There is a misconception in the RC community that receiver antenna length is critical. It is not. I believe this misconception results from the fact that transmitter antennas ARE length critical. One of the reasons for this misconception is that people read on the internet about tuning transmitter antennas, and assume it applies to receiver antennas as well. The actual science is quite involved, very math intensive, so few delve deep into it. Here's my attempt to explain what I know of it in a easy to understand framework. I hope it helps out. Transmitter antennas radiate a certain portion of the power fed to them, and reflect the balance back to the transmitter - this is why operating a transmitter with the wrong or without an antenna can damage it, nearly all the power is reflected back to the output transistor and it overheats. Some of the power is transmitted and some is reflected back to the transmitter. This forms a ratio. This ratio is known as the Standing Wave Ratio (SWR) which is a ratio of power sent to the antenna (forward) vs the power radiated by the antenna (1:1 being ideal - where all the power is radiated). SWR is determined by measuring the power being reflected back from the antenna and comparing it to the power being sent to the antenna (Reverse : Forward). An example would be a system with a SWR of 1.2:1 and a transmitter power of 1200mw. 1000mw will be radiated, and 200mw is returned to the transmitter and generates heat. If the antenna is tuned perfectly, with a 1:1 SWR, all 1200mw will be transmitted. This is a compromise on tunable systems, as the antenna is tuned for a specific length, and is only perfect (or nearly so) at one specific frequency. This antenna tuning can be achieved by cutting the antenna to a resonant length, but can also be done by matching it to the antenna with a inductor/capacitor network. Known in HAM circles as a “Transmatch” this network allows a single antenna to be used with a very wide range of transmitted frequencies and provides very good performance. It uses variable inductors and capacitors and is re-tuned for each desired frequency. A small specific version of this network (with fixed components) is often integrated into small transmitters to make them work better with less than perfect antennas. This is a simplification for the sake of brevity (like I warned about) and any time a a signal 'changes hands' some of it is as accepted and some reflected (which is why I went into all that detail above). This is all about matching impedances, and that's a big subject, worthy of entire books. Here's why it matters: In any case, on a receiver antenna, the only concern it that it draws the signals you want from the air as efficiently as possible. In doing so, it draws in virtually all radio signals, with the length determining which are best received. THIS is why we care about the length of a receiver antenna. It's a broad range, not a specific channel, or even a band (such as they are with our 5.8ghz video signals), covering many MHz. We finally get to why the length matters: The receiver uses a series of tuned circuits to eliminate the unwanted adjacent signals, and focus on the desired channels. The 1st stage of this process occurs where the antenna meets the receiver (IE “front end”). To remove the bulk of the unwanted radio energy, the receiver works with the antenna to create an impedance mismatch at frequencies above and below the desired target. This forms a broad tuned circuit, and reflects a large portion of the unwanted signals back to the antenna, where it is re-radiated (disposed of). The desired signal is thus 'pre-cleaned' allowing more sophisticated circuit to further refine it. The antenna plays as part of this tuned circuit, and contributes it's capacitance and inductance (components of impedance) the the components in the receiver circuitry. It's length effects it's impedance, and how it interacts with the receiver. This is WHY antennas will vary in length on different receivers, and why it helps, but is NOT critical to have them 'just right'. Each receiver has it's own specific front end impedance, and wants a antenna that works with it to produce the best performance. I apologize for the length of this comment, and have to qualify that I had to simplify some details to keep it this size, so don't beat me up if there's some minor inaccuracy or imprecision in it somewhere. I tried to make the basic principles easy to understand, and hope it helps out.
Scott Marshall there was only one model that was a receiver. All the other units were transceivers. For telemetry to work the receiver switches to a transmitter in between packets to send telemetry. Therefore matching impedance is important during the telemetry cycle. This was not specifically tested.
The runs are so short and the wavelength @ 5.GHz both work to make it effects minimal for our uses, and it can be disregarded. In long cable runs as in RF CATV it's the bane of engineers trying to keep the system in phase. I wrote a long explanation of multipath and analog TV ghosting, but I spared you the gory details. (I tend to run on on these sort of subjects) Short answer - NO, The long answer is too long for this section. If you want the gory details, PM or email me.
This is pretty accurate, but there's one thing you haven't mentioned: the sensitivity to resonant frequency shifts, both on the TX and RX side, is inversely proportional to the bandwidth of the antenna. If you have an extremely narrowband antenna, a slight change in center frequency can produce so much mismatch that you get essentially no signal at all. A wideband antenna is far less sensitive to this. Dipoles and cloverleafs have a relatively wide bandwidth, which is why you won't see that much difference. But some other antennas such as PCB-based patch antennas without an air gap or miniaturized ('meander') antennas (especially PCB-based) can become completely nonfunctional. Also, in terms of signal strength, mismatch at the RX and TX side has the exact same effect. One is no more critical than the other. The difference between RX and TX matching lies only in the fact that the TX can be damaged by a mismatched antenna whereas the RX can't. Variable matching networks are popular for amateur radio which uses lower frequencies, but they aren't practical for high frequencies such as 2.45 GHz or 5.8 GHz. We get away with this because the transmitters that are used for FPV are relatively low power compared to those that are used in amateur radio, so they don't overheat that easily.
Great testing Joshua! I'm a young RF enthousiast, and I have access to a network analyzer that goes up to 3GHz. I was thinking, since all these antennas have U.FL connectors, I can just hook one up to the VNA and see at what frequency it resonates. The resonant frequency of these antennas are dependent on the length of the exposed antenna wire (outside of the coax) but the size and location of the RF ground will also be important for antenna tuning. One thing that is also interesting to note, is that the receiver chips cannot do any antenna tuning, the only option is that they put passive components at the feedpoint of the antenna, to change the antenna impedance. However, any time you are tuning an antenna with inductors or capacitors, the components MUST be right before the radiating side of the antenna. You cannot really tune an antenna at the feedpoint of the coax cable, since the 50-ohm coax will transform the impedance, which is dependant on the coax length. I think I'm going to try to hook up my VNA to one of these 'whip' antennas, and I also want to see if I can see any matching components on the PCB. The best test to measure the antenna resonant frequency would be to use hot air to remove the RF chip on the receiver, and solder a tiny coax cable directly on the pads where the signal to the antenna goes. Then you can perform a calibration, and actually measure the impedance that the chip is 'seeing' with all the RF ground and matching components on the PCB.
Please do. I recently lost TX to a quad running a new R-XSR. It could have been anything but suddenly dropped out of the sky before the "Telemetry lost" warning. I never found the quad to do an investigation. I've looked through ham forums but found zilch apart from 5/8λ Vs 1/4λ - 1/2λ
Well, my plan went to hell. The U.FL connector on my Frsky receiver are actually not standard U.FL, but some sort of micro U.FL. So I cannot connect the antenna to my VNA without cutting off the connector and soldering the coax directly on a connector I can connect to my VNA. However, this could in theory mess up any test results. Not really sure how to proceed.
I think I'll do a video on this but basically there are many factors that influence those antennas and it's not just "tuning" but also radiation pattern, impedance and other things which determine how the antenna will actually perform. Joshua's test environment had too many variables to get consistent and therefore accurate results so I'll try to come up with something that mitigates (as far as possible) those variables. It's an interesting subject and one that's worth more discussion. I'll try to get my vid done tomorrow.
Looking forward to it - while you are testing - I'd be really curious to see if there is any point in using something like this over the stock antenna: www.amazon.com/Irctek-Connector-Omnidirectional-Replacement-Receiver/dp/B01L7IW6VE
I'm looking forward to this too. When you do, cover the topic of the length of the antenna and ground plane. www.flitetest.com/articles/taranis-upgrade-speaker-shielding-sma-connector This is my article that covers the lengths of the Taranis antenna and the 5 dBi replacement... Enjoy
In the last 30 minutes between watching your video and reading the comments I've learned more about antennas then I've ever wanted to know. People use to pay for this type of information! Thanks for your time in all the work that you do to promote the hobby and teach everyone about the facts you've uncovered.
As a layman I conclude two things from these tests. 1: the engineers at FrSky seem to know what they're doing when they design these antennas and receivers. 2: no matter which receiver you get, unless you're going for super long range or something, you're going to have a very solid reliable signal. I was slightly concerned when I bought the R-Xsr that it may have less range than a full size receiver, but it seems I was worried for nothing!
Hi Joshua, again a very cool video from you. I have to thank you for all your content, which made me learn a lot! This time I think I can give you some informations: To get the best reception from an existing combination of components, the length of the antenna-"core" (unshielded part) needs to be perfekt for a combination of: actual wave-speed, frequenz, antennas environment, ... I would advise you to: put the TX far away ... if you like in your car ... and you stay at your quad. Yes I saied quad, because of the antennas environment! you want it to work in you quad, so tune it there ! As someone in the comments already saied, your ground-part (which is also the carbon of your frame) of the antenna and anything in its neighborhood will change its behavior. So use the place where it will have to work. The only thing you should change during the tuning is the length of the antenna. You changed the distance and as you have seen in your chart, you got hills and valleys. I'm shure this was because of interferences. You got bounced signals on from what ever which de- or increased you signal. Don't use several recievers, there can be so much different ( quality of rf-chip, antenna-matching, groundarea ...) Also, the recievers working next to each other can change there behavior. So put the TX in the car far away, go into the house with your quad and your goggles ( need to see rssi in the osd !). Put the quad somewhere where it can't move, but it should sit on something similar to air, because your quad normaly is in the air ... I usally use styrofoam, this has nearly no effect on the resonance and matching. Use only one of the antenna cables, disconnect the other. To get great range of frequency in your test, use a antenna with a very long core. measure the antennacore-length, step back from the quad, read the RSSI ... put values into chart .. cut a little piece of the antenna, as small as possible (I like to use a fingernail clipper) measure the antennacore-length, step back , read the RSSI ... put values into chart .. cut a little piece . . . everything should be the same, even your position while reading the RSSI You will get a nice graph! greetings, Max
Any time I have ever chopped or messed up an antenna, or as soon as I get a new antenna, I always measure it myself. It’s simple, use this formula to expose the proper amount of the inner wire. 300 divided by your radio frequency, then divide that by 4 to give you a 1/4 wave length. I typically try to use 2.44. This hits in the middle of the frequency and seems to work best for me. I’ve seen so many ppl who do not peel back the coating, and mesh to expose the inner wire and they think its a bad receiver when its simply not peeled back.
Thank you for taking the time to do these types of tests. I don't currently run FrSky, but its nice to know that you have done the actual work behind figuring this out and might help in the future if I go to FrSky. Great video as always!
Alex knows his stuff; be been keeping up with him 6 years now for his antenneas.i use blue beam exclusively.Thank u for this video I like your approach.
Thank you Josh, we all appreciate the time and effort you put into all of your videos. I watch them regularly and look forward to new ones just because I know I am going to learn something :) I know most people might question why you do some of these videos, but I wish I had the time and knowledge you did on the subject because I wonder these things often but am unsure how to answer them. Thank you for putting into words some of the answers we wonder when no one is looking :)
I hate to admit it, but my new favorite go-to receiver is the XSR-E (the Banggood Eachine/Frsky collaboration). It's damn near the same size as an XM+, but has telemetry. Also, according to my sketchy/not quite Bardwell-quality field testing, it performs almost identical to my old go-to X4R SB's. I have the XSR-E's on 7-8 models right now, and they're performing great!
Nice tests! How ever, I'm a ham radio operator and I once had a chat with an engineer who said "Don't bother with the antenna. Design the apparatus and match the antenna after." Antenna is irrelevant at these frequencies (With in limits) The matching is crucial. Keep up the great work, buddy!!
In my company we design and test antennas as a part of daily development. You are certainly right that the tune of the antenna comes not only from conductor length. Type of surrounding materials, including the "plastic" shroud, distance to all objects in near field (first wavelength), and antenna material itself can all affect how well the antena is tuned to certain frequencies, heck - even the quality of solder joints can. The tune itself is not a 0/1 factor. There are all possible shades of gray between optimal tune to a frequency and ignoring it all together. Moreover you can have a decently tuned antenna using certain fractions of a wavelength physical size. I'd assume that manufacturers test and use the right combination of length and materials to get desired results. To do those tests right you'd need to use only one receiver, cause differences between them can be too big to see changes made by the antenna, and be very precise with physical distance and orientation between tested devices, avoid multi path transmission as much as possible etc. Short story - You are right, but you'd probably see something consistent only in an anechoic chamber.
You need more distance to see the real results. Stinger swarm did this same test on RotorRiot and his results were inconclusive due to his cat chewing his antennas haha. But from what I can see is that even in Range mode, you need a lot more distance then you have in your yard to get any solid results from the data, they were in a pretty big field. Maybe you could use a dummy load to cut down the range? I was thinking of doing the same test on a dry lake bed here in California, but have since switched to Crossfire, so there is no need. Great video, very informative!
Hey JB. Good to see that somebody, atleast ;), tries to get some reliable data out of this. After everybody was going crazy for the 5dbi mod and it's range extension, I did some tests on my own using the Taranis logging. Unfortunately the results between 2dbi and 5dbi where so close that you couldnt tell a difference in just range - and until now, no one seem to investigate it and providing detailed data like in this test. You seem to stick to the stock antenna with your x9D SE, but maybe there will be a science-like-validation by you at some point. Always appreciate your content and way to present and explain stuff!
Nice test. I’m interested in when the FC sees a failsafe based on range. It would be nice to have a FC logging to blackbox RSSI values and when a Failsafe occurs. Add a beeper for the fun of it and you walk until and audible tells you a failsafe took place and put a flag in the ground.
SUGGESTION! How about use one receiver. strip the antenna to a long length (say 4or5mm) walk a range test, then trim 1mm walk the same range test, repeat the test taking 1mm off until the antenna each time record and share the results 😄. P.S. I love your goats 🐐😜
This might be picking a nit but the worst reception would be if the RX antennae were not end on but horizontal. In this case your transmitter signal is vertically polarized. A horizontal RX antenna would have the lowest received signal. A good reason for antenna diversity , mount one antenna horizontal and one vertical. Of course microwave signals tend to rotate over distance so this may not be true in this case.BTW consider the lowly AM radio that operates in the MF band, the antenna is invariable coiled up on a ferrite bar and has no external wire. A long wire AM radio antenna would be about 200 meters long.
I am new to the hobby (Loving it by the way), and I am also an active Ham Radio operator. With Radio frequencies, antennas perform the best if they are trimmed for a specific frequency in a band. So for example with the 2-meter band in ham radio (144 to 148 MHz) you would make the antenna 1/4 or 1/2 wavelength to 146 MHz. The antenna will be usable on the whole 2-meter band, but will still be most efficient on 146 MHz, so in the case of FPV, technically those antennas will perform best on whatever frequency (Channel) that the antenna is actually trimmed for. So if you know the frequency you always use you can trim them for it. I enjoy your videos. You sold do similar tests on all of the different channels to find out on which channel each antenna is most resonate. Impedance and radiation pattern also affect it like xjet said. I do think it would be interesting to test them all throughout the band though. I am sure that then each one would perform better on different channels.
Ah okay. I was just curious. Like I said I am new to the RC and FPV community but have been in Ham radio for a few years. With Ham Radio though I have only worked in MHz and have not done much with digital. Keep up the good work. I enjoy your videos and learn a lot. Most channels I find on here besides yours and UAVFeaturs are nothing but VLOG or Freestyle. I do enjoy them, but don't learn much at all. Thanks for what you do, and for explaining.
My best signal stability and range have come from running a pair of sleeve dipole antennas on my xm+ receivers. Just standard wifi rubber ducky antennas with the housing removed and direct soldered to the Rx. compared to stock antenna lengths, 1/4 wave re-trimmed, or anything else, they perform much more stably. I got the idea from an rcgroups page on testing the taranis range with different antenna setups, and I'll stick with it for my freestyle quads, since they're the only ones that I ever really push the range with anyway.
I have them shrink sleeved to big-ass zip ties, sticking out the back end of my quad, haven't had a problem yet at least! They use rg316 coax for the leads, so they're a fair bit tougher in that regard, the only part I'd be concerned with is the metal ground plane thing getting dinged up
HI, What a nice test, love it. I think the way of your test is fantastic. And especially the energy that you put into it. But I always thought that the antenna length of the transmitter was more critical. I may see that wrong. Again, love your approach and the way you attack a Challenge!
Thank you for the wicked analogy about radio frequency vs actual frequency of updates we throw out into the airways. helped me understand the tuning a hint more.
oh as for longer range while on that pole? A) heaps above the ground plane compared to when we test in a quad, and B) heaps less RF noise due to no carbon fibre surrounding.
Hey Joshua, I know that its pointless for multirotors to say, but it prooves my point (and fun-fact). Ive stopped trying to find out the range of these guys a long time ago. I gave up :) First one does not support SBUS but "my friend" was flying 3.3km long range with old v8r7-ii. Other friend tried the "real" test, left transmitter at home and went hiking to near mountain - he did nearly 6km when he gave up. (X8R using SBUS). Not mentioning that our EURO version of taranis is on 100mW (not sure if 100 but I know is less then US). I guess these newer guys will stick to that performance.
Don't forget that these receivers transmit (sending RSSI data) as well as receive, so undoubtedly the manufacture will have tuned the antenna length to match the RF output stage, this will usually give a length that is not a factor of wavelength, for example a capacitive output they will have needed to lengthen the antenna and if inductive shorten it.For receive only the length is nominally the same as for transmit and a general rule of thumb for receive only is usually 95% of the wavelength. The antenna length is best left as the manufacturer / designer set it.
I used to work in an office next door to an rf guy. He tuned his antennas by connecting them to am rf network analyzer, holding his tongue a little to the left, and then with his side cutters, cut a few molecules off the end of the wire.
Thanks, yes some of us watched the whole video. Ok maybe I am a nerd engineer, whatever. I do like the attempt you made doing this. I think what we learned was that as long as your antenna is connected and the exposed part 28, 32, 33mm or whatever is not totally chopped off, it is going to work for most of our needs. These are not space shuttle o'rings, we are not sending our quads to Mars.
Good test - I was wondering this too. The thing is, that the monopole antennas does not have good VSWR in first place, so a few milimeters does not matter. If you want a good antenna you need a dipole.
I just had an image of JB in a cloak and hat like Gandalf, carrying around his staff of FRSKY receivers. "You wouldn't deny an old man his staff, hmm?"
A great video! With 20:20 hindsight you should have included two identical receiver/antenna combinations to compare their differences with the differences between different receivers and antennae. That's just me being a smartass after the event.
thanks for going through with this, i tried to grasp this stuff right when i started flying toys, using 3/4 length transmitter and half wavelength receivers... its a really complex topic to even try and reduce to laymans terms,. iforce2d had very similar outcomes to you with flysky set ups, using the 26mm and the 28mm? sleeved dipole version versus a strict 31.125mm.. he said if anything the stock antennae were the best, he tested in open countryside with actual return to home quads. so all we can do is go with the manufacturers choices, they seem to know best....
I have seen more than one, genuine XM Plus do in excess of 5.5kms with 70% RSSI still showing on the OSD of a Fixed Wing...I think the fact it does NOT have telemetry is what gives it an advantage...The long range FrSky L9R also does NOT have telemetry. Had mine out beyond 3 kms and the RSSI barely dropped with its antenna amplifiers. The noise floor in a Fixed Wing scenario, is much lower than a Quad of course, even so, it's a great advert for the (Tiny) XM Plus, popular with Mini Quads today. The other thing to note, and you can easily prove this yourself, is that separating the antennas on a typical telemetry capable receiver will drastically increase its range. We did some ground tests and proved this beyond doubt. The antenna placement in your clip is way too close together for a telemetry capable receiver. Goat one says to Goat 2, WTF is he doing with that big stick then..?....lol.
Joshua, the XM+ does have rssi, it broadcasts it on a channel and works wonderfully. I've actually flown the longest distance with an XM+ compared to a stock x4r-sb and xsr. It's too bad you didn't get that in the video but very cool video anyways!
Joshua Bardwell ahh, I didn't realise there wasn't a flight controller. I can understand why you had trouble walking so far and still didn't have signal drop, I put my taranis in range mode and flew 300 meters with rssi still in the 70s/80s. Are you going to do a video on frskys new telemetry protocol :D?
Depends on where you bought it. If it comes from China most likely it has stock firmware. EU and US stock sometimes come with it already flashed with rssi firmware. I've bought several now from China and US stock and some have RSSI working out of the box without flashing the firmware. I always flash them anyways just so I know what firmware is installed.
Nice test, but maybe should be done on open field with various receivers of each model because there is lot of bad FrSky receivers. Your job is great and I understand the financial limits of that kind of test. I get the best results with R-XSR and redundancy xm+ and changing antennas with replacement ones. I fly more than 2km without any warning. Of course that doesn't mean nothing but it's good for me.
Good idea for testing this, and thanks for doing it, but the methodology I don't understand. Why cross-test different receivers and different antenna lengths in random associations? It makes it so much more difficult to isolate the effect of the antenna length. I would focus on the receivers where it's easy to get rssi value. For each of those (there is only 3), test 3 different antenna lengths and record rssi for each combination. Then you have 9 curves that you can compare. You can see, for each receiver, which antenna length works best, and you can maybe then see a trend towards one antenna length giving better range in general.
I tested different receivers because most people who report short range are using a receiver-and-antenna combination. So the claim is not just, the shorter antenna on the R-XSR is worse. The claim is, the shorter antenna AND the R-XSR is worse. So I start by testing the actual claim that is being made: is one receiver+antenna combo worse than another? Then from there, I branch out to explore the effect of different antennas. Which is why in the 2nd test I had both a 1/4 and a 1/5 wave antenna on the same receiver (R-XSR).
Alright, I now understand a bit better the thought process. But in the end, the video's title question is not answered: "wrong antenna length?" With this method, no way of knowing if the receivers' hardware or antennas are at fault. Sorry for being picky, I'm an engineer too, after all :)
I think what you shod test is antenna placement on the Quad. and then walk away with the Quad. To many is stuffing the antenna in there and having problems. This works great: www.thingiverse.com/thing:2147201
My hypothesis: EMI decreases the reception range. The reason you are getting such great range - compared to what you expect - is probably the lack of noisy electronics in proximity to the receiver.
Same experience here... I have 3 XSRs and experimented with different antenna lengths. Bone stock seems to work best and better than my two X4R SBs. I have two R-XSRs sitting on my bench for new builds - lets see how this goes.
Good Test ! They all seem pretty close .......... Have you tried different antennas on One Receiver unit ? ........ or different Transmitter antennas ? Also increasing your height was interesting ........ it would be beneficial to know the optimum height for your TX antenna . Great show !
I think the antenna location on the craft is as important as antenna length. Personally, I have the antennas coming out of the bottom of my quads-planes, for the simple reason that they are flying above me and if the antennas are on the bottom of the craft there is a line of sight with no obstructions, back to my transmitter. Always place the antennas so they are no obstructions between you and the craft that you are flying to get max range.
We have similar discussions in our german forums without clear solutions too... One short tip maybe you test it. If i shred a antenna i removed the jacket and then i evert the helical ground wire. Its then like a dipol told me an skilled "antennaguy". Functioned for me but not range tested because the Video Signal was the limited factor.
All of the antenna wires are about the same (wire type), not sure your long velocity factor analogy applies here. There's still quite a few variables (as you mentioned), antenna length, polarization, possible lumped elements changing the tune of the antenna, and SNR differences in the receivers, overall great video as usual
It may have been a slip of the tongue Joshua but loss of telemetry is not the same as loss of control link. This seems to be a common misconception among users and a point you could help clarify in one of your videos. This can be demonstrated when the TX is too close and the radio alternately reports Tele lost Tele recovered and yet all the while you have perfect control. FrSky are actually taking a few liberties with the RX antennas since even with a perfect 1/4 wave radiator length the whisker design has no counterpoise element to complete the dipole leading to induced current in the sheild. FlySky on the other hand often fit their RX's with outer sleeves which acts as the second half of a 1/2 wave dipole and also acts to decouple the shield which as you know is a much better design choice. In fact It may be interesting to fit a pair of those sleeved dipoles on the FrSky RX's and compare to the FrSky whiskers. One last thing PLEASE start complaining long and loud about the new style antenna coax and IPX connectors FrSky are now using. Hopefully they will take note and revert to the older style coax and uFL's. It would seem in striving for ultra miniaturisation they have forgotten the real world implications and practicalities of (to use the Australian vernacular) Piss weak coax.
Great effort trying to figure this out. With my time spent in CB radio hobby before, my experience is that the antenna length vs range is not that critical for receivers than transmitters. I.e. If you got the a bad antenna, say a wrong length antenna, even using a bare wire cut casually will still receive considerbly good. However the same antenna if used in a transmitter, can have high SWR and thus affect range considerably at the receiver side. Regarding to why the antenna length is different for different receivers, I AM PURELY GUESSING HERE, are they related to the telemetry transmission freq of these receivers? Since these receivers are with telemetry, meaning that these antennas are also used as transmitter antenna, is Frsky use a length that is particularly tuned to the freq(s) used by these different receivers? If that is true, then we should see fixed, say 32mm antenna for all non-telemetry receivers caused they are just for receiving ..... hmmm, looking forward for your further reseach into it.
Makes me wonder what is the difference in using numbers in between the 23mm to 33mm. Meaning is there any difference between 23,24,25 and so on or if there are specific number that make the antenna good and just so happens that the 23,26,32 you tested is about there... Great video and its awesome that you put all this thought on questions people have but most dont bother to investigate!!!
33/23 is approx 1.41 (or square root of 2). On a pure sine wave it's the RMS to Peak ratio - 1:1.41. On a sine wave carrier + signal they probably tweak this number for tuning. Complicated maths involved? Yes. Do I understand it all? 'Fraid not!
Had to jump back in time... I placed an R-xsr in my newest build for Lua and such, problem is I failsafed in a spot that I had no issues with my other build using Xm+. I noticed the length of the R-xsr antenna. It is much shorter. So I ordered some after market antennas to test. Thank you for this video! Seems to explain why this crash happened.. ScooterPie
Let the goats help ! I think it would have been good to put a 1/3 wavelength antenna in there and maybe even a no-antenna receiver in there just for a control.
Also dont forget that frsky receiver work as transmitter for telemetry and this antenna must be matched by SWR. Change length may degrade telemetry performance.
Worth mentioning that an antenna for receiving only is not as critical as an antenna that broadcasts. In a badly adapted antenna, the radio energy will bounce back into the transmitter, and thus the range will decrease. That would be interesting to measure swr on the counterparts with telemetry.
You've got to love how RC hobbyist with no background in RF electronics think that because they read a few blogs that they now know more then the RF engineers who designed the system. The only proper way to tune an antenna like these is with a VNA but then you would actually need to know how to read a smith chart LOL .... Kep in mind that these are not simple receivers but transceivers that send back telemetry so they antenna is tuned for the transmit signal not the RX signal.. My advice is unless you have a VNA and know how to use it then you have no business cutting your antennas.
I have to say that the stripped area of the frsky antenna is not the only active part. The entire length is the working area. I don’t know why they even strip the end, radio waves go right through plastic so there is no need to strip it. I don’t think that is coaxial wire, I may be wrong, I’ve looked at it closely. Maybe they strip it so we will wear it out and buy more? Also those antenna pairs should all be at 90 degrees to each other.
I think you're mistaken about this. It is absolutely coaxial cable. It has a braided grounding shield along the length, which is why the exposed section radiates and the shielded section does not.
I have had 2 different XSR receivers bought from two different vendors and both times these receivers have given me RSSI warnings at 150-200 feet. I'd also have failsafe triggers on occasion. I have tried these XSR on different builds with different Flight Controllers to see if it makes a difference. It worked the same. I switched over to a XM+ with the 1/4 wave antennas to see if it worked better, and now I can easily go 500-600 feet and still have 70% RSSI. I don't know if the antenna is the difference or not, but I have no faith in the XSR receivers. I might try to pop on some 1/4 wave antennas on my old XSR to see if it makes a difference. It's possible I got two dud receivers, but I find it hard to believe it could happen 2 times from two different vendors. I am going to try the R-XSR with swapped out 1/4 wave antennas to see if it will work better. I really like having Telemetry, but the XM+ has RSSI on channel 16 which is sufficient as long as I am paying attention to the value in my googles, however having Telemetry and the Taranis calling out warings is much preferred.
You mentioned in an earlier reply that you cut your antennas to 33mm. I thought 1/4 wave was 31.25mm. Also, the X4R-SB was at 33.2mm and the XM+ was 32.7mm according to the video. That's odd. I guess all the receivers are tuned like Joshua suggests, but if you say 33mm worked for your XSRs, I will give it a shot.
Thanks for your testing but you haven't tried an exact 1/4 wavelength. I would use one rx with a short antenna and measure range , then change to a long antenna on the same rx and check range and then a 1/4 wave antenna on same rx and check range.
I have very good experience with XM+ stock antennas (almost 1km of solid signal). XSR with stock antennas was much worse (around 700m when it already started giving me alerts). XSR with aftermarket antennas from banggood worked better but still lacking behind the XM+.
All 2nd generation of Tinyhawks including the Freestyle have RSSI as high as 99, while the older Tinyhawks maxes at 70. I wonder what Emax did to tune it. But even with the 30% difference, the distance was only 5 meters before failsafe, but it seems the new versions have better consistency in holding connections within range. The only thing I noticed is the new antennas on the 2nd generation TH are thicker and stiffer.
I watched Stingy and Steele test this in real flight test. An XSR with 32mm mod outperformed x4R with 32mm stock, which both whooped an xsr with 23mm stock, therefore I run a modded XSR.
+gabbo2027 no offense to them but they did one test. Do you know that your antennas are the same coax as theirs, with the same velocity factor? Just one reason why multiple tests are valuable.
8:13 fu#$% you touched my heart there joshua haha nicest self advertising ever all the way!! I just bought 2 parallel chargers from your brand, and also i am thi king in patreon you, bjt first i have to understnd how does that work, you helped me so much in the hobby ... i trully owe you a beer at least! Thanks for all mate
Nominal Velocity of Propagation is the key to your shorter antenna. Also interesting: Most of these receiver antennas are only a half dipole. So if you measure 1/4 length of unshielded wire it might be actually 1/2 wavelength antenna design but its only half of the antenna present.
Joshua, did you test using a standard Taranis antenna or did you use an upgraded antenna? I am trying to decide if it's worth modding my Taranis to upgrade the coax cable and use a higher dbi antenna to get more range.
Joshua, you are wrong about VF (Velocity Factor) of the coaxial cables. The formula is VF=1/SQRT(K), where K is dielectric constant of the central insulator. For example, if it is Polyethylene, then VF=1/SQRT(2.3)= 0.65938..., but we can round it to 0.66, or 66% of speed of light. If the same polyethylene has air bubbles inside in form of foam, then it can't be calculated because then K is between 1 and 2.3, for which only manufacturer can provide you the right information. VF for antenna element itself depend of ratio between wavelength and diameter of the radiating element and it is NEVER 1, only by math. Usually it is around 0.95, or 95% speed of light. The thicker the radiating material - the shorter antenna should be. But if radiating element is too thick, although it will be tuned, its Q factor (Quality factor) will be low, thus gain will be low as well. And, you can't measure antennas that way - by placing more antennas next each to other. Also, 'whisker' antennas are bad - it need at least 'sleeve' over the coaxial cable, and will be good that it has the other end of the dipole.
Make antennas like this, and you will be VERY surprised how much range you will get. :wildlab.org/index.php/2016/03/02/tutorial-how-to-increase-range-of-your-quadcopter/
Did you test with a standard taranis also? Wouldn’t the coax you put in that, improve your signal? Didn’t you have a previous video showing that you swapped out the aftermarket coax cable within the taranis to a even better/thicker one? Great video as always!
I would like your opinion Joshua. I am having range problems with my XSR receiver. The receiver is new and I have a Taranis X9D Plus. The radio works fine with other receivers, only this one has range problems. How could I improve that? Is that a receiver problem? Tks a lot!
Thank you Josh, I appreciate your videos and tests. Silly question, have you ever used a rubber coated paper clip for antennae? I have tons of questions actually, I'm wondering what the pitch/gauge of the antenna changes about the signal efficiency as well as wether or not a wispy versus firm antenna matters at all either. :) I want more range for my drone and the vtx without having to double the literal strength and diverting it with anything aftermarket. (who knows, there may be a way with math on our side)
I've been messing with this for awhile after noticing on my xsr the RSSI was fluctuating a lot. So I tried putting new 31.2mm then shortened those antennas to 29.5mm then 26mm. What I noticed was the 29.5mm length for the antenna seemed to have a more stable RSSI then the the other 2 lengths and I have been able to fly in places where normally I would have failsafed. I think it is cause it is about 95% the wave length and copper wire has a velocity factor of about 0.95. My best explanation is that the antenna needs shorter to compensate for the difference in speed between charges in the wire and the radio waves interacting with the wire.
Sometimes science tells us that we aren't asking the right questions. The important take away is that everything seems very forgiving. Try it with a spektrum and you'll be done in an hour.
Nice testing. I'm not sure that I would expect RSSI between different receivers to be calibrated the same. I think you need to test to failure in a straight line test - seems like that requires a longer runway. :-)
The "low RSSI" and "critical RSSI" warning level is defined on the Taranis at the same level for each receiver.... Anyway, in the 2nd test I was not reading RSSI values but physically watching for the LED to indicate a failsafe. That's functionally equivalent to testing to failure. The moment the RX light goes red, that's when the quad would fall out of the sky.
Thanks for all the work you put into your videos, but I want to learn more today! I just bought the Kopis 1 (love it) and have a question about mounting the receiver antennas. I was able to squeeze in a spektrum SPM4649T in there, but due to wiring restrictions I ended up with the antennas pointing towards the back from the front arms. You gave directions to mount them pointing to the front from the back arms. Can you tell me if that matters. I can make a longer harness to flip the receiver around if you think there will be an improvement. And could you explain why it matters if they point to the back or front? Please and thank you from an appreciative viewer.
I have a somewhat relevant question: Why do you mount your Rx antennas on your quads to back motor arms facing forward? Why not front arms facing backwards? Wouldn't that minimize the chances of the antennas acting like anchors in a crash and thus decrease the chances of damaging the antennas?
VilleF1N when you mount youre antennes facing forwards regardless on which arm. You will have the worst link, which is a good thing because then youre certain of a good link when you return.. jn stead of flying to that point of bad link, you turn around, and loose link completely..
There's a wire length coefficient in the circuitry of every receiver. Otherwise, I wouldn't be able to fly around with a completely disconnected RX antenna.
Looking at your first graph, would it be right to conclude that a stock XSR is better than an R-XSR reception wise ? The blue line is above the orange line only between 60-80ft. 20ft interval over 120ft total
None of these receivers use true dipole antennas. There is no ground plane. Therefore the receivers use the attached circuitry and coax length as the ground plane. This probably affects the results. Also, the antennas should probably be perfectly straight and vertical for repeatable measurements.
There is a misconception in the RC community that receiver antenna length is critical. It is not. I believe this misconception results from the fact that transmitter antennas ARE length critical.
One of the reasons for this misconception is that people read on the internet about tuning transmitter antennas, and assume it applies to receiver antennas as well. The actual science is quite involved, very math intensive, so few delve deep into it. Here's my attempt to explain what I know of it in a easy to understand framework. I hope it helps out.
Transmitter antennas radiate a certain portion of the power fed to them, and reflect the balance back to the transmitter - this is why operating a transmitter with the wrong or without an antenna can damage it, nearly all the power is reflected back to the output transistor and it overheats.
Some of the power is transmitted and some is reflected back to the transmitter. This forms a ratio.
This ratio is known as the Standing Wave Ratio (SWR) which is a ratio of power sent to the antenna (forward) vs the power radiated by the antenna (1:1 being ideal - where all the power is radiated). SWR is determined by measuring the power being reflected back from the antenna and comparing it to the power being sent to the antenna (Reverse : Forward).
An example would be a system with a SWR of 1.2:1 and a transmitter power of 1200mw. 1000mw will be radiated, and 200mw is returned to the transmitter and generates heat. If the antenna is tuned perfectly, with a 1:1 SWR, all 1200mw will be transmitted. This is a compromise on tunable systems, as the antenna is tuned for a specific length, and is only perfect (or nearly so) at one specific frequency.
This antenna tuning can be achieved by cutting the antenna to a resonant length, but can also be done by matching it to the antenna with a inductor/capacitor network. Known in HAM circles as a “Transmatch” this network allows a single antenna to be used with a very wide range of transmitted frequencies and provides very good performance. It uses variable inductors and capacitors and is re-tuned for each desired frequency. A small specific version of this network (with fixed components) is often integrated into small transmitters to make them work better with less than perfect antennas.
This is a simplification for the sake of brevity (like I warned about) and any time a a signal 'changes hands' some of it is as accepted and some reflected (which is why I went into all that detail above). This is all about matching impedances, and that's a big subject, worthy of entire books.
Here's why it matters:
In any case, on a receiver antenna, the only concern it that it draws the signals you want from the air as efficiently as possible. In doing so, it draws in virtually all radio signals, with the length determining which are best received. THIS is why we care about the length of a receiver antenna. It's a broad range, not a specific channel, or even a band (such as they are with our 5.8ghz video signals), covering many MHz.
We finally get to why the length matters:
The receiver uses a series of tuned circuits to eliminate the unwanted adjacent signals, and focus on the desired channels. The 1st stage of this process occurs where the antenna meets the receiver (IE “front end”).
To remove the bulk of the unwanted radio energy, the receiver works with the antenna to create an impedance mismatch at frequencies above and below the desired target. This forms a broad tuned circuit, and reflects a large portion of the unwanted signals back to the antenna, where it is re-radiated (disposed of). The desired signal is thus 'pre-cleaned' allowing more sophisticated circuit to further refine it. The antenna plays as part of this tuned circuit, and contributes it's capacitance and inductance (components of impedance) the the components in the receiver circuitry. It's length effects it's impedance, and how it interacts with the receiver.
This is WHY antennas will vary in length on different receivers, and why it helps, but is NOT critical to have them 'just right'. Each receiver has it's own specific front end impedance, and wants a antenna that works with it to produce the best performance.
I apologize for the length of this comment, and have to qualify that I had to simplify some details to keep it this size, so don't beat me up if there's some minor inaccuracy or imprecision in it somewhere. I tried to make the basic principles easy to understand, and hope it helps out.
Scott Marshall there was only one model that was a receiver. All the other units were transceivers. For telemetry to work the receiver switches to a transmitter in between packets to send telemetry. Therefore matching impedance is important during the telemetry cycle. This was not specifically tested.
Man! That was awesome! I'm so glad I bothered to read the comments. I'd always wondered what it was that killed VTX's that didn't have an antenna.
As an Extra class amateur radio operator, I applaud the clarity of your explanation. Well done.
The runs are so short and the wavelength @ 5.GHz both work to make it effects minimal for our uses, and it can be disregarded.
In long cable runs as in RF CATV it's the bane of engineers trying to keep the system in phase.
I wrote a long explanation of multipath and analog TV ghosting, but I spared you the gory details. (I tend to run on on these sort of subjects)
Short answer - NO,
The long answer is too long for this section.
If you want the gory details, PM or email me.
This is pretty accurate, but there's one thing you haven't mentioned: the sensitivity to resonant frequency shifts, both on the TX and RX side, is inversely proportional to the bandwidth of the antenna. If you have an extremely narrowband antenna, a slight change in center frequency can produce so much mismatch that you get essentially no signal at all. A wideband antenna is far less sensitive to this. Dipoles and cloverleafs have a relatively wide bandwidth, which is why you won't see that much difference. But some other antennas such as PCB-based patch antennas without an air gap or miniaturized ('meander') antennas (especially PCB-based) can become completely nonfunctional.
Also, in terms of signal strength, mismatch at the RX and TX side has the exact same effect. One is no more critical than the other. The difference between RX and TX matching lies only in the fact that the TX can be damaged by a mismatched antenna whereas the RX can't.
Variable matching networks are popular for amateur radio which uses lower frequencies, but they aren't practical for high frequencies such as 2.45 GHz or 5.8 GHz. We get away with this because the transmitters that are used for FPV are relatively low power compared to those that are used in amateur radio, so they don't overheat that easily.
Great testing Joshua! I'm a young RF enthousiast, and I have access to a network analyzer that goes up to 3GHz. I was thinking, since all these antennas have U.FL connectors, I can just hook one up to the VNA and see at what frequency it resonates. The resonant frequency of these antennas are dependent on the length of the exposed antenna wire (outside of the coax) but the size and location of the RF ground will also be important for antenna tuning.
One thing that is also interesting to note, is that the receiver chips cannot do any antenna tuning, the only option is that they put passive components at the feedpoint of the antenna, to change the antenna impedance. However, any time you are tuning an antenna with inductors or capacitors, the components MUST be right before the radiating side of the antenna. You cannot really tune an antenna at the feedpoint of the coax cable, since the 50-ohm coax will transform the impedance, which is dependant on the coax length.
I think I'm going to try to hook up my VNA to one of these 'whip' antennas, and I also want to see if I can see any matching components on the PCB. The best test to measure the antenna resonant frequency would be to use hot air to remove the RF chip on the receiver, and solder a tiny coax cable directly on the pads where the signal to the antenna goes. Then you can perform a calibration, and actually measure the impedance that the chip is 'seeing' with all the RF ground and matching components on the PCB.
Well tell me your results I'm dying to know!
I'll have time to do the measurements in a couple of hours, can I email the results to you? I'll be generating some plots and stuff..
Yes, and make a video out of it, please :-)
Please do. I recently lost TX to a quad running a new R-XSR. It could have been anything but suddenly dropped out of the sky before the "Telemetry lost" warning. I never found the quad to do an investigation. I've looked through ham forums but found zilch apart from 5/8λ Vs 1/4λ - 1/2λ
Well, my plan went to hell. The U.FL connector on my Frsky receiver are actually not standard U.FL, but some sort of micro U.FL. So I cannot connect the antenna to my VNA without cutting off the connector and soldering the coax directly on a connector I can connect to my VNA. However, this could in theory mess up any test results. Not really sure how to proceed.
I think I'll do a video on this but basically there are many factors that influence those antennas and it's not just "tuning" but also radiation pattern, impedance and other things which determine how the antenna will actually perform. Joshua's test environment had too many variables to get consistent and therefore accurate results so I'll try to come up with something that mitigates (as far as possible) those variables. It's an interesting subject and one that's worth more discussion. I'll try to get my vid done tomorrow.
Looking forward to it - while you are testing - I'd be really curious to see if there is any point in using something like this over the stock antenna: www.amazon.com/Irctek-Connector-Omnidirectional-Replacement-Receiver/dp/B01L7IW6VE
I'm looking forward to this too. When you do, cover the topic of the length of the antenna and ground plane.
www.flitetest.com/articles/taranis-upgrade-speaker-shielding-sma-connector
This is my article that covers the lengths of the Taranis antenna and the 5 dBi replacement... Enjoy
+xjet looking forward to your take on this!
Looking for this too uncle Bruce :)
But your 'news' videos are full of empty promises... so I won't hold my breath! Sorry to be so blunt Bruce.
In the last 30 minutes between watching your video and reading the comments I've learned more about antennas then I've ever wanted to know. People use to pay for this type of information! Thanks for your time in all the work that you do to promote the hobby and teach everyone about the facts you've uncovered.
As a layman I conclude two things from these tests. 1: the engineers at FrSky seem to know what they're doing when they design these antennas and receivers. 2: no matter which receiver you get, unless you're going for super long range or something, you're going to have a very solid reliable signal. I was slightly concerned when I bought the R-Xsr that it may have less range than a full size receiver, but it seems I was worried for nothing!
Hi Joshua,
again a very cool video from you. I have to thank you for all your content, which made me learn a lot!
This time I think I can give you some informations:
To get the best reception from an existing combination of components, the length of the antenna-"core" (unshielded part) needs to
be perfekt for a combination of:
actual wave-speed, frequenz, antennas environment, ...
I would advise you to:
put the TX far away ... if you like in your car ... and you stay at your quad.
Yes I saied quad, because of the antennas environment! you want it to work in you quad, so tune it there !
As someone in the comments already saied, your ground-part (which is also the carbon of your frame) of the
antenna and anything in its neighborhood will change its behavior.
So use the place where it will have to work.
The only thing you should change during the tuning is the length of the antenna.
You changed the distance and as you have seen in your chart, you got hills and valleys.
I'm shure this was because of interferences. You got bounced signals on from what ever which de- or increased you signal.
Don't use several recievers, there can be so much different ( quality of rf-chip, antenna-matching, groundarea ...)
Also, the recievers working next to each other can change there behavior.
So put the TX in the car far away, go into the house with your quad and your goggles ( need to see rssi in the osd !).
Put the quad somewhere where it can't move, but it should sit on something similar to air, because your quad normaly is in the air ...
I usally use styrofoam, this has nearly no effect on the resonance and matching.
Use only one of the antenna cables, disconnect the other.
To get great range of frequency in your test, use a antenna with a very long core.
measure the antennacore-length, step back from the quad, read the RSSI ... put values into chart ..
cut a little piece of the antenna, as small as possible (I like to use a fingernail clipper)
measure the antennacore-length, step back , read the RSSI ... put values into chart ..
cut a little piece
.
.
.
everything should be the same, even your position while reading the RSSI
You will get a nice graph!
greetings, Max
If I remember, Stingy and Steele did a test, and found that using the 31mm on the XSR gave the best results over the stock antenna.
Any time I have ever chopped or messed up an antenna, or as soon as I get a new antenna, I always measure it myself. It’s simple, use this formula to expose the proper amount of the inner wire. 300 divided by your radio frequency, then divide that by 4 to give you a 1/4 wave length. I typically try to use 2.44. This hits in the middle of the frequency and seems to work best for me. I’ve seen so many ppl who do not peel back the coating, and mesh to expose the inner wire and they think its a bad receiver when its simply not peeled back.
Thank you for taking the time to do these types of tests. I don't currently run FrSky, but its nice to know that you have done the actual work behind figuring this out and might help in the future if I go to FrSky. Great video as always!
12 and a half minutes into the video and I think you gave me the answer to my range issues. Thank you.. I think..😉👍
Alex knows his stuff; be been keeping up with him 6 years now for his antenneas.i use blue beam exclusively.Thank u for this video I like your approach.
Thank you Josh, we all appreciate the time and effort you put into all of your videos. I watch them regularly and look forward to new ones just because I know I am going to learn something :) I know most people might question why you do some of these videos, but I wish I had the time and knowledge you did on the subject because I wonder these things often but am unsure how to answer them. Thank you for putting into words some of the answers we wonder when no one is looking :)
I hate to admit it, but my new favorite go-to receiver is the XSR-E (the Banggood Eachine/Frsky collaboration). It's damn near the same size as an XM+, but has telemetry. Also, according to my sketchy/not quite Bardwell-quality field testing, it performs almost identical to my old go-to X4R SB's. I have the XSR-E's on 7-8 models right now, and they're performing great!
Nice tests! How ever, I'm a ham radio operator and I once had a chat with an engineer who said "Don't bother with the antenna. Design the apparatus and match the antenna after." Antenna is irrelevant at these frequencies (With in limits) The matching is crucial. Keep up the great work, buddy!!
Your testing is very very appreciated, you have spent hours making tests that not anyone can easily do. Thanks Joshua
Greetings from Italy
I asked about this maybe a year ago when you posted the FrSky antenna mod. This explanation was a bit more in depth. Thanks. :)
In my company we design and test antennas as a part of daily development. You are certainly right that the tune of the antenna comes not only from conductor length. Type of surrounding materials, including the "plastic" shroud, distance to all objects in near field (first wavelength), and antenna material itself can all affect how well the antena is tuned to certain frequencies, heck - even the quality of solder joints can. The tune itself is not a 0/1 factor. There are all possible shades of gray between optimal tune to a frequency and ignoring it all together. Moreover you can have a decently tuned antenna using certain fractions of a wavelength physical size. I'd assume that manufacturers test and use the right combination of length and materials to get desired results. To do those tests right you'd need to use only one receiver, cause differences between them can be too big to see changes made by the antenna, and be very precise with physical distance and orientation between tested devices, avoid multi path transmission as much as possible etc. Short story - You are right, but you'd probably see something consistent only in an anechoic chamber.
So for now... buy what you like and use that till more testing is completed. Got it. Thanks Jb
You need more distance to see the real results. Stinger swarm did this same test on RotorRiot and his results were inconclusive due to his cat chewing his antennas haha. But from what I can see is that even in Range mode, you need a lot more distance then you have in your yard to get any solid results from the data, they were in a pretty big field. Maybe you could use a dummy load to cut down the range? I was thinking of doing the same test on a dry lake bed here in California, but have since switched to Crossfire, so there is no need. Great video, very informative!
Hey JB.
Good to see that somebody, atleast ;), tries to get some reliable data out of this.
After everybody was going crazy for the 5dbi mod and it's range extension, I did some tests on my own using the Taranis logging.
Unfortunately the results between 2dbi and 5dbi where so close that you couldnt tell a difference in just range - and until now, no one seem to investigate it and providing detailed data like in this test.
You seem to stick to the stock antenna with your x9D SE, but maybe there will be a science-like-validation by you at some point.
Always appreciate your content and way to present and explain stuff!
Nice test. I’m interested in when the FC sees a failsafe based on range. It would be nice to have a FC logging to blackbox RSSI values and when a Failsafe occurs. Add a beeper for the fun of it and you walk until and audible tells you a failsafe took place and put a flag in the ground.
When the RX light goes red, that's a failsafe. But the channels would be periodically locking out earlier than that.
Thank you for going out and spearmint. it's people like you that we learn from thank you
SUGGESTION! How about use one receiver. strip the antenna to a long length (say 4or5mm) walk a range test, then trim 1mm walk the same range test, repeat the test taking 1mm off until the antenna each time record and share the results 😄. P.S. I love your goats 🐐😜
this question mark had been on my mind for a long time until just now. great video!
This might be picking a nit but the worst reception would be if the RX antennae were not end on but horizontal. In this case your transmitter signal is vertically polarized. A horizontal RX antenna would have the lowest received signal. A good reason for antenna diversity , mount one antenna horizontal and one vertical. Of course microwave signals tend to rotate over distance so this may not be true in this case.BTW consider the lowly AM radio that operates in the MF band, the antenna is invariable coiled up on a ferrite bar and has no external wire. A long wire AM radio antenna would be about 200 meters long.
I am new to the hobby (Loving it by the way), and I am also an active Ham Radio operator. With Radio frequencies, antennas perform the best if they are trimmed for a specific frequency in a band. So for example with the 2-meter band in ham radio (144 to 148 MHz) you would make the antenna 1/4 or 1/2 wavelength to 146 MHz. The antenna will be usable on the whole 2-meter band, but will still be most efficient on 146 MHz, so in the case of FPV, technically those antennas will perform best on whatever frequency (Channel) that the antenna is actually trimmed for. So if you know the frequency you always use you can trim them for it. I enjoy your videos. You sold do similar tests on all of the different channels to find out on which channel each antenna is most resonate. Impedance and radiation pattern also affect it like xjet said. I do think it would be interesting to test them all throughout the band though. I am sure that then each one would perform better on different channels.
The 2.4 GHz control systems we use are digital frequency-hopping. They're all over the band.
Ah okay. I was just curious. Like I said I am new to the RC and FPV community but have been in Ham radio for a few years. With Ham Radio though I have only worked in MHz and have not done much with digital. Keep up the good work. I enjoy your videos and learn a lot. Most channels I find on here besides yours and UAVFeaturs are nothing but VLOG or Freestyle. I do enjoy them, but don't learn much at all. Thanks for what you do, and for explaining.
Ive been wondering about this for a while. Thanks for doing this test!
I like the way you explain things. It makes it easy to understand.
My best signal stability and range have come from running a pair of sleeve dipole antennas on my xm+ receivers. Just standard wifi rubber ducky antennas with the housing removed and direct soldered to the Rx. compared to stock antenna lengths, 1/4 wave re-trimmed, or anything else, they perform much more stably. I got the idea from an rcgroups page on testing the taranis range with different antenna setups, and I'll stick with it for my freestyle quads, since they're the only ones that I ever really push the range with anyway.
I would be pretty suspicious of the durability on those.
I have them shrink sleeved to big-ass zip ties, sticking out the back end of my quad, haven't had a problem yet at least! They use rg316 coax for the leads, so they're a fair bit tougher in that regard, the only part I'd be concerned with is the metal ground plane thing getting dinged up
like always, I appreciate the time and sound science. Well done and 'I did learn something'. keep them coming.
HI,
What a nice test, love it. I think the way of your test is fantastic. And especially the energy that you put into it.
But I always thought that the antenna length of the transmitter was more critical.
I may see that wrong.
Again, love your approach and the way you attack a Challenge!
Thank you for the wicked analogy about radio frequency vs actual frequency of updates we throw out into the airways. helped me understand the tuning a hint more.
oh as for longer range while on that pole? A) heaps above the ground plane compared to when we test in a quad, and B) heaps less RF noise due to no carbon fibre surrounding.
Thank you for this great test. I know it took a whole bunch of time to prepare and analyse everything, but in the end we got a quite good answer!
Hey Joshua, I know that its pointless for multirotors to say, but it prooves my point (and fun-fact). Ive stopped trying to find out the range of these guys a long time ago. I gave up :) First one does not support SBUS but "my friend" was flying 3.3km long range with old v8r7-ii. Other friend tried the "real" test, left transmitter at home and went hiking to near mountain - he did nearly 6km when he gave up. (X8R using SBUS). Not mentioning that our EURO version of taranis is on 100mW (not sure if 100 but I know is less then US). I guess these newer guys will stick to that performance.
Don't forget that these receivers transmit (sending RSSI data) as well as receive, so undoubtedly the manufacture will have tuned the antenna length to match the RF output stage, this will usually give a length that is not a factor of wavelength, for example a capacitive output they will have needed to lengthen the antenna and if inductive shorten it.For receive only the length is nominally the same as for transmit and a general rule of thumb for receive only is usually 95% of the wavelength.
The antenna length is best left as the manufacturer / designer set it.
You know you’re doing good work when Martin Baert gets in on the comments section. Great work and an awesome video!
I used to work in an office next door to an rf guy. He tuned his antennas by connecting them to am rf network analyzer, holding his tongue a little to the left, and then with his side cutters, cut a few molecules off the end of the wire.
Thanks, yes some of us watched the whole video. Ok maybe I am a nerd engineer, whatever. I do like the attempt you made doing this. I think what we learned was that as long as your antenna is connected and the exposed part 28, 32, 33mm or whatever is not totally chopped off, it is going to work for most of our needs. These are not space shuttle o'rings, we are not sending our quads to Mars.
Good test - I was wondering this too. The thing is, that the monopole antennas does not have good VSWR in first place, so a few milimeters does not matter. If you want a good antenna you need a dipole.
I think this is the takeaway: that other things are more important to range than the length of the antenna.
I just had an image of JB in a cloak and hat like Gandalf, carrying around his staff of FRSKY receivers.
"You wouldn't deny an old man his staff, hmm?"
I'm never disappointed with your Videos!
A great video! With 20:20 hindsight you should have included two identical receiver/antenna combinations to compare their differences with the differences between different receivers and antennae. That's just me being a smartass after the event.
Good test Josh , keep doing these vids .
thanks for going through with this, i tried to grasp this stuff right when i started flying toys, using 3/4 length transmitter and half wavelength receivers... its a really complex topic to even try and reduce to laymans terms,. iforce2d had very similar outcomes to you with flysky set ups, using the 26mm and the 28mm? sleeved dipole version versus a strict 31.125mm.. he said if anything the stock antennae were the best, he tested in open countryside with actual return to home quads. so all we can do is go with the manufacturers choices, they seem to know best....
What should the quarter wave length be for 27mhz
I have seen more than one, genuine XM Plus do in excess of 5.5kms with 70% RSSI still showing on the OSD of a Fixed Wing...I think the fact it does NOT have telemetry is what gives it an advantage...The long range FrSky L9R also does NOT have telemetry. Had mine out beyond 3 kms and the RSSI barely dropped with its antenna amplifiers.
The noise floor in a Fixed Wing scenario, is much lower than a Quad of course, even so, it's a great advert for the (Tiny) XM Plus, popular with Mini Quads today.
The other thing to note, and you can easily prove this yourself, is that separating the antennas on a typical telemetry capable receiver will drastically increase its range. We did some ground tests and proved this beyond doubt. The antenna placement in your clip is way too close together for a telemetry capable receiver.
Goat one says to Goat 2, WTF is he doing with that big stick then..?....lol.
Great info JBard, thanks for doing the hard science and getting us this tasty tasty data!
Joshua, the XM+ does have rssi, it broadcasts it on a channel and works wonderfully. I've actually flown the longest distance with an XM+ compared to a stock x4r-sb and xsr. It's too bad you didn't get that in the video but very cool video anyways!
I know that but you need a flight controller to read that data and obviously I didn't have one in this test.
Joshua Bardwell ahh, I didn't realise there wasn't a flight controller. I can understand why you had trouble walking so far and still didn't have signal drop, I put my taranis in range mode and flew 300 meters with rssi still in the 70s/80s. Are you going to do a video on frskys new telemetry protocol :D?
they have rssi but don't they ship with non rssi firmware?i had to flash mine to get rssi.
I had to flash a rssi firmware to mine, too
Depends on where you bought it. If it comes from China most likely it has stock firmware. EU and US stock sometimes come with it already flashed with rssi firmware. I've bought several now from China and US stock and some have RSSI working out of the box without flashing the firmware. I always flash them anyways just so I know what firmware is installed.
Nice test, but maybe should be done on open field with various receivers of each model because there is lot of bad FrSky receivers. Your job is great and I understand the financial limits of that kind of test.
I get the best results with R-XSR and redundancy xm+ and changing antennas with replacement ones. I fly more than 2km without any warning. Of course that doesn't mean nothing but it's good for me.
Love this sort of testing. Thanks Joshua
Good idea for testing this, and thanks for doing it, but the methodology I don't understand. Why cross-test different receivers and different antenna lengths in random associations? It makes it so much more difficult to isolate the effect of the antenna length. I would focus on the receivers where it's easy to get rssi value. For each of those (there is only 3), test 3 different antenna lengths and record rssi for each combination. Then you have 9 curves that you can compare. You can see, for each receiver, which antenna length works best, and you can maybe then see a trend towards one antenna length giving better range in general.
I tested different receivers because most people who report short range are using a receiver-and-antenna combination. So the claim is not just, the shorter antenna on the R-XSR is worse. The claim is, the shorter antenna AND the R-XSR is worse.
So I start by testing the actual claim that is being made: is one receiver+antenna combo worse than another?
Then from there, I branch out to explore the effect of different antennas. Which is why in the 2nd test I had both a 1/4 and a 1/5 wave antenna on the same receiver (R-XSR).
Alright, I now understand a bit better the thought process. But in the end, the video's title question is not answered: "wrong antenna length?" With this method, no way of knowing if the receivers' hardware or antennas are at fault. Sorry for being picky, I'm an engineer too, after all :)
I think what you shod test is antenna placement on the Quad. and then walk away with the Quad.
To many is stuffing the antenna in there and having problems.
This works great:
www.thingiverse.com/thing:2147201
My hypothesis: EMI decreases the reception range. The reason you are getting such great range - compared to what you expect - is probably the lack of noisy electronics in proximity to the receiver.
Same experience here... I have 3 XSRs and experimented with different antenna lengths. Bone stock seems to work best and better than my two X4R SBs.
I have two R-XSRs sitting on my bench for new builds - lets see how this goes.
Wow that's a test rig! Amazing effort mate. Thanks.
Good Test ! They all seem pretty close .......... Have you tried different antennas on One Receiver unit ? ........ or different Transmitter antennas ? Also increasing your height was interesting ........ it would be beneficial to know the optimum height for your TX antenna . Great show !
Thanks for the great video Mr. Bardwell.
I think the antenna location on the craft is as important as antenna length. Personally, I have the antennas coming out of the bottom of my quads-planes, for the simple reason that they are flying above me and if the antennas are on the bottom of the craft there is a line of sight with no obstructions, back to my transmitter. Always place the antennas so they are no obstructions between you and the craft that you are flying to get max range.
We have similar discussions in our german forums without clear solutions too...
One short tip maybe you test it. If i shred a antenna i removed the jacket and then i evert the helical ground wire. Its then like a dipol told me an skilled "antennaguy". Functioned for me but not range tested because the Video Signal was the limited factor.
All of the antenna wires are about the same (wire type), not sure your long velocity factor analogy applies here. There's still quite a few variables (as you mentioned), antenna length, polarization, possible lumped elements changing the tune of the antenna, and SNR differences in the receivers, overall great video as usual
It may have been a slip of the tongue Joshua but loss of telemetry is not the same as loss of control link. This seems to be a common misconception among users and a point you could help clarify in one of your videos. This can be demonstrated when the TX is too close and the radio alternately reports Tele lost Tele recovered and yet all the while you have perfect control.
FrSky are actually taking a few liberties with the RX antennas since even with a perfect 1/4 wave radiator length the whisker design has no counterpoise element to complete the dipole leading to induced current in the sheild. FlySky on the other hand often fit their RX's with outer sleeves which acts as the second half of a 1/2 wave dipole and also acts to decouple the shield which as you know is a much better design choice. In fact It may be interesting to fit a pair of those sleeved dipoles on the FrSky RX's and compare to the FrSky whiskers.
One last thing PLEASE start complaining long and loud about the new style antenna coax and IPX connectors FrSky are now using. Hopefully they will take note and revert to the older style coax and uFL's. It would seem in striving for ultra miniaturisation they have forgotten the real world implications and practicalities of (to use the Australian vernacular) Piss weak coax.
Great effort trying to figure this out. With my time spent in CB radio hobby before, my experience is that the antenna length vs range is not that critical for receivers than transmitters. I.e. If you got the a bad antenna, say a wrong length antenna, even using a bare wire cut casually will still receive considerbly good. However the same antenna if used in a transmitter, can have high SWR and thus affect range considerably at the receiver side. Regarding to why the antenna length is different for different receivers, I AM PURELY GUESSING HERE, are they related to the telemetry transmission freq of these receivers? Since these receivers are with telemetry, meaning that these antennas are also used as transmitter antenna, is Frsky use a length that is particularly tuned to the freq(s) used by these different receivers? If that is true, then we should see fixed, say 32mm antenna for all non-telemetry receivers caused they are just for receiving ..... hmmm, looking forward for your further reseach into it.
Great vid JB, an interesting watch. Perfect for the lunch break. Eating and learning. Win win!
Makes me wonder what is the difference in using numbers in between the 23mm to 33mm. Meaning is there any difference between 23,24,25 and so on or if there are specific number that make the antenna good and just so happens that the 23,26,32 you tested is about there... Great video and its awesome that you put all this thought on questions people have but most dont bother to investigate!!!
33/23 is approx 1.41 (or square root of 2). On a pure sine wave it's the RMS to Peak ratio - 1:1.41. On a sine wave carrier + signal they probably tweak this number for tuning. Complicated maths involved? Yes.
Do I understand it all? 'Fraid not!
Excellent. Thanks for your time and efforts! You are the best!!
I was just comparing the replacement antenna to the originals on my R-XSR, replacement is right about 31mm, brand new R-XSR: 23mm...
Had to jump back in time...
I placed an R-xsr in my newest build for Lua and such, problem is I failsafed in a spot that I had no issues with my other build using Xm+. I noticed the length of the R-xsr antenna. It is much shorter. So I ordered some after market antennas to test. Thank you for this video! Seems to explain why this crash happened.. ScooterPie
Let the goats help ! I think it would have been good to put a 1/3 wavelength antenna in there and maybe even a no-antenna receiver in there just for a control.
Also dont forget that frsky receiver work as transmitter for telemetry and this antenna must be matched by SWR.
Change length may degrade telemetry performance.
Worth mentioning that an antenna for receiving only is not as critical as an antenna that broadcasts. In a badly adapted antenna, the radio energy will bounce back into the transmitter, and thus the range will decrease.
That would be interesting to measure swr on the counterparts with telemetry.
I was looking forward to seeing this video. Great
You've got to love how RC hobbyist with no background in RF electronics think that because they read a few blogs that they now know more then the RF engineers who designed the system. The only proper way to tune an antenna like these is with a VNA but then you would actually need to know how to read a smith chart LOL ....
Kep in mind that these are not simple receivers but transceivers that send back telemetry so they antenna is tuned for the transmit signal not the RX signal..
My advice is unless you have a VNA and know how to use it then you have no business cutting your antennas.
Very interesting results. Somewhat counterintuitive to preconceptions. Either way, great data to have.
I have to say that the stripped area of the frsky antenna is not the only active part.
The entire length is the working area. I don’t know why they even strip the end, radio waves go right through plastic so there is no need to strip it.
I don’t think that is coaxial wire, I may be wrong, I’ve looked at it closely.
Maybe they strip it so we will wear it out and buy more? Also those antenna pairs should all be at 90 degrees to each other.
I think you're mistaken about this. It is absolutely coaxial cable. It has a braided grounding shield along the length, which is why the exposed section radiates and the shielded section does not.
I have had 2 different XSR receivers bought from two different vendors and both times these receivers have given me RSSI warnings at 150-200 feet. I'd also have failsafe triggers on occasion. I have tried these XSR on different builds with different Flight Controllers to see if it makes a difference. It worked the same. I switched over to a XM+ with the 1/4 wave antennas to see if it worked better, and now I can easily go 500-600 feet and still have 70% RSSI. I don't know if the antenna is the difference or not, but I have no faith in the XSR receivers.
I might try to pop on some 1/4 wave antennas on my old XSR to see if it makes a difference. It's possible I got two dud receivers, but I find it hard to believe it could happen 2 times from two different vendors. I am going to try the R-XSR with swapped out 1/4 wave antennas to see if it will work better. I really like having Telemetry, but the XM+ has RSSI on channel 16 which is sufficient as long as I am paying attention to the value in my googles, however having Telemetry and the Taranis calling out warings is much preferred.
yep. the XSR stock antennas are just crap. i experienced the same.
You mentioned in an earlier reply that you cut your antennas to 33mm. I thought 1/4 wave was 31.25mm. Also, the X4R-SB was at 33.2mm and the XM+ was 32.7mm according to the video. That's odd. I guess all the receivers are tuned like Joshua suggests, but if you say 33mm worked for your XSRs, I will give it a shot.
great video brother
Thanks for your testing but you haven't tried an exact 1/4 wavelength. I would use one rx with a short antenna and measure range , then change to a long antenna on the same rx and check range and then a 1/4 wave antenna on same rx and check range.
I have very good experience with XM+ stock antennas (almost 1km of solid signal). XSR with stock antennas was much worse (around 700m when it already started giving me alerts). XSR with aftermarket antennas from banggood worked better but still lacking behind the XM+.
All 2nd generation of Tinyhawks including the Freestyle have RSSI as high as 99, while the older Tinyhawks maxes at 70. I wonder what Emax did to tune it. But even with the 30% difference, the distance was only 5 meters before failsafe, but it seems the new versions have better consistency in holding connections within range.
The only thing I noticed is the new antennas on the 2nd generation TH are thicker and stiffer.
Learnt loads. Many thanks. Ray
Love the Goats, they are notorious RF signal boosters though...
I've tested XM+ receiver range with stock 25mm antennas, then with 31mm antennas and last one was much better
I watched Stingy and Steele test this in real flight test. An XSR with 32mm mod outperformed x4R with 32mm stock, which both whooped an xsr with 23mm stock, therefore I run a modded XSR.
+gabbo2027 no offense to them but they did one test. Do you know that your antennas are the same coax as theirs, with the same velocity factor? Just one reason why multiple tests are valuable.
8:13 fu#$% you touched my heart there joshua haha nicest self advertising ever all the way!! I just bought 2 parallel chargers from your brand, and also i am thi king in patreon you, bjt first i have to understnd how does that work, you helped me so much in the hobby ... i trully owe you a beer at least! Thanks for all mate
www.patreon.com/thedroneracingengineer
Click "become a Patron" and go from there.
Nominal Velocity of Propagation is the key to your shorter antenna. Also interesting: Most of these receiver antennas are only a half dipole. So if you measure 1/4 length of unshielded wire it might be actually 1/2 wavelength antenna design but its only half of the antenna present.
My Futaba receiver is 144 total and 28mm at the end works good
Don't forget the antanna on the drone/plane must also be the same size as the transmitter to get full benafits.
Joshua, did you test using a standard Taranis antenna or did you use an upgraded antenna? I am trying to decide if it's worth modding my Taranis to upgrade the coax cable and use a higher dbi antenna to get more range.
It's a standard antenna.
Joshua, you are wrong about VF (Velocity Factor) of the coaxial cables. The formula is VF=1/SQRT(K), where K is dielectric constant of the central insulator. For example, if it is Polyethylene, then VF=1/SQRT(2.3)= 0.65938..., but we can round it to 0.66, or 66% of speed of light. If the same polyethylene has air bubbles inside in form of foam, then it can't be calculated because then K is between 1 and 2.3, for which only manufacturer can provide you the right information. VF for antenna element itself depend of ratio between wavelength and diameter of the radiating element and it is NEVER 1, only by math. Usually it is around 0.95, or 95% speed of light. The thicker the radiating material - the shorter antenna should be. But if radiating element is too thick, although it will be tuned, its Q factor (Quality factor) will be low, thus gain will be low as well. And, you can't measure antennas that way - by placing more antennas next each to other. Also, 'whisker' antennas are bad - it need at least 'sleeve' over the coaxial cable, and will be good that it has the other end of the dipole.
Make antennas like this, and you will be VERY surprised how much range you will get. :wildlab.org/index.php/2016/03/02/tutorial-how-to-increase-range-of-your-quadcopter/
Sorry I'm confused. Which part did I get wrong?
Did you test with a standard taranis also?
Wouldn’t the coax you put in that, improve your signal? Didn’t you have a previous video showing that you swapped out the aftermarket coax cable within the taranis to a even better/thicker one?
Great video as always!
I thought this as you made a comment roughly half way through about it always resulting better than expected for you
+Tip Nadroj this isn't that radio. This is stock.
Joshua Bardwell interesting, still love the tastings
Joshua Bardwell and testings... hah
I would like your opinion Joshua. I am having range problems with my XSR receiver. The receiver is new and I have a Taranis X9D Plus. The radio works fine with other receivers, only this one has range problems. How could I improve that? Is that a receiver problem? Tks a lot!
Thank you Josh, I appreciate your videos and tests.
Silly question, have you ever used a rubber coated paper clip for antennae?
I have tons of questions actually, I'm wondering what the pitch/gauge of the antenna changes about the signal efficiency as well as wether or not a wispy versus firm antenna matters at all either. :) I want more range for my drone and the vtx without having to double the literal strength and diverting it with anything aftermarket. (who knows, there may be a way with math on our side)
Jb another info spectacular! Thx
I've been messing with this for awhile after noticing on my xsr the RSSI was fluctuating a lot. So I tried putting new 31.2mm then shortened those antennas to 29.5mm then 26mm. What I noticed was the 29.5mm length for the antenna seemed to have a more stable RSSI then the the other 2 lengths and I have been able to fly in places where normally I would have failsafed. I think it is cause it is about 95% the wave length and copper wire has a velocity factor of about 0.95.
My best explanation is that the antenna needs shorter to compensate for the difference in speed between charges in the wire and the radio waves interacting with the wire.
So you suggest 29.5mm worked best for you instead of the of the longer 1/4 wave. Interesting.
Sometimes science tells us that we aren't asking the right questions. The important take away is that everything seems very forgiving. Try it with a spektrum and you'll be done in an hour.
Nice testing. I'm not sure that I would expect RSSI between different receivers to be calibrated the same. I think you need to test to failure in a straight line test - seems like that requires a longer runway. :-)
The "low RSSI" and "critical RSSI" warning level is defined on the Taranis at the same level for each receiver.... Anyway, in the 2nd test I was not reading RSSI values but physically watching for the LED to indicate a failsafe. That's functionally equivalent to testing to failure. The moment the RX light goes red, that's when the quad would fall out of the sky.
Thanks for all the work you put into your videos, but I want to learn more today! I just bought the Kopis 1 (love it) and have a question about mounting the receiver antennas. I was able to squeeze in a spektrum SPM4649T in there, but due to wiring restrictions I ended up with the antennas pointing towards the back from the front arms. You gave directions to mount them pointing to the front from the back arms. Can you tell me if that matters. I can make a longer harness to flip the receiver around if you think there will be an improvement. And could you explain why it matters if they point to the back or front? Please and thank you from an appreciative viewer.
It doesn't matter. Your way is fine.
I have a somewhat relevant question: Why do you mount your Rx antennas on your quads to back motor arms facing forward? Why not front arms facing backwards? Wouldn't that minimize the chances of the antennas acting like anchors in a crash and thus decrease the chances of damaging the antennas?
VilleF1N when you mount youre antennes facing forwards regardless on which arm. You will have the worst link, which is a good thing because then youre certain of a good link when you return.. jn stead of flying to that point of bad link, you turn around, and loose link completely..
There's a wire length coefficient in the circuitry of every receiver. Otherwise, I wouldn't be able to fly around with a completely disconnected RX antenna.
Looking at your first graph, would it be right to conclude that a stock XSR is better than an R-XSR reception wise ? The blue line is above the orange line only between 60-80ft. 20ft interval over 120ft total
None of these receivers use true dipole antennas. There is no ground plane. Therefore the receivers use the attached circuitry and coax length as the ground plane. This probably affects the results. Also, the antennas should probably be perfectly straight and vertical for repeatable measurements.