EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: TheUnnamedNewbie on November 18, 2015, 12:58:39 pm
-
For a project at uni I have to work with the following antenna: http://www.digikey.com/product-detail/en/ANT-433-SP/ANT-433-SP-ND/1679578 (http://www.digikey.com/product-detail/en/ANT-433-SP/ANT-433-SP-ND/1679578)
I don't understand how you could fit a 433MHz quater wavelength antenna (wavelength at 443MHz is roughly 1m) in a 3 cm long package. Can someone with better knowledge of these matters explain this rf voodoo? Would a quater wave antenna not need to be about 25cm long?
-
Dielectric makes things shrink
-
As well as the dielectric, there might be a bit of a wiggly line in there too.
Or, it might just be a resistor ;-)
-
What the other posters have said: 1) a dielectric can be used to shrink the wavelength (the frequency remains the same, but the speed of light inside the dielectric is reduced; and the wavelength will be reduced by the same factor). 2) The shape of the antenna can help a lot. Look up fractal antennas.
Also, just because you don't have a quarter wave antenna, a smaller fraction-of-a-wave antenna will still generate some signal. It won't be as effective at capturing the signal as a tuned quarter, half or full-wave receiver but it may still pick up enough signal for many applications.
PS. There is no physics defiance going on. It's all good smart applied physics/engineering. ;)
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
I'm firmly of the opinion that no-one fully understands the physics of antennas. I've met a few good antenna designers and they make microwave feritte designers seem normal.
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
I'm firmly of the opinion that no-one fully understands the physics of antennas. I've met a few good antenna designers and they make microwave feritte designers seem normal.
One of the problems is that antenna design, for the most part, does not lend itself to an analytic result: you don't put in a dozen requirements and a single answer pops out on a plate. Instead it is steeped in numerical analysis, with the inputs having, to a greater or lesser extent, mutual effects. It's like a squidgy balloon where you push in one side and it pops out the other, or, in the case of antennas, it pops out in several places. To get reasonable results you need to rely on a combination of prior art and experience.
But yes, the ferrite voodoo in the realms of phase adjusters, isolators and circulators, to me that makes antennas relatively simple.
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
I'm firmly of the opinion that no-one fully understands the physics of antennas. I've met a few good antenna designers and they make microwave feritte designers seem normal.
Actually, firmly understanding the physics of antennas requires 4 straightforward differential equations. They're named after this guy Maxwell. Perhaps you've heard of him ;).
Indeed, the idea of an antenna arose entirely from Maxwell's equations, and the first ones ever built (by our homeboy Hertz) were built with the very specific purpose of proving the existence of electromagnetic waves and their behavior, thus validating the theory.
So, antennas are not mysterious. In fact, they exist entirely because our understanding is so complete that we predicted their existence and how they would work, as well as the existence and behavior of electromagnetic waves years before we even knew they existed.
Of course once the frequency gets high enough, the quantized nature of electromagnetic waves begins causing some strangeness. It starts with molecular rotational modes of water in the atmosphere, and by 300GHz, the atmosphere is totally opaque. And it remains so until an island of visibility we call near infrared then visible light. But for any application (like antennas) below that frequency where electromagnetic waves behave like waves, not photons, all you need is Maxwell's equations.
The monuments we've built to those equations were not done lightly:
(http://i.imgur.com/8cngZTa.jpg)
-
It's not too difficult to make any system complex enough so it becomes difficult to understand in practice. Sure, one can easily think up an antenna whose properties are easier to measure than to calculate. But in principle we have a very good handle on the physics as metacollin points out. The rest, clever and complicated as it may get, is bean counting. The answer to the OP's question is not particularly difficult to understand from first principles.
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
I'm firmly of the opinion that no-one fully understands the physics of antennas. I've met a few good antenna designers and they make microwave feritte designers seem normal.
Actually, firmly understanding the physics of antennas requires 4 straightforward differential equations. They're named after this guy Maxwell. Perhaps you've heard of him ;).
Indeed, the idea of an antenna arose entirely from Maxwell's equations, and the first ones ever built (by our homeboy Hertz) were built with the very specific purpose of proving the existence of electromagnetic waves and their behavior, thus validating the theory.
That's as maybe, but I defy any individual on their own to be able to look at Maxwell's equations and say "I know, I'll make a Yagi-Uda antenna from that".
-
That's as maybe, but I defy any individual on their own to be able to look at Maxwell's equations and say "I know, I'll make a Yagi-Uda antenna from that".
You're defying Uda and Yagi then.
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
If it's really +3dBi then you've just invented a perpetual motion machine.
-
That's as maybe, but I defy any individual on their own to be able to look at Maxwell's equations and say "I know, I'll make a Yagi-Uda antenna from that".
You're defying Uda and Yagi then.
That was two people, not an individual, and without doubt they'd based their findings on prior art, not just a few abstract equations on their own.
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
If it's really +3dBi then you've just invented a perpetual motion machine.
dBi is the gain WRT an isotropic radiator that radiates power equally in all directions. 3dBi is absolutly possible by concentrating that power over a smaller pertion of a sphere. A dipole antenna doesnt have any gain in the axis of the elements so hence has a posative gain figure if it is efficient. Yagi antennas achieve gains up to about 20dBi by being highly directional.
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
It's there in the datasheet under Electrical Specifications
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
If it's really +3dBi then you've just invented a perpetual motion machine.
dBi is the gain WRT an isotropic radiator that radiates power equally in all directions. 3dBi is absolutly possible by concentrating that power over a smaller pertion of a sphere. A dipole antenna doesnt have any gain in the axis of the elements so hence has a posative gain figure if it is efficient. Yagi antennas achieve gains up to about 20dBi by being highly directional.
DOH! had a bit of a brain fart there
-
I cant see anything that states that its a quarter wave antenna. It just says that it works at 433 MHz and has a gain of minus 6dBi. That seems perfectly possible. A quarter wave monopole would be closer to plus 3dBi.
It's there in the datasheet under Electrical Specifications
You are correct it does state that. I only looked at the discription in digikey.
It may be 1/4 wave in a high dielectric as others have mentioned. Looks more like a PIFA if i had to guess.
I note they show an excellent VSWR and state that this means that most of the power is being transferred to the antenna. What it doesnt state is thst much of this transferred power is turned to heat. How else would the gain be less than 0dBi.
Im not saying this is a bad antenna. Small antennas have a small aperture so its not possible to get good efficiency from them. A small antenna that doesnt reflect lots of power back into your transmitter may well represent a good tradeoff.
Its also worth noting that it only gives the quoted gain when used with a large groundplane.
-
I note they show an excellent VSWR and state that this means that most of the power is being transferred to the antenna. What it doesnt state is thst much of this transferred power is turned to heat. How else would the gain be less than 0dBi.
I've read something about this type of scenario - they add resistance to increase the bandwidth of a shortened antenna. Which would make sense only if it wasn't 1/4 wavelength after all.
The tradeoff is a loss in efficiency as the resistive losses become significant compared with the radiation resistance. The radiation resistance and bandwidth drops as the antenna becomes shorter.
He states that the practical lower limit is about 10% of the wavelength.
[from Thomas Lee, Planar Microwave Engineering, The Dipole Antenna]
I always thought it was fascinating that antennas could (theoretically) be arbitrarily short.
-
To the OP:Take the datasheet to read "Electrical Wavelength"
Depending on the frequency you can get "pretty close" with a measuring tape. Then you have to put the measuring tape away .
-
I am quite unfamiliar with the workings of antennas and didn't realise a dipole antenna was allowed to be constructed with a curvy line, or that dielectrics could be involved. In my head there had to be a 15cm long copper wire in there somehow (naive as I am), which is why I asked the question.
-
[implies that that Maxwell created 4 equations to describe electromagnetism, instead of the actual 20, and conflates the Maxwell and Heaviside sets of equations]
The monuments we've built to those equations were not done lightly:
Monuments are great for mythmaking, not so much for accurate history.
-
As stated before it has a gain of -6.4 db. I have used this type of antenna from Linx before and they are not 1/4 wave. They are a zig zag inside that approximates a 50 ohm load but are not very efficient.
-
I always thought it was fascinating that antennas could (theoretically) be arbitrarily short.
I'm not familiar with the antenna design referred to by the OP but it's possible to model a short monopole using a few crude/basic design equations. I've some experience of doing this stuff at work but I'm definitely NOT an antenna engineer :)
If you were to shorten a 433MHz 1/4wave monopole antenna down to about 1/20 of a wavelength the crude equations predict a radiation resistance of about 1 Ohm and this is in series with a tiny capacitance. The equations predict about 0.8pF.
So to get any kind of efficiency out of this antenna you would have to add series inductance (i.e. a loading coil) to resonate with this capacitance and also you would have to (efficiently) match to the 1 ohm radiation resistance.
In reality, the matching will not be efficient and there will be losses in the series inductor and so the overall antenna efficiency will degrade and also the bandwidth will be very small due to the high Q of the structure. It will also be prone to detuning effects from nearby objects.
i.e. a fair bit of transmitter power would get converted to heat in the ohmic losses in the resonating inductor and the matching network because these resistances will be significant wrt the (undesirably low) radiation resistance of just 1 ohm.
There are tricks you can do to raise the radiation resistance of short antennas like this (so you can design for better efficiency and bandwidth) but you still end up with a relatively inefficient antenna and the bandwidth will still be very narrow.
-
He states that the practical lower limit is about 10% of the wavelength.
In this case of a 433MHz monopole of length 1/10 wavelength, the radiation resistance would be up near 3.5 Ohms and this would be in in series with 1.2pF making this much easier to (efficiently) match into.
The bandwidth would be wider than the 1/20 wavelength antenna but this would still be a very narrow bandwidth antenna. However, I'd expect the efficiency to be much better than the 1/20 wave monopole because the ratio of radiation resistance to the other ohmic losses will be so much better :)
-
I always thought it was fascinating that antennas could (theoretically) be arbitrarily short.
I'm not familiar with the antenna design referred to by the OP but it's possible to model a short monopole using a few crude/basic design equations. I've some experience of doing this stuff at work but I'm definitely NOT an antenna engineer :)
If you were to shorten a 433MHz 1/4wave monopole antenna down to about 1/20 of a wavelength the crude equations predict a radiation resistance of about 1 Ohm and this is in series with a tiny capacitance. The equations predict about 0.8pF.
So to get any kind of efficiency out of this antenna you would have to add series inductance to resonate with this capacitance and also you would have to (efficiently) match to the 1 ohm radiation resistance.
In reality, the matching will not be efficient and there will be losses in the series inductor and so the overall antenna efficiency will degrade and also the bandwidth will be very small due to the high Q of the structure. It will also be prone to detuning effects from nearby objects.
i.e. a fair bit of transmitter power would get converted to heat in the ohmic losses in the resonating inductor and the matching network because these resistances will be significant wrt the (undesirably low) radiation resistance of just 1 ohm.
There are tricks you can do to raise the radiation resistance of short antennas like this (so you can design for better efficiency and bandwidth) but you still end up with a relatively inefficient antenna and the bandwidth will still be very narrow.
Since a dipole has symmetry in its favor, my physics professor solved maxwells equations for a dipole as an example, and the resulting pattern matches the donut/torus pattern you'd expect from a dipole. But forget trying to set up the integral for anything but a dipole or monopole. I showed him a RCHP patch on a choke ring and asked if he could do that, and he told me to get out :-DD
-
I am quite unfamiliar with the workings of antennas and didn't realise a dipole antenna was allowed to be constructed with a curvy line, or that dielectrics could be involved. In my head there had to be a 15cm long copper wire in there somehow (naive as I am), which is why I asked the question.
As a general rule you are about right, using dielectrics and scrunching up the antenna makes them less efficient, otherwise all those TV antennas you we on roofs would be a lot less unsightly. No such thing as a free lunch!
-
Another possibility is a small magnetic loop. I recommend reading TI's AN058 to start. Fully understanding the physics of antennas can take a lifetime...
I'm firmly of the opinion that no-one fully understands the physics of antennas. I've met a few good antenna designers and they make microwave feritte designers seem normal.
Actually, firmly understanding the physics of antennas requires 4 straightforward differential equations. They're named after this guy Maxwell. Perhaps you've heard of him ;).
Indeed, the idea of an antenna arose entirely from Maxwell's equations, and the first ones ever built (by our homeboy Hertz) were built with the very specific purpose of proving the existence of electromagnetic waves and their behavior, thus validating the theory.
So, antennas are not mysterious. In fact, they exist entirely because our understanding is so complete that we predicted their existence and how they would work, as well as the existence and behavior of electromagnetic waves years before we even knew they existed.
Of course once the frequency gets high enough, the quantized nature of electromagnetic waves begins causing some strangeness. It starts with molecular rotational modes of water in the atmosphere, and by 300GHz, the atmosphere is totally opaque. And it remains so until an island of visibility we call near infrared then visible light. But for any application (like antennas) below that frequency where electromagnetic waves behave like waves, not photons, all you need is Maxwell's equations.
The monuments we've built to those equations were not done lightly:
(http://i.imgur.com/8cngZTa.jpg)
You do realise those aren't Maxwell's equations?
They are derived / recast directly from the work of Maxwell, but those you see above in vector form are due to Oliver Heaviside.
-
You'll find xray imaging of the 403MHz version of this antenna on the last page of this PDF : http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1062&context=eesp (http://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1062&context=eesp)
-
That's as maybe, but I defy any individual on their own to be able to look at Maxwell's equations and say "I know, I'll make a Yagi-Uda antenna from that".
You're defying Uda and Yagi then.
That was two people, not an individual, and without doubt they'd based their findings on prior art, not just a few abstract equations on their own.
Well, it was mostly Uda, and sure the shoulders of giants are there for standing on. One could not very well base an antenna on Maxwell's equations without using the work of Maxwell, who himself relied on what came before him. @Zad As for those equations on the side of the building not being Maxwell's equations: they are in fact what are called Maxwell's equations, written in modern integral form.
-
Those 4 simple equations have simple closed form solutions for a very small number of cases. The easiest cases have conducting boundaries in a rectilinear arrangement. With a step up in math complexity you can add cylindrical systems, and with another step up you add systems with spherical symmetry.
Beyond those simple cases you have two choices, experience and intuition is one choice, numerical simulation is the other. In both cases there is a lot of cut and try, local optimization and swearing. In the past as guys like Yagi refined their concepts rule of thumb design equations could be developed. They don't directly solve the big four equations, just provide rules for fabrication that give good results. The numerical simulation approach is getting far better over the last few decades, and has enabled much broader experiments in antenna configuration. It is a way of fast cheap cut and try guided by experience and imagination. There is hope that concepts like genetic algorithms may supercede the experience and imagination part, but at the moment that approach is still computationally challenging.
Back to the dimensions of this quarter wave antenna. Remember that quarter wave is a concept describing how many waves are "resident" on the antenna at resonance. Many things can reduce this number, including an RLC circuit and changes in dielectric. The RLC concept is often used to reduce the physical length of whip antennas. That is the "loading coil" seen at the base of many antenna designs.
-
Beyond those simple cases you have two choices, experience and intuition is one choice, numerical simulation is the other. In both cases there is a lot of cut and try, local optimization and swearing. In the past as guys like Yagi refined their concepts rule of thumb design equations could be developed. They don't directly solve the big four equations, just provide rules for fabrication that give good results. The numerical simulation approach is getting far better over the last few decades, and has enabled much broader experiments in antenna configuration. It is a way of fast cheap cut and try guided by experience and imagination.
+1, that about sums it up! :-+