dBi gain.  It's about 2.15 dBi.

https://en.wikipedia.org/wiki/Dipole_antenna

[fourfathom]

Well, at 0.6 Lambda height it's more than 7 dBi, now maybe you'll need to search for dBi

A dipole is not 7 dBi gain.  It's about 2.15 dBi.

https://en.wikipedia.org/wiki/Dipole_antenna

That's gain in free space. 

The height above ground of a half-wave dipole will cause the gain to increase at some elevation angles, and decease at others.  I don't have the numbers handy, but I believe that the 7dBi value is possible.  Obviously the quality of the ground also is a factor.

[uer166]

Well, at 0.6 Lambda height it's more than 7 dBi, now maybe you'll need to search for dBi

A dipole is not 7 dBi gain.  It's about 2.15 dBi.

https://en.wikipedia.org/wiki/Dipole_antenna

That's gain in free space. 

The height above ground of a half-wave dipole will cause the gain to increase at some elevation angles, and decease at others.  I don't have the numbers handy, but I believe that the 7dBi value is possible.  Obviously the quality of the ground also is a factor.

Would this be freznel zone interaction? I thought antennas were always specced in free space, since the distance to GND is not really a controlled parameter. If you have obstructions in Freznel zone (like a ground), then the pattern gets pretty arbitrary.

Edit: pulled out Kraus book, 2-19 claims that directivity of a Sine power pattern (I assume a 1/2 wave dipole is pretty close to that) has directivity of 1.27, which is 2.08dB.

[Fried Chicken]

I'm not sure what dipole radiation resistance is...

https://en.wikipedia.org/wiki/Radiation_resistance
Radiation resistance is that part of an antenna's feedpoint electrical resistance caused by the emission of radio waves from the antenna.

In other words, this is a part of antenna input resistance which leads to energy losses for EM radiation.
The antenna also has a part of resistance due to unwanted energy loss due to heating.

For receiver antenna you can view it as a signal source resistance.

So kind of like eddy currents?

I'm also not sure what's meant by feeder input.

Feeder input is a transmission line input end point which is used to put RF signal from source.
For receiver antenna this is transmission line endpoint which is used to connect to antenna.
If you're using coax cable as transmission line, it will be coax cable end which is connected to antenna.

I'm confused.  "Transmission line input end point" - so the end of the antenna?  Where the antenna connects to the receiver?  That's where the resistance in a coax cable is put, at the endpoint?

[uer166]

I'm confused.  "Transmission line input end point" - so the end of the antenna?  Where the antenna connects to the receiver?  That's where the resistance in a coax cable is put, at the endpoint?

It's the location where the transmission line ends, and the antenna element(s) begin. In your case it's where Coax is connected to the 2 dipole elements.

[Fried Chicken]

Well, if you are only interested in SHF signals up to the GHz, that one may be a start, but then don't expect it to give you some decent reception down at UHF/VHF or HF, they just won't perform; if you're going to install an antenna up the roof, my suggestion is to start with a discone, most of them will offer acceptable performance from 25MHz up to about 2GHz

Willing to get down to HF/VHF with decent performance, here's my suggestion

Pick this small 9:1 BalUn https://www.amazon.com/Balun-One-Nine-v2-Barebones/dp/B07XJRTF94 and put it inside a plastic w[e]at[h]erproof junction box, then use some 16AWG insulated wire and lay it out over the roof (or inside attic if you have a wooden roof) forming an horizontal loop and connect the ends of the wire to the balun, cut the loop wire wire at the opposite side from the Balun and insert a resistor with a value around 500...600 Ohm, the latter will flatten the impedance curve allowing to obtain a good match over a wide bandwidth, such a terminated  loop will also offer the advantage of being less "noisy" than other types of antennas, yet it will be able to set the SNR and allow you to pick up signals from MW up to around 50MHz at least

I'm not wanting to spend more money beyond what I have.  I've now purchased these pizza discs, and plan on putting two of them together as shown.  This is a limitation I have placed on myself.  I currently use the SDR to listen to ATC and pick up ADSB signals.

[Fried Chicken]

I'm confused.  "Transmission line input end point" - so the end of the antenna?  Where the antenna connects to the receiver?  That's where the resistance in a coax cable is put, at the endpoint?

It's the location where the transmission line ends, and the antenna element(s) begin. In your case it's where Coax is connected to the 2 dipole elements.

On the dinky included dipole antenna, they pulled the shielding and put it on one dipole as the negative, and then took the inside and merely connected it to the other dipole antenna.

There's a resistance here?  I mean it must certainly be negligible?

[uer166]

I'm confused.  "Transmission line input end point" - so the end of the antenna?  Where the antenna connects to the receiver?  That's where the resistance in a coax cable is put, at the endpoint?

It's the location where the transmission line ends, and the antenna element(s) begin. In your case it's where Coax is connected to the 2 dipole elements.

On the dinky included dipole antenna, they pulled the shielding and put it on one dipole as the negative, and then took the inside and merely connected it to the other dipole antenna.

There's a resistance here?  I mean it must certainly be negligible?

If you're asking about the radiation impedance, then that "resistance" as you called, is a physical manifestation of the wave interaction. You want to have it either match to your coax transmission line impedance (50Ohms), as well as your receiver input (also 50Ohms), or convert it to 50Ohms if it isn't already. You place that matching conversion thingy at the antenna feed point as discussed.

There is another thing called loss resistance, which you want as low as possible, since it just converts current to heat. You can't do much about this other than minimize in antenna design. You can expect 50-80% antenna efficiency (50-20% of energy lost to heat) in general. It's caused mostly by skin effect where current in antenna elements crowds to the metal surface.

Comment about pizza pan antenna: since you're not using a balun (and an image in one of your links showed coax soldered to one of the pans), it is not a full dipole. It's a simple monopole antenna with a GND plane of a specific shape. Since the element is round, you get lots of bandwidth out of it, though efficiency probably not the best.

If you make it out of carbon steel, which has a resistance many times that of copper, the efficiency is yet more shit.