As I have started using the UHF 433 systems for range, I have been experimenting with various TX and RX Antennae.
I came across this link http://www.fpvmanuals.com/category/manuals/equipment-manauls/antennas/ which explains how to build what they call a turn stile antenna. Could anyone try and explain why this antenna would work as I had really poor performance with it. I ended up using a simple cross dipole on the RX side with much better results.
Someone I know in the electronics field suggested that the two V's could in fact partially cancel out and in doing so, produce unusual impedance plots at the required resonant frequency. Anyway... any comments would be useful...
Antenna have been the mainstay of my career and have remained an incessant passion with me. I do have some considerable experience in all aspects of antenna design, from the low HF region ( sub 1MHz) up to the mid microwave region , around 8GHz.
So please forgive me if I become a little vocal on the subject..I will try to explain the Turnstile antenna, and what its characteristics are, but need to start with some basics first.
The two most common antenna ( in VHF, UHF and lower microwave regions) are probably the vertical radiator above a ground plane ( the 1/4wave or 5/8wave vertical) and the dipole antenna.
A dipole is a 1/2wavelength at the operating frequency, radiates linearly polarised electromagnetic waves, and typically has an ideal gain of 3dB over the simple 1/4wave radiator. The dipole will be vertically polarised if placed vertical, and visa versa.
a 1/4wave radiator is simply half a dipole, placed above a reflecting ground plane ( either radial wires or a flat sheet of conducting material).
An important consideration of any antenne is its feedpoint impedance. If the antenna were a perfect 50ohm restive input, and you used 50ohm coax to feed it, and your transmitter ( or receiver) has a 50ohm coax connection socket, all is well 0 a perfect match and maximum transfer of energy, with no standing wave ratio ( the measurement of energy reflected back by the antenna to the transmitter, versus that which the antenna radiates outwards.)
Now, a Dipole has a natural feedpoint impedance of around 72ohms, not 50ohm.
a 1/4wave vertical has a feedpoint impedance of around 75ohms, if the ground plane upon which it is mounted is flat (at 90deg to the 1/4 antenna element). If the ground plane were a cone, swept downwards, at 45deg below the horizontal, the feedpoint impedance drops to close to 50 ohms. So this is a demonstration of how to improve the match between the TX and the antenna. ( 50ohm is the accepted impedance standard to which most commercial equipment complies)
But there is another criteria to be met to insure maximum radiation of the TX energy by the antenna, and that is that the antenna element MUST be resonant. This is achieved by ensuring the element is the correct length at the frequency of interest. SO a dipole must be EXACTLY a 1/2 wave length at the frequency of interest with a small correction factor. That factor is dependent on the diameter of the material ( wire, tubing, etc) used to make the element. A dipole at 2.4GHz made from 6mm diameter tubing will be shorter at resonance than one made from 0.8mm wire, by a factor of around 0.93 X the halfwave length.
Now, if you manage to make the element the correct length to be nicely resonant, and the feedpoint impedance is a good match, you have the most efficiency and max radiation of the signal.
Placing the antenna element against any plastic or fibreglass ( for example) will raise the resonate frequency so detuning the antenna from the intended frequency and lowering its efficiency. Placing the antenna adjacent to ANY metalic object will detune the antenna similarly, induce eddy currents in the adjacent metal, which consumes radiated energy from the antenna, reducing the energy actually radiated out to the receiver. In addition, any adjacent metal will act as reflectors, directing the antenna radiation away from it, and shielding reception on the metal side of the antenna. So, keep the antenna in the open, away from other dialectric materials and metals.
Now a dipole, as discussed, has a feedpoint impedance of around 70ohms. You cannot change that. It does not present a good match to your TX. So, what some folk do, is they trim the dipole length, till the SWR measurment approaches 1:1, ie the dipole presents and impedance of 50ohms to the transmitter. Only the transmitter is happy. The antenna is a disaster, because it is no longer resonant, and radiates poorly. You cannot correct an impedance mismatch by adjusting the antenna resonant length!
I said the dipole radiates linearly. If we want circular polarized radiation patterns, there are a number of antenna options - The helical - a very good circular antenna, but very directional, and the turnstile, and omnidirection antenna.
The Turnstile achieves this by feeding two crossed dipoles in phase quadrature. What does this mean? Two dipoles are placed in the same horizontal plane, at 90deg too each other, in a cross or X pattern. The one element is fed with the direct RF energy, while the other is fed with that energy 90 deg out of phase. This will cause each element to radiate in a phase relationship to the other element, such that the resultant combined radiated wave is circularly polarised.
There are two basic ways to achieve this phase ship between elements - One is to use an additional length of coax ( 1/4wave long) between the one element and the feed energy, and feed the other direct. Since a 1/4wave length of coax is electrically 90deg long, it will introduce the phase delay. However, simply connecting the two dipole elements in parralel like this, results in half the feedpoint impedance, ie, around 30 to 35 ohms, giving around 1.4:1 SWR. We cannot trim the element length to tune this, so we must match the feedpoint by other means - this is done by using coax cables of the correct length and impedance, as shown in the diagram below:
Another way ( the 'IBcrazy way', although he did not invent it - It was done invented by a radio ham in the early 60's) is to slightly lengthen the one element ( becomes more inductive) and shorten the other( becomes more capacitive) - this also introduces the required phase difference between the elements. However, in the IBcrazy way, and ALL the derivatives sold in all the shops, there are two fatal flaws.
One , the two elements are simply connected in parallel, resulting in a mismatch, and then the elements are trimmed to obtain a good match, totally upsetting the 90deg phase relationship between the two elements. The result is NOT circular polarisation, but a mix of linear and elliptical, with very poor radiation pattern uniformity.
Two, a fatal mistake in many antenna configurations, the coax is connected directly to the radiation elements without any RF balun or choke. This results in the RF energy from the antenna elements 'running' down the coax as well. The RF does not care where is is in the antenna, and any rf on any metallic part of the antenna will radiated from that part, so the coax, which is supposed to keep the rf 'inside' , radiates as well. This radiation is linearly polarised, and mixes constructively and destructively with the main antena radiation, destroying the radiation pattern and polarization altogether. So such a turnstile construction is a waste of time. Many of the users of the IBcrazy turnstile achieve 'good' range, a few km, etc. A proper turnstile construction, either based on the image above, or my images below, give me 15km line of sight, with identical antenna at each end and 500mwatts at 2.4GHz.
Also, thecoax should depart from the soldered connection at the antenne elements, at 90 deg. In other words, the coax must not lie adjacent to any of the elements, or in the same plane as any of the elements - it will just act as a radiator element and distort the antenna radiation patter completely.
An alternative feed match for the dual crossed dipole turnstile is the split tube balun - a more complex and demanding construction, but very rigid and accurate, as below.
Thats my 10cents worth..
Joe, thanks for a very informative post. Can you provide details on the tube balun calculations needed to determine dimensions for different frequencies? I am very familiar with conventional balun construction but had not heard of this concentric tube version so I looked it up this afternoon. I found some good information, but not how to calculate the lengths.
Nice looking antennas there - would love to hear more about them.
Thank you for the informative post. I really appreciate the time you spent on it. This kind of real info in invaluable!
I will try to put up a blog on this for info - it is a bit of effort to present something decent, and then in a simple reply it is lost from the page in 5 minutes! A blog maybe at least lasts a half day, for more folk to see...
That would be excellent! Thank in advance!
Thanks for this very useful and clear info.
Could you elaborate a bit more for an antenna noob like me about "Balun".
I know what a balloon is but not what a balun is ;)
Another request : what would be the simplest diy 433Mhz I could build and how ? (for DTF UHF on a UAV)
James and the rest, blog done and loaded - may vanish from the page fast , chased by all the lawnmowers..
I think the problem you're having with this antenna may be due to the description of it on fpvmanuals. It says "the turnstile antenna is dual polarized and will help mitigate issues with polarization". Not only is that not true, that's not even a thing. The turnstyle is circularly-polarized. To get the best reception, the antenna at the other end must also be circularly polarized with the same rotation. If the other end is linearly polarized you will have reduced performance and if it is circularly polarized in the opposite direction you will have terrible performance.
That is probably why things got better when you went to a simple dipole.
For 433 MHz, your best bet would be a dipole with a ferrite choke balun. Even without the ferrite it's still very good.
If you still need more range, the Yagi-Uda antenna is a good bet. It's just a dipole with some carefully-sized wires placed near it, they're not even electrically connected to each other.
I'm not sure that anything Joe wrote has anything to do with your question, and I would get second opinions before I spent a lot of time following his advice. I don't know if you noticed this, but the antenna he includes pictures of is not actually a turnstile. A real turnstile uses a carefully-sized length of coax between the two dipoles to delay one of the signals by 90°, so the peaks of the waveform arrive in order at each wire in a perfectly-timed clockwise or counter-clockwise order, hence, circular polarization.
The antenna Joe shows is not something you would show to a beginner, and frankly even if you could do all the math required and built it correctly (which he hasn't) it's still a low-efficiency abomination.
Hope that helps.
I have build this dipole with no balun out of a RG316 coax cable and servo wires. I think I am well tuned on the frequency resonance (length of the dipole branches were cut to 150 mm each to account for a 0,91 velocity factor due to the plastic tube surrounding the servo wires). But it is 72 ohm probably so I'd like to bring its reactance down to 50 ohm. How to ?