Jeff asked about adding a 'Balun' to a simple dipole, so I thought to disertate a little once more on the subject...
The purpose of a Balun is to allow the connection of a coaxial feed line ( unbalanced feedline) to a balanced antenna , such as a dipole.
Picture at left is a Circularly polarised turnstile antenna. Although this is for 75ohm coax, similar strusture are possible for 50ohm as well. Significant in this particular design is the fact that the 75ohm phasing cable section serves the function of the Balun as well.
What we are trying to achieve with a Balun, or its electrical equivalent, is to prevent common mode currents from flowing down the coax feedline. These currents are induced when a dipole is simply connected directly to a coax feedline, one element to the inner conductor, and one to the outer. The currents flowing in the dipole elements result in the electromagnetic wave being generated and radiated by the dipole elements. However, the current flowing in the element connected to the coax outer sleeve has to return to the generator ( the transmitter) and so does this by flowing on the outer sleeve of the coax cable. This current induces radiation from the coax outer sleeve, thereby distorting and destructively interfering with the dipole radiation pattern. This current can also return all the way back to the transmitter, and induced energy flows in adjacent cables and looms - In a small UAV with wiring close to everything, this sometimes manifests as servos twitching in sync with the video or datalink transmissions, etc, among other phenomena.
The balun on the left could be used to feed a folded dipole. Such a dipole has a feedpoint impedance of around 300ohms, and this balun exhibits an impedance step-up ratio pf 4:1. SO a 75ohm feedline would result in a good match to the 300ohm dipole, with a unbalanced to balanced trasformation in the deal.
However, there exist a number of alternative 'baluns' that can be used where an impedance transformation is not desired, a sort of 1:1 Balun.
Note that these are not really baluns in the true sense, but actualy perform the same duty by acting as a choke for the RF currents flowing on the coax outer sleeve.
The left image is often referred to as the 'Bazooka' Balun. It is a 1/4 wavelength of tubing, snugly fitted over the coax sleeve insulation, with the bottom end of the tube soldered all the way around to the coax sleeve braid. The top end of the tube is open and insulated from the rest of the antenna. This works by the 1/4 wave section forming a short circuit to the flowing currents at the base, and a high impedance at the top, choking of said currents.
The following are variations of the theme:
Where in the Bazooka Balun a tubular sleeve surrounds the Coax, a single 1/4wave length of conductor can be substituted in the following manner.
The Bazooka balun is preferred and is more efficient.
In order to not distort the antenna radiation patterns ans not cause EMI with other on-board electronics, it is always desirebale to use a Balun type feed for balanced antenna such as dipoles, Turnstiles, etc. The examples shown can be used with 1:1 and 4:1 impedance match for all dipole types.
Yes, any piece of copper wire is ok - not too thick so that is difficult to heat and solder without melting the rest of the installation, if not Teflon coax, and not too thin so that is is difficult to keep snag against the feed line - around 1 to 1.5mm is fine.
I appreciate your detailed answer to my question. It has been very helpful in my own antenna design. In your example to a alternate for the bazooka balun, you said "a single 1/4wave length of conductor can be substituted in the following manner.". Can that conductor be any piece of wire? I see in many examples people use a section of the feeder coax and only use the outer conductor for the balun. I would rather just use a section of 22ga wire.
Also, I forgot to mention if your goal is just a flat antenna, that can easily be done with regular round copper wire and dollar store foam board. I have seen Yagi-Uda, Moxon, and dipole loop antennas in this configuration,
There are already many printed antennas available and in use. For example:
It isn't very hard to make a PCB antenna, in fact when designing PCBs it can be hard not to make a PCB antenna that radiates harmful interference from your circuit.
The problem is that many PCB antennas are not very efficient. Antennas work better and more predictably when the elements are made of a generous amount of very conductive material. In contrast, the copper on a PCB must be quite thin. While the wires in a cheap 5.8 GHz cloverleaf antenna might be thick enough to carry 20 Amps, a PCB trace that could carry 20 amps would have to be about 15mm wide, obviously too wide to fold into a 5.8 GHz loop. Many PCB antenna designs get around this problem by wetting the antenna traces with solder, thereby adding an extra-thick layer of conductive metal.
Another problem is that PCB traces are formed by covering the entire board in copper and then chemically etching off the copper you don't want. This can present an environmental problem, and it can be difficult to get the expensive dissolved copper out of the etchant. Maybe this is why cell phones almost universally now use antennas that are punched out of a sheet of solid copper
If you still want to try to make a PCB antenna, Iulian Rosu has a collection of Design Ideas for Printed and Microstrip Antennas which should fit a range of needs. The Antenna Theory website also has a List of Antennas whish can provide inspiration.
Hope this helps :)
Many thanks for explanation :)
Horisontal F antenna?
433 band is great for long distance fly...
The calculations for these antenna are 'simple' , but are affected by so many factors that they rarely work from the calculations! For the wire version, the wire thickness has an effect, the proximity of the 1/4wave 'spokes' of the wheel to each other have an effect, etc, none of which are easily compensated for in the maths. A simple solution for the PCB version is as elusive, for the same reasons as the wire version, but further complicated by the PCB laminate characteristics, permittivity, etc. They also end up quite large for the lower frequencies - about 350mm diameter at 433MHz, and only become practical from maybe the 868MHz frequencies upwards. However, PCB laminate losses ( the resins in ordinary FR4 PCB laminates) begin to affect the higher frequency implementations detrimentally - esp from 2.4GHz and up.
There is a fellow marketing/selling these still I believe - I think it is KENT Electronics..
I originally designed our SurVoyeur UAV to be fitted with a planar wheel on 868MHz, the frequencies we are legally able to use in our country, and after many trial and error attempts achieved a PCB design that worked very well - it was designed to fit over the hole/bulge seen in the aircraft image below, as a cover and antenna. But when we tried to make more, from a different batch of PCB laminate, they did not tune properly and we discovered it was the difference in permittivity that was messing it all up - so we gave that idea up and went to a Skew Planar Wheel antenna, as seen in the cavity in the fuselage - that works very well. The picture shows measurements being done to see if the proximity of the fibreglass fuselage proximity affected the antenna tuning at all.
How about PCB planar wheel for 433 MHz? I never seen a really simple calculation for this type of antenna.
(are you in RSA?) You are quite correct. There is a range of antenna that could be easily duplicated and provide consistent performance when manufactured on of from PCB materials. The question really is what type of antenna constitutes a 'simple, good antenna'? A dipole is simply and quite easy to make conventionally, with wire, etc. For the DIY enthusiast it is not so easy to fabricate this on PCB, in fact probably for most , PCB manufacture is not really a DIY matter. The more complex antenna types certainly make it worthwhile moving to a sort pf PCB construction, especially where assembly and test man-hours are saved.
For example, the collinear antenna, which consists of a number of vertically orientated antenna, such as dipoles, stacked one above the other, suitably spaced. This increases the antenna gain over that of a single dipole by squashing the dipole doughnut into a more flattened and broader 'wheel'. Such collinear antenna can be made from conventional dipoles, fed with coax cable, and suitably spaced in the vertical plane, or by means of 1/4 and 1/2wave sections of coax cable connected criss cross end to end, and also by microstip techniques on PCB laminate.
Here are some examples of the latter two collinears mentioned:
The Coax collinear ( sometimes referred to as the CoCo antenna ?!??)
The RF input is on the right side, and the antenna would be placed vertically, with the right side of the above inage at he bottom.
A PCB stripline version, of two different implemenations.
Without discussing how the collinear works, as
that was not your question, it is easy to see that the PCB version is a lot easier to fabricate in production.
The one on the right is fed at the bottom.
The effects of the PCB laminate ( the fibreglass or teflon impregnated laminate) on the element dimensions is critical in the sizing of the copper elements on the laminate. As you would have to modify wire element lengths when encased in plastic, teflon, etc, according to the velocity factor, so the velocity factors related to the PCB laminate, materials used, etc, also have to factored in.
And therein lies a problem. The normal FR4 type PCB laminates used in most electronic equipment PCB's is designed for use in such electronics, not for antenna or stripline use, and as such the material permittivity is not closely controlled - for FR4 it varies anywhere from around 3 to 5. This means that once you have designed and tuned an antenna on FR4, got it all working, produced a working batch, and then decide to go into production, there is no guarantee the next batch will work. The permittivity of the PCB laminate you purchase for your next production run will most likely be different, and the antenna will not tune properly...
A further problem is that the fibreglass is laminated in place using epoxy resins, and the resulting structure is very lossy as the RF frequency goes up. FR4 is not to be recommend at 2.4GHz and certainly not at 5GHz up. Although the antenna made on FR4 at 5.6GHz may see to have a good SWR, and be broad banded, chances are that is appears good due to the looses in the FR4 material, or at least the 'good' characteristics are enhanced by the FR4 losses - any losses will broaden bandwith and reduce emissions and performance. Put a few hundred watts into such an antenna at 2.4GHz and in most cases you see the FR4 laminate go up in flames due to the laminate losses. And the losses remain losses - the same losses are there all the time, not only with high power.
FR4 is the PCB laminate commonly and inexpensively available to all. However, there are special RF laminates available, made by DuPont and others. There are a variety of laminates, thicknesses, etc, designed for optimal use at different frequencies. Some of these have glass fibres embedded in Teflon ( instead of epoxies) and are very low loss - also available in various permittivity values, all very well controlled and you are able to purchase identical materials each time, BUT they are pricey with minimum sizes, etc - not really in the DIY domain.
The cost of manufacture, the cost of the process, the materials, etc, does not extend that technology down into the 'cheap' DIY world - the sellable quantities are to low and the prices acceptable to the user would be to low for the producer.
Having said all that, certainly of 433MHz this can be done with FR4, even on 1.2GHz, with no problem. For hobby applications, it is even possible on 2.4GHz, but the producibility of product due to the varying FR4 permittivity makes it a nightmare for the producer.
"should provide consistent performance (when made from the same material)"
hits the nail on its head! when made from the same material!!
Bottom line, yes, it can be done, but it is not feasible in the average DIY workshop, is not easily controlled, and the market is too small to attract the bigger manufactures sufficiently. There may come the day when UAV's are so commonplace, that more of the military manufacturing concepts reaches the civil sector, at understandable prices!
Ok... This has been rather informative, from the both of you... Here is my question...
Why can't we just have simple good antennas that can be manufactured on PCB? That way everyone could have access to good designs that should provide consistent performance (when made from the same material) ?? The size on both 1.2 2.4 and 5.8 shouldn't be a problem.. Thinking of the V (Dipole) antennas that one used to get...
Never claimed to be one either, just sharing the flavour of the day - and certainly few could approve of your attitude, but the world is filled with different folk, so I am sure there is space for your type. I did not start this process - never accused anyone of dishonesty, etc. Seem the internet's anonymity brings out odd traits in some folk. May I suggest, again, that the bantering desist, and if you feel you have superior knowledge , bring it to the forum in a practical way where folk may benefit - since you believe I cannot you must obviously know how to do it - please do so, I gracefully bow out to your intellect Jonathan.
Have a good day!