Page 1 of 23 12311 ... LastLast
Results 1 to 10 of 230

Thread: MYTHBUSTERS: Letís talk seriously about antennas, or 10 myths busted.

  1. #1

    MYTHBUSTERS: Letís talk seriously about antennas, or 10 myths busted.

    Over the last couple of months it seems a war has been started against the new SpiroNET antennas from Fatshark/ImmersionRC. Although I have been mentioned a couple of times in the related threads, I have refused to jump into the discussions. But I cannot avoid getting uncomfortable each time a read a "not accurate" (I wouldn't like to say "false") technical claim, and the reality is that around the SpiroNET discussions a lot of confusion has been spread regarding antennas, especially from some self-proclaimed "experts".

    I suppose I will get some strong responses in this thread, so I would like first to clarify:

    - I am not a Sander's groupie. It is just a question about respecting someone else's job, especially when he has brought so many brilliant designs to the FPV arena.

    - Without any doubt, the main responsible for the fact that we all fly with Circular Polarization antennas is Alex (IBCrazy). His tutorials at RCG really spread the word, and we all should be grateful to him for his efforts and generosity. But, he hasn't invented Circular Polarization, nor owes the Skew Planar Wheel concept, and he wasn't the first to suggest the use of CP for FPV, just learned from some of the pros, like OldManMike, Devonboy/Devonian, Mr. RC-Cam and some others. Alex is too honest to claim any of this, but I have seen some others (wanting to protect/defend him) making similar claims on his behalf.

    - Both Alex (IBCrazy) and Hugo (Hugeone) have proven an enormous level of generosity. I am confident that the fact they started to share and spread technology brought them in an unplanned way into selling antennas. They both welcomed me when I started to follow exactly the same path, and we have always respected each other and even helped. I cannot say the same for some of the newcomers. No more comments...

    - I have got no intention to start endless discussions. I created this thread just because I felt a bit guilty for not giving my own point of view about some technical discussions which are confusing people. I will politely answer technical questions, but I won't lose my time trying to justify endless times technical facts.

    - I drive Circular Wireless as a hobby. The incomes are welcome, 80% reinvested, and enjoyed, but I am lucky enough to not rely on them. Ours being the most expensive antennas, I forecast that with this thread I will help sales of cheaper options, but I don't care too much.

    - My background is Telecommunications &Electronics Engineer (not sure about worldwide degrees equivalences, let's say is the top level, just below Doctor). My investigations on wireless video systems brought me to recover some old relationships from my University times, and right now I have the luck to work together with Antennas Doctors, and get constant access to a very modern antenna laboratory, which cost way more than my house. Not wanting to swank about it, just to explain that I am not talking nonsense.

    So, let's start to bust some myths...

  2. #2

    1. VSWR IS THE MAIN INDICATOR OF ANTENNA'S PERFORMANCE

    FALSE. First of all, VSWR is for HAMs. Microwave engineers prefer to use the Return Loss parameter. There is a direct correlation, and you can even find in Internet dozens of calculators. VSWR, Return Loss, S11 or whatever you prefer to work with, is just an indicator about the antenna impedance matching with the active equipment input or output impedance. In our case, our goal is to get an antenna as close as possible to 50 Ohms. The issue is that the impedance has a real part (resistive) which is constant, but also a complex part (inductive and capacitive) which varies a lot with the frequency. For this reason, what we want to do is get 50 Ohms at our frequency of interest.

    The Return Loss parameter measures the amount of energy reflected back by the antenna at different frequencies. There is an industrial consensus about considering that the antenna is matched at any frequency where Return Loss is <-10dB, which means that only 10% of the power is reflected back (i.e. 90% of the power is "profited" by the antenna). In Circular Wireless we warrant -13dB, which means that only 5% of the power (maximum) is reflected back. You can achieve a perfect match at a specific frequency, and get almost 0% reflections. This would give approximately a 2,5% extra range... Variations in ambient humidity will give you much bigger range variations. So, let's say that becoming obsessed with the VSWR or Return Loss value is just useless.

    The fact to get the antenna 90%, 95% or 99% impedance matched would affect exclusively to the range, with 5% range variations in the best theoretical case. Just to give you an example, a 50 Ohms resistor would make a perfect impedance match across the whole spectrum, but this does not mean it is an antenna...

    OK, but yes: 2.5% more range is 2.5% more range. No doubt. BUT, there are way more important parameters, and no one is talking about them, just because no one is testing them (the equipment to do so is 6 digits figure at least...).

    The purity of the circular polarization (which can go from a line, through an ellipsoid, to a perfect circle) also varies with the frequency, and not always happens at the "best impedance matching" frequency. This purity will make the antenna better for barrels, as the X and Y components will be the same, and you won't lose range when flying knife edge.

    The uniformity of the radiation pattern in the XY plane would mean that you can do 360ļ turns without losing signal.

    The uniformity of the radiation pattern, together with the lowest possible gain, will mean that the antenna's axis radiation nulls will be narrower, and you will lose signal very briefly when doing loops.

    So, VSWR is not so important, and it shows a very small part of the picture regarding the performance of the antenna.
    Last edited by sircana; 30th August 2012 at 02:24 AM.

  3. #3
    2. THE RESONANCE FREQUENCY OF AN ANTENNA MUST MATCH THE EXACT OPERATION FREQUENCY (I.E. THE SELECTED CHANNEL)

    FALSE. As explained in the Busted Myth #1, the antenna is valid for operation across a frequency range, where Return Loss <-10dB. Inside this frequency range (which is the bandwidth of the antenna), performance variations in terms of range are impossible to notice.

    To prove this, I am going to confess that the Return Loss graph we show in our webpage for the SPW58 does not correspond to the current design, but to the earlier prototypes. The ones we are supplying right now have got their lowest Return Loss (i.e. their resonance frequency) above 6 GHz.
    Why? Just because size matters, both in terms of drag and strength. We decided to shift the bandwidth of the SPW58 upwards, in a way that we still can warrant -13dB Return Loss in the band of interest (5645-5945MHz, much smaller than the total bandwidth of the SPW58), but making the antenna much smaller. The below photo shows the difference between the current product (left) and a hand made one perfectly tuned to 5800MHz.



    So, becoming obsessed about matching the antenna's resonance peak to a specific frequency is just useless.

    P.S. At this point I would like to invite any other antenna manufacturer to do range test comparison between a pair of Circular Wireless SPW58 and their own Omni antennas combo. If anyone can prove that their antennas are noticeably better (in exactly the same test conditions) I will not only refund him whatever he would have paid for the SPW58s, but also send free of charge a full set of CW products.
    Last edited by sircana; 30th August 2012 at 02:27 AM.

  4. #4
    3. PERFECTLY TUNED OMNIDIRECTIONAL ANTENNAS CAN DOUBLE THE RANGE OVER NOT SO PERFECT ONES

    FALSE. The range depends basically on the following factors:

    - Transmitter power.
    - Power coupled to the transmitter antenna (i.e. the impedance match = VSWR = Return Loss)
    - Transmitter antenna directivity (what we use to call -wrongly- "gain")
    - Propagation losses (mainly depending on the distance, but also with atmospheric conditions)
    - Receiver antenna directivity
    - Receiver sensitivity

    Pretending that only improving the impedance match can give you the extra 6dB you need to double the range is just hilarious.
    Last edited by sircana; 29th August 2012 at 06:52 PM.

  5. #5

    4. THE NUMBER OF LOBES HAS A HUGE IMPACT ON PERFORMANCE

    FALSE. The Skew Planar basic concept works for any number of lobes equal or bigger than 3. With 3 lobes the uniformity of the radiation pattern and the axial ratio in the XY plane is slightly affected, but it works. With 4 lobes you achieve almost perfect uniformity in the XY plane. Anything above that (5, 6, 12 lobes) just adds nothing in terms of performance.
    Last edited by sircana; 29th August 2012 at 06:51 PM.

  6. #6
    5. EVERY SMALL LOSS REDUCTION FROM THE COAXIAL CABLE, CONNECTORS OR ADAPTERS IS VERY IMPORTANT

    FALSE. I have read somewhere a discussion about using a different coaxial cable, just because the "bad" one had 300dB/100m losses at 5.8GHz.

    OK, let's assume that we change this "bad" coax, and we use a perfect one, with no losses. That means that for 10cm of cable (I cannot find any reason to use longer lengths, but OK, let's make it 20cm...), so for 20cm of cable you get 0.6dB losses. This means in terms of range that if you avoid this unwanted 0.6dB losses, you will get a theoretical range increase of 7,15%. WOW!!!

    I am going to confess another thing: We have changed the design of our SPWs, and instead of giving the choice of using straight or angled connector, now we only produce straight SMA antennas, and we include a SMA angled adaptor.

    Do you know the difference between these two adaptors? (the one we include is the shiny one, on the right):



    The left one is suited for 5.8GHz. The right one works fine for 1.2 and 2.4GHz, but no so fine at 5.8GHz.

    The left one cost more than our SPW selling price (35 euro), the right one is affordable.

    Do you know the impact of using the right one instead of the left one, in terms of range, image glitches when performing loops, barrels, turns, etc.? NONE.

    If someone wants to go the extra mile, he can always buy the left one. However, I wouldn't recommend wasting the money. Not for 5.8GHz, and not for wireless analogue video transmission.
    Last edited by sircana; 29th August 2012 at 06:51 PM.

  7. #7
    6. IT IS DIFFICULT TO PUSH VIDEO RANGE LIMITS

    FALSE. Just buy a linear polarized omni antenna on the plane with high gain (let's say 6-8dB), fly flat, buy the bigger dish you can find, and keep it perfectly pointed to the plane. Easy.

    Considering that the 5.8GHz range record was hold for a long time by a guy who lost the plane in that flight, and that the current one has been made with a quite stable plane, equipped with RTH, I honestly believe that they doesn't add too much to improving FPV experience.

    Flying with a guess of where the horizon is, and where the return home OSD arrow is pointing, without being able to see anything else because of the glitches and the snow over a B&W image, and knowing that in case of failing to guess where the plane is going you can always switch on the RTH, IMHO is useless.

    I am way more interested in improving video quality, which implies improving the immersive FPV experience. For me the maximum range of any setup is when you get the first hint of noise or the first glitch.
    Last edited by sircana; 30th August 2012 at 08:24 AM. Reason: not being respectful with other people achievements

  8. #8

    7. A HIGHER GAIN ANTENNA ON THE TRANSMITTER GIVES YOU MORE RANGE

    FALSE (WELL, EXCEPT IF YOU FLY LIKE A GIRL...). For omnidirectional passive antennas, there is only one way to increase the gain: flattening the radiation pattern. This means that when pushing the range limits you will lose signal every time you bank, climb or dive the plane. Useless IMHO.

    I prefer to design omni antennas with the lowest possible gain, and increase the gain in the receiver end. This way you can fly much more acrobatic without video glitches (except if you point the radiation null directly to the receiver, what is quite difficult to achieve in normal flight).

  9. #9

    8. USING FREEWARE SIMULATION TOOLS IS ENOUGH TO DESIGN OR OPTIMIZE AN ANTENNA

    FALSE. Software simulation tools do not always tell the truth, specially the simpler ones like NEC.

    Just to give an example, there is a circular polarized design which shows a wonderful NEC simulation, with a perfectly uniform radiation pattern (and no, I am not talking about Cloverleaf...).

    This is a plot of the real radiation pattern of this antenna, tested in an anechoic chamber. Not so nice...




    Simulation software does not always tell the truth. In fact for professional designs it is quite usual to simulate in different software platforms, as each one implement different algorythms.
    Last edited by sircana; 30th August 2012 at 11:32 AM.

  10. #10

    9. DEVELOP A NEW ANTENNA (DESIGN OR PRODUCTION METHOD) IS NOT SO DIFFICULT

    FALSE. At Circular Wireless we have invested a significant amount of resources trying to develop a "flat" Omnidirectional Circularly Polarized antenna.

    A successful antenna development is not something you get by chance. There are several steps to follow:

    DESIGN



    SIMULATION



    PROTOTYPE BUILD



    LAB TESTING... OUCH!! FAILED. AND IT LOOKED NICE IN THE BEST SOFTWARE SIMULATION...




    This is regarding design. Improving production methods over existing designs is not obvious...

    Industrial processes are industrial processes: accurate, consistent, efficient, etc.

    They can never compare to hand build. But you need to develop these industrial processes.

    Just an example (let's show some more "secrets"...):



    Not as easy as it looks. 3 different suppliers evaluated, and many tests until finding the perfection.

    The result? Perfect axial symmetry. Just a comparison between our SPW58 and a handmade SPW:




    Last edited by sircana; 29th August 2012 at 07:21 PM.

Similar Threads

  1. LRS UHF: Letís talk about antennas
    By sircana in forum RC TECH
    Replies: 256
    Last Post: 9th June 2016, 06:16 PM
  2. servo talk
    By FPV-PLASTICS in forum RC TECH
    Replies: 14
    Last Post: 9th September 2012, 01:12 AM
  3. Replies: 3
    Last Post: 22nd May 2012, 03:40 PM
  4. Replies: 0
    Last Post: 1st April 2012, 10:22 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •