-
Any RF antenna design boffins around?
Posted by
icon
on 25 Nov, 2015 23:49
-
Hi
I have made, and am making some more, 5.8GHz antennas of a circularly polarised, 'clover leaf' design. These are used to transmit video from... well anything really.
The first thing that surprised me (I don't think in GHz) was that these are a 'short' at DC - the centre conductor of a length of RG316 is simply soldered via 3 short wires arranged at 120deg, to the shielding braid.
My first go was with 1.6mm silver solder wire, which I happened to have lying around. Surprise, surprise it actually worked! I haven't tested the range yet.
I'm going to have another go, this time with much thinner copper covered steel (actually MIG welding wire). at 5.8GHz, I presume the 'skin' is thin enough that the signal is carried in the copper.
My questions are these: what actual effect does the thickness of the material have? 1.6mm vs 0.6mm? Thicker wire is obviously less prone to physical distortion, but would it be better (better how?) to accept the fragility of thinner wire? Second question; the silver will tarnish - the copper is very thin so eventually the steel underneath will rust. Wouldn't it be great to dip it in Plastidip, or spray it with lacquer? Nay! This will disrupt the pixie dust which is essential to its function. What effect does this kind of thing have? Does it affect the frequency the antenna is 'tuned' to (if that even means anything)? If so, is it possible to guess how much, and make changes accordingly, or do you need fabulously expensive, multi-GHz-capable equipment?
I realise that this probably requires a three-year course to explain, but I'd be grateful if someone had a go.
Cheers
John
-
#1 Reply
Posted by
DimitriP
on 26 Nov, 2015 00:07
-
-
#2 Reply
Posted by
vk6zgo
on 26 Nov, 2015 00:19
-
There is copper & copper----the copper layer on the welding wire may have impurities due to being optimised for other than conductivity.
The layer is also quite thin & subject to corrosion.
A sprayed layer of lacquer shouldn't detune the antenna much----a heavy layer of Plastidip probably will.
Basically,you have to work out which way the resonant frequency changes,& correct for it.
No "pixie dust",but you probably need some method of checking resonance,if you want to optimise the things.
-
#3 Reply
Posted by
retrolefty
on 26 Nov, 2015 00:24
-
Many decades ago when more active in ham radio I build several wire and metal antennas. One thing I can say is that none flew right off the design/calculation board. A SWR meter was always used/needed to 'trim' for best performance.
Testing and adjustments is more important then what wire gauge you use in my opinion.
-
#4 Reply
Posted by
Howardlong
on 26 Nov, 2015 00:28
-
It will have an effect on the effective length of the antenna elements and have an effect on the impedance.
You might want to invest some time in a modelling program like EZNEC. I haven't done much antenna work for a while, but I did use modelling programs like EZNEC quite extensively some years ago, there may be more modern alternatives nowadays. You can see the effect of changing things like length and thickness.
Personally speaking, I'm a bit old in the tooth and like to do final impedance tweaking using a VNA. There are some wizards these days who can model the whole thing and be confident it'll be pretty much there from their simulation alone. I found once I'd used a VNA and could see interactively the effects of changing things in real time I just didn't want to go back to random and somewhat aimless tweaking. Luckily these days there are some pretty reasonably priced PC based VNAs, but I'm not sure they go up to 5 or 6 gigs.
One final thing is to be very wary of mechanical supporting structures of any kind at microwaves. Insulators have dielectric and loss properties not the same as free space and can be catastrophic on radiation patterns for example, not to mention matching. Note that your antenna might, of you're lucky, present a reasonable match... but then so will a 50 ohm resistor!
-
#5 Reply
Posted by
T3sl4co1l
on 26 Nov, 2015 10:03
-
Yeah, if your only gauge of performance is "hey, it worked, I guess" -- just put up a 1/4 wave whip or something like that, and leave it be. It's not that you won't do better -- you certainly can -- but you won't
know, and without measurement, any conclusion is meaningless.
Checking signal strength is helpful, but not ideal. It has to be done carefully. Example: an antenna which radiates modestly in the forward direction, but strongly in the sideways direction, might still measure better than a merely "good" unidirectional antenna. Or worse. How? That strong side lobe could take a long path back to the receiver, thanks to reflections in the test area. The test site is critical. An open air test site is preferable, and an anechoic test chamber even better still!
Signal strength indicators are also not consistent. RSSI is an arbitrary scale. Bandwidth is adaptive, and often lazy (i.e., it might not renegotiate for a faster channel, unless it has to). Power level might be adaptive too (i.e., it only uses what it needs), which would *really* throw off your measurement!
So it's very tricky to do, and really needs to have good equipment to get a sound answer. Without that equipment, be very weary of, and careful about, any possible sources of interference: side lobes, reflections, mismatched loads, and dubious measurements. Doubt everything -- yourself foremost, as you are the easiest person to fool -- and also remember that, if it's not falsifiable, it's not science.
Tim
-
#6 Reply
Posted by
hendorog
on 26 Nov, 2015 10:35
-
Have you seen this forum?
http://fpvlab.com/forums/forumdisplay.php?16-DIYAccording to ibcrazy, who popularised the cloverleaf for FPV, thicker wire on a cloverleaf increases the bandwidth of the antenna - i.e. essentially the antenna becomes a bit more tolerant of manufacturing errors.
There is also a cheap and cheerful SWR meter you can build for 5.8GHz which will allow you to test that theory.
Using that device you could at least adjust for the frequency shift effect of the effect of plastic dip/lacquer.
You can get somewhere with very cheap DIY gear. There are other options as well before you get to an Agilent VNA for $$$
-
#7 Reply
Posted by
icon
on 26 Nov, 2015 12:08
-
There is also a cheap and cheerful SWR meter you can build for 5.8GHz which will allow you to test that theory.
You couldn't point me to the one you're thinking of?
Thanks
J
-
#8 Reply
Posted by
PA0PBZ
on 26 Nov, 2015 12:18
-
-
#9 Reply
Posted by
icon
on 26 Nov, 2015 20:14
-
Ah, right. The sticking point there is the directional coupler, which might spoil the notion of 'cheap'. Also opens the possibility of getting the wrong sort? Is there any particular thing I should be looking for (or avoiding)? What's a 'negative detector'? ('Directional coupler with negative detector')
Cheers
J
-
#10 Reply
Posted by
hendorog
on 26 Nov, 2015 20:56
-
Ah, right. The sticking point there is the directional coupler, which might spoil the notion of 'cheap'. Also opens the possibility of getting the wrong sort? Is there any particular thing I should be looking for (or avoiding)? What's a 'negative detector'? ('Directional coupler with negative detector')
Yeah there aren't many on ebay at the moment. Just a case of waiting and jumping on one when it comes up cheap.
It's quite possible to build one as they are very simple, but brace yourself for a lot of trial and erroring. There are a couple of examples in that thread.
You want something which covers the frequency range of interest and has the right coupling to get the best out of the detector.
4-8GHz is the octave which covers 5.8GHz. Wider frequency ones are available but the directivity is generally worse. They tend to be more expensive on ebay as well.
For a 200mW Tx and the standard diode detector you want about 10dB coupling.
You probably want SMA connectors, although I did notice some reasonably cheap N type DC's if you already have or can get hold of adapters.
e.g.
http://www.ebay.com/itm/Narda-3044-10-4-to-8-GHz-10-dB-Type-N-F-Coaxial-Directional-Coupler-/221906405127?hash=item33aaa70707:g:lkEAAOSwNNxWFB4jThe Directional coupler with negative detector you mentioned sounds to me like a coupler which already has a detector built into it.
I don't know for sure but it might/should be possible to just use that without needing the detector circuit from fpvlab.
I'm sure someone here will know exactly how it works
-
#11 Reply
Posted by
Howardlong
on 26 Nov, 2015 21:54
-
If it's any help, I started my microwave career on directional couplers and crystal detectors, with AC millivoltmeters on the detector and an AM modulated signal source. While it's definitely possible to make measurements with this technique, in these days of instant gratification I'm not sure how many people will have the patience to want to follow it through. Certainly when I was using that technique, the very same people who'd suggested and explained it to me were using VNAs and RF power meters with calibrated heads themselves!
I now understand why. It's fine for the odd rare measurement, but for more than that, especially for frequency sweeps, I'd not want to go back.
-
#12 Reply
Posted by
LukeW
on 26 Nov, 2015 23:42
-
Regarding the OP antenna being a DC ground short, that's not unusual for many antennas.
-
#13 Reply
Posted by
icon
on 26 Nov, 2015 23:58
-
Regarding the OP antenna being a DC ground short, that's not unusual for many antennas.
I'm sure you're right, but as a non-RF bod it caused me a "Wait...what?" moment.
J
-
#14 Reply
Posted by
TheSteve
on 27 Nov, 2015 01:12
-
Having had a go at building some cloverleaf antennas myself I can say that the odds of it actually being tuned exactly where you want it is slim to nil. I haven't even found a commercially built one that was on frequency yet(note that I have not tried one from ibcrazy). They were all several hundred MHz from the ideal spot. Of course they still work, but not quite as well as they actually can. I doubt very few of the antennas sold commercially are ever tested on a proper machine, they just measure, cut, bend, solder and ship - hoping to get them in the ballpark. I did build a 2.4 GHz one that was nearly bang on - they are much more forgiving at lower frequencies.
My test device is a calibrated Agilent N1996A 6 GHz spectrum analyzer with built in return loss bridge and SWR measurement suite.
-
#15 Reply
Posted by
hendorog
on 27 Nov, 2015 01:55
-
Having had a go at building some cloverleaf antennas myself I can say that the odds of it actually being tuned exactly where you want it is slim to nil. I haven't even found a commercially built one that was on frequency yet(note that I have not tried one from ibcrazy). They were all several hundred MHz from the ideal spot. Of course they still work, but not quite as well as they actually can. I doubt very few of the antennas sold commercially are ever tested on a proper machine, they just measure, cut, bend, solder and ship - hoping to get them in the ballpark. I did build a 2.4 GHz one that was nearly bang on - they are much more forgiving at lower frequencies.
My test device is a calibrated Agilent N1996A 6 GHz spectrum analyzer with built in return loss bridge and SWR measurement suite.
They can be tuned by increasing/reducing the feedpoint gap. The idea is to not cut the centre conductor too short until you've finished tuning.
-
#16 Reply
Posted by
TheSteve
on 27 Nov, 2015 02:29
-
Having had a go at building some cloverleaf antennas myself I can say that the odds of it actually being tuned exactly where you want it is slim to nil. I haven't even found a commercially built one that was on frequency yet(note that I have not tried one from ibcrazy). They were all several hundred MHz from the ideal spot. Of course they still work, but not quite as well as they actually can. I doubt very few of the antennas sold commercially are ever tested on a proper machine, they just measure, cut, bend, solder and ship - hoping to get them in the ballpark. I did build a 2.4 GHz one that was nearly bang on - they are much more forgiving at lower frequencies.
My test device is a calibrated Agilent N1996A 6 GHz spectrum analyzer with built in return loss bridge and SWR measurement suite.
They can be tuned by increasing/reducing the feedpoint gap. The idea is to not cut the centre conductor too short until you've finished tuning.
I might try making more 5.8 GHz models when I get some time, see how long it takes to get one bang on frequency.
-
#17 Reply
Posted by
T3sl4co1l
on 27 Nov, 2015 04:01
-
I would caution that, if mere 5% errors in frequency are substantial in performance/characteristics, it's a dangerously precise and high Q antenna, and probably won't serve you well anyway for those reasons. A 1/4 wave whip will cover a much wider bandwidth (particularly with a relatively wide, or flared, element), and has only one element to tune.
Tim
-
#18 Reply
Posted by
hendorog
on 27 Nov, 2015 07:09
-
I would caution that, if mere 5% errors in frequency are substantial in performance/characteristics, it's a dangerously precise and high Q antenna, and probably won't serve you well anyway for those reasons. A 1/4 wave whip will cover a much wider bandwidth (particularly with a relatively wide, or flared, element), and has only one element to tune.
Tim
The point of using these for fpv video is the circular polarisation. The multi-pathing is pretty terrible on 5.8GHz without that.
Its not that hard to make and tune for good performance as long as there is some way of measuring it.
The 1/4 wave antennas that come with the tx's need tuning with an SWR meter as well.
The other common option is a helical. It is forgiving in bandwidth and easy to build.
-
#19 Reply
Posted by
icon
on 27 Nov, 2015 09:41
-
Having had a go at building some cloverleaf antennas myself I can say that the odds of it actually being tuned exactly where you want it is slim to nil. I haven't even found a commercially built one that was on frequency yet(note that I have not tried one from ibcrazy). They were all several hundred MHz from the ideal spot. Of course they still work, but not quite as well as they actually can.
And then the 'band' I'm working with runs from 5.645GHz to 5.945GHz, which is 300MHz anyway. I suppose that allows for some selective assembly!
J
-
#20 Reply
Posted by
Zero999
on 27 Nov, 2015 13:08
-
Second question; the silver will tarnish
I doubt that will be a problem. If the silver on the surface tarnishes, then it will become an insulator and the current will continue to flow underneath it. The only time there would be a problem is if the dielectric loss of the silver oxide layer had a high loss at the frequency of interest or if all the silver oxidised.
-
-
http://www.urc.cc/pubs/URC-2002a.pdfWhile silver oxide does seem to be kind of lossy, the thickness of the layout would be very thin. This is no different than a dielectric loaded wire antenna. First component as one would expect in this relationship is the thickness of the coating.
No good engineer in this feed does it with out the aid of software to simulate designs. It is just like building things but cheaper, faster to do, and ... . Look up a software called NEC (Numerical Electromagnetic Code). Versions 1 and 2 are open sourced and free.
https://en.wikipedia.org/wiki/Numerical_Electromagnetics_CodeThere are many ways to make an antenna wider bandwidth. If that is the only problem my I suggest the conical dipole. That said antennas for these frequencies are so cheap now days. Unless than you are wanting to learn then grab a good book on the subject, good simulation software, and some free time and go to town.
Testing antenna performance is much much more than just the input match. A 50 ohm termination has a great VSWR, but a real crappy far field pattern for some applications
.
-
#22 Reply
Posted by
TheSteve
on 28 Nov, 2015 02:27
-
The clover leaf is already modeled and verified to be a solid antenna, especially for FPV video uses at 5.8 GHz. The only real challenge is to build one that is properly tuned as some pretty detailed precision is needed.
-
-
The clover leaf is already modeled and verified to be a solid antenna, especially for FPV video uses at 5.8 GHz. The only real challenge is to build one that is properly tuned as some pretty detailed precision is needed.
I have no dough there is a close form solution to the ideal clover leaf antenna. I am a little confused if you have it already modeled and simulated why is this not just a monte carlo simulation problem.
If you have a physical antenna the thing to do is mount it on the air craft and use a signal generator and a spectrum analyzer or a vector network analyzer. Use these to measure 2 things the S11 or the VSWR and far field pattern(This would be S21 on the VNA). With antennas where circular polarization is desired this very important know what polarization is as a function of the angle. If is not just a constant.
You have to include the air craft as it is most certainly in the reactive near field of the antenna and will act as part of the antenna system.
If it is a tuning problem, what previsions are there for tuning? Matching networks, trimmable parts of the antenna, and so on.
-
#24 Reply
Posted by
T3sl4co1l
on 28 Nov, 2015 07:29
-
The clover leaf is already modeled and verified to be a solid antenna, especially for FPV video uses at 5.8 GHz. The only real challenge is to build one that is properly tuned as some pretty detailed precision is needed.
That sounds rather contradictory though: if it's "solid", then why are people apparently having trouble building it to the exact frequency? Why is the frequency so particular?
Like I said earlier: if the bandwidth is so narrow that it's hard to get it right, it's not a good design. Good design means good everything: not just "oh well, you didn't build it right" -- a good designer
must accommodate the tolerances of the manufacturing process!
Offhand, I don't see actual measurements and discussion of this, so I have no idea how troublesome this is.
Tim
-
#25 Reply
Posted by
hendorog
on 28 Nov, 2015 10:02
-
The clover leaf is already modeled and verified to be a solid antenna, especially for FPV video uses at 5.8 GHz. The only real challenge is to build one that is properly tuned as some pretty detailed precision is needed.
That sounds rather contradictory though: if it's "solid", then why are people apparently having trouble building it to the exact frequency? Why is the frequency so particular?
Like I said earlier: if the bandwidth is so narrow that it's hard to get it right, it's not a good design. Good design means good everything: not just "oh well, you didn't build it right" -- a good designer must accommodate the tolerances of the manufacturing process!
Offhand, I don't see actual measurements and discussion of this, so I have no idea how troublesome this is.
Tim
Probably because we are talking about home made antennas, made using sidecutters. High precision is required at that frequency. This ain't HF band where a meter here or there doesn't matter. Materials matter too.
Construction and tuning empirically makes sense for these.
I think modelling means its been simulated in 4NEC.
I think solid means that thousands have been built and flown with good results.
-
#26 Reply
Posted by
T3sl4co1l
on 28 Nov, 2015 10:50
-
Sidecutters are one thing, but there are simple antennas (like conicals) with bandwidth of octaves. You have to go out of your way to screw that up!
(Obvious problem: wideband antennas center around 150-200 ohms. You need a matching circuit for 50 ohms. Which tends to restrict bandwidth...)
Specifically for circular polarization and low directivity, I don't know exactly what would be most appropriate (a helical might be too directional, but is otherwise pretty wide band, and not hard to make). Maybe a pancake spiral design?
Tim
-
#27 Reply
Posted by
hendorog
on 29 Nov, 2015 00:29
-
Sidecutters are one thing, but there are simple antennas (like conicals) with bandwidth of octaves. You have to go out of your way to screw that up!
(Obvious problem: wideband antennas center around 150-200 ohms. You need a matching circuit for 50 ohms. Which tends to restrict bandwidth...)
Specifically for circular polarization and low directivity, I don't know exactly what would be most appropriate (a helical might be too directional, but is otherwise pretty wide band, and not hard to make). Maybe a pancake spiral design?
Tim
Yep - I mentioned helicals earlier for exactly that reason. They are often used for the ground station but never on the aircraft.
Anything that goes on the plane needs to be reasonably omni and of course CP.
The other antennas sound interesting, I'll have a search around but do you have any links handy?
-
#28 Reply
Posted by
T3sl4co1l
on 29 Nov, 2015 03:34
-
This is the kind of principle that gets really high bandwidth,
https://en.wikipedia.org/wiki/Self-complementary_antennahttp://www.radioeng.cz/fulltexts/2006/06_03_15_19.pdfA tapered impedance transformer / balun shouldn't be too bad at these frequencies, for only modest bandwidth. And probably just one turn's worth of spiral would get the desired polarization, while having a more or less dipole radiation pattern. It would be easy to cut from foil following a printout, or fabricated on PCB directly. It could also be made from wires tacked together, but you'd need to make both sides of each element (i.e., trace the outer perimeter), plus cross pieces for stiffness and such.. it would be tedious to fabricate.
Regarding general bandwidth, the dimension of the elements is the first determinant there. Using very thin wires leads to a high Q (and higher losses, and poorer coupling to space). This is the typical case with simple wire designs, e.g. monopole, dipole, L, folded, rhomb, etc. And, I'm guessing, the cloverleaf in this thread. Wider, especially tapered elements help a lot, even just in one dimension (as the planar construction above). Conical constructions, and fractal self-similar and self-complementary designs like the log spiral, log periodic, and so on, get very wide bandwidth, which might not be needed here, but shows the ultimate evolution of the concept, and the motivating geometrical principles.
Tim
-
#29 Reply
Posted by
Howardlong
on 30 Nov, 2015 07:52
-
Sidecutters are one thing, but there are simple antennas (like conicals) with bandwidth of octaves. You have to go out of your way to screw that up!
(Obvious problem: wideband antennas center around 150-200 ohms. You need a matching circuit for 50 ohms. Which tends to restrict bandwidth...)
Specifically for circular polarization and low directivity, I don't know exactly what would be most appropriate (a helical might be too directional, but is otherwise pretty wide band, and not hard to make). Maybe a pancake spiral design?
Tim
Yep - I mentioned helicals earlier for exactly that reason. They are often used for the ground station but never on the aircraft.
Anything that goes on the plane needs to be reasonably omni and of course CP.
The other antennas sound interesting, I'll have a search around but do you have any links handy?
FWIW I spent a good deal of the early 2000s working on circular polarised antennas for space applications. To make a reasonable helical you need to take care on the matching and feed point, and the mechanical fix. Be very careful of plastic standoffs that will have unknown dielectric and loss effects. That will substantially affect both the axial ratio and the directionality. I know that because of work we did on the antenna range with many antennas at the time particularly including helicals and CP patches.
In addition, the off-axis radiation will not be very circular for a patch or axial helical. In fact, the polarisation sense will typically reverse off-axis. That is why Lindenblads are often used in aviation on the ground segment.
One further thing, I don't think CP is used to reduce multipath for FPV, it's to reduce polarisation mismatch fading. I assume you're using linear polarisation on one antenna and CP on the other? That being the case, you already have a 3dB loss. However you are much less likely to see deep fades due to polarisation mismatch.
If both antennas were CP, multipath would still be a problem as reflected CP will be in the opposite polarisation sense, nulling the direct path.
-
#30 Reply
Posted by
T3sl4co1l
on 30 Nov, 2015 08:53
-
-
#31 Reply
Posted by
Howardlong
on 30 Nov, 2015 14:17
-
Ooh, that's pretty neat:
http://ww2.amsat.org/wordpress/wp-content/uploads/2014/01/70ParaLindy.pdf
Indeed, Tony was a bit of a polymath, he had his fingers in a lot of pies from antennas to embedded DSP, in many respects a man after my own heart. We used to meet up once or twice a year to exchange notes, and imbibed more than is strictly necessary as a result. Very regrettably he was taken from us quite early in life last year, a true gent who distributed knowledge and know-how, and is sadly missed.
-
#32 Reply
Posted by
hendorog
on 30 Nov, 2015 19:34
-
FWIW I spent a good deal of the early 2000s working on circular polarised antennas for space applications. To make a reasonable helical you need to take care on the matching and feed point, and the mechanical fix. Be very careful of plastic standoffs that will have unknown dielectric and loss effects. That will substantially affect both the axial ratio and the directionality. I know that because of work we did on the antenna range with many antennas at the time particularly including helicals and CP patches.
In addition, the off-axis radiation will not be very circular for a patch or axial helical. In fact, the polarisation sense will typically reverse off-axis. That is why Lindenblads are often used in aviation on the ground segment.
Thanks for the info.
The parasitic Lindenblad below looks interesting. I do recall reading about it previously in the context of modifying the stock dipoles which come with the low cost fpv gear.
One further thing, I don't think CP is used to reduce multipath for FPV, it's to reduce polarisation mismatch fading. I assume you're using linear polarisation on one antenna and CP on the other? That being the case, you already have a 3dB loss. However you are much less likely to see deep fades due to polarisation mismatch.
AFAIK most people use CP on both ends for multipath rejection. However, I'm sure some do what you suggest when doing FPV aerobatics.
If both antennas were CP, multipath would still be a problem as reflected CP will be in the opposite polarisation sense, nulling the direct path.
But surely the CP rx antenna and the reflection are cross polarised and so the reflection is attenuated?
I need to blow the dust off some of my gear and do some proper tests. I got a bit sidetracked by this new RF world I discovered, building test gear and experimenting and forgot about actually flying the planes around
Edit: The helical design which became popular for FPV is here:
http://www.rcgroups.com/forums/showthread.php?t=1377791And the cloverleaf:
http://www.rcgroups.com/forums/showthread.php?t=1388264
-
#33 Reply
Posted by
moya034
on 30 Nov, 2015 19:38
-
Not sure if someone said this already, but in generally, thicker diameter elements in an antenna will make it more broad-banded. (It will cover a wider frequency range at an acceptable VSWR).
(I'm not really sure if this still true in the microwave freqs, but I know it to be true at HF/VHF/UHF)
-
#34 Reply
Posted by
Howardlong
on 30 Nov, 2015 21:17
-
One further thing, I don't think CP is used to reduce multipath for FPV, it's to reduce polarisation mismatch fading. I assume you're using linear polarisation on one antenna and CP on the other? That being the case, you already have a 3dB loss. However you are much less likely to see deep fades due to polarisation mismatch.
AFAIK most people use CP on both ends for multipath rejection. However, I'm sure some do what you suggest when doing FPV aerobatics.
Hmm, I'd have thought polarisation mismatch more than multipath.
If both antennas were CP, multipath would still be a problem as reflected CP will be in the opposite polarisation sense, nulling the direct path.
But surely the CP rx antenna and the reflection are cross polarised and so the reflection is attenuated?
On reflection(!) you are absolutely right. The reflected reverse sense is strongly attenuated due to the polarisation mismatch, it does not destructively interfere. CP at both ends makes sense for mitigating multipath in this case.
One thing to consider is that the further the drone is away, the lower the elevation angle for a given static height. My understanding is that there's a 500' AGL flying limit (at least here in the UK), so realistically having helicals or patches pointing up and down with (say) a 45 degree beamwidths isn't going to be great when the a/c is a little way away: you want greater gain at the lower angles, given inverse square law. Plus there's the complication that off-axis, the polarisation won't be very circular, and in some cases will even be in the opposite sense. Directly overhead, the distance is at its minimum, you don't need much gain overhead.
A Lindenblad is a good idea, but fabricating one for 5GHz will be mechanically challenging.
As a matter of interest are all FPV video links analogue? I can see why of you're trying to minimise latency of you're controlling the a/c, but I thought drones were self flying, or is this RC? If latency isn't a problem, a FEC digital stream would typically provide a better range and be less prone to multipath fading, but it would innevitably introduce latency as part of the temporal diversity scheme introduced to mitigate against burst interference and fading.
Sorry, I have no direct experience of drones and RC. (FWIW, I fly real aeroplanes ;-) )
-
#35 Reply
Posted by
hendorog
on 30 Nov, 2015 22:56
-
One thing to consider is that the further the drone is away, the lower the elevation angle for a given static height. My understanding is that there's a 500' AGL flying limit (at least here in the UK), so realistically having helicals or patches pointing up and down with (say) a 45 degree beamwidths isn't going to be great when the a/c is a little way away: you want greater gain at the lower angles, given inverse square law. Plus there's the complication that off-axis, the polarisation won't be very circular, and in some cases will even be in the opposite sense. Directly overhead, the distance is at its minimum, you don't need much gain overhead.
The altitude limit here is 400' - which gives 100' separation in theory. Not sure about the UK but given the common history its probably the same there.
The horizontal coverage is more important than vertical, as you say the angles become narrow the further away you are and so the area an antenna can cover without moving it is actually quite large. People use antenna trackers or multiple antennas and diversity to increase that area.
I gave up on all that and just mounted the antenna on the googles so I could turn my head when the picture got fuzzy
There has been a big revamp of the rules here in NZ, and flying beyond line of sight is now not permitted. Exemptions can be applied for but are costly and require safety plans etc.
As a matter of interest are all FPV video links analogue? I can see why of you're trying to minimise latency of you're controlling the a/c, but I thought drones were self flying, or is this RC? If latency isn't a problem, a FEC digital stream would typically provide a better range and be less prone to multipath fading, but it would innevitably introduce latency as part of the temporal diversity scheme introduced to mitigate against burst interference and fading.
Yep, in general all analogue, although digital is becoming more popular as people experiment with HD. Latency is seen as quite important generally.
The quadcopter types all have a flight controller of some sort, usually with an autopilot. Depending upon the goal latency may or may not matter. Dodging in and out of trees is pretty latency sensitive, whereas flying a pre-programmed path for mapping isn't.
Sorry, I have no direct experience of drones and RC. (FWIW, I fly real aeroplanes ;-) )
Yeah I've dabbled with that too
Ironically I had a near miss with an RC plane when I was a newbie solo student pilot.
-
-
I have found QFH antennas to be excellent. There's an excellent chapter on them in "Reflections 3" by Walt Maxwell.
-
#37 Reply
Posted by
Psi
on 01 Dec, 2015 06:36
-
There's new super sensitive receiver preamps you can get that work remarkably well to improve the signal.
Normal receiver = almost complete snow
Receiver with preamp = almost perfect picture.
-
#38 Reply
Posted by
Psi
on 01 Dec, 2015 06:49
-
As a matter of interest are all FPV video links analogue? I can see why of you're trying to minimise latency of you're controlling the a/c, but I thought drones were self flying, or is this RC? If latency isn't a problem, a FEC digital stream would typically provide a better range and be less prone to multipath fading, but it would innevitably introduce latency as part of the temporal diversity scheme introduced to mitigate against burst interference and fading.
Sorry, I have no direct experience of drones and RC. (FWIW, I fly real aeroplanes ;-) )
99.9% is standard definition analog. The few digital systems that exist are for when you absolutely need HD and they all cost far to much. eg, HD system $2000, analog SD system $10
The industry is still waiting for a low latency digital system.
Latency really is a serious issue, you cant delay it even by 1 frame when realtime control is needed at the speeds we fly around at. See video below.
PAL 25FPS is 40ms per frame. At 100kph if you delay the video stream by 40ms you are flying more than 1meter behind your actual position.
This is from onboard HD video camera, guy was flying from separate analog video feed.
Set youtube to 1080p 60fps if your PC can handle it, its worth it.
-
#39 Reply
Posted by
Howardlong
on 01 Dec, 2015 07:30
-
I am not sure of the legislation outside the UK, but my understanding is that here the a/c must be within line of sight vision of the operator, so under these regulations what would be the need for low latency video?
I am sure that the cost of digital will reduce, possibly a problem as well as latency is reducing power consumption and therefore battery mass on a light weight design. If extremely low latency is a requirement for remote RC flying, I'm not sure how realistically a digital system could improve over analogue in that scenario: digital systems are dependent on FEC to combat bit loss in both burst and gaussian scenarios, and work by sending redundant data with time separation, hence the latency. In analogue there's no attempt at error correction, which is why you get the fading and fuzziness, and your brain is left to interpolate the errors and see through the noise. In digital, well as we all know, it either works or it doesn't, there's nothing in between for us to interpret.
-
#40 Reply
Posted by
hendorog
on 01 Dec, 2015 08:12
-
There's new super sensitive receiver preamps you can get that work remarkably well to improve the signal.
Normal receiver = almost complete snow
Receiver with preamp = almost perfect picture.
Sounds like an LNA. I have one I got from ebay but never got around to trying it out. Do you have any specs for noise figure to get those good results?
-
#41 Reply
Posted by
Psi
on 01 Dec, 2015 08:19
-
I am not sure of the legislation outside the UK, but my understanding is that here the a/c must be within line of sight vision of the operator, so under these regulations what would be the need for low latency video?
bit of a grey area depending on country.
In NZ you must have an observer who can see the aircraft in order to fly it FPV and be legal.
-
#42 Reply
Posted by
Howardlong
on 01 Dec, 2015 08:23
-
There's new super sensitive receiver preamps you can get that work remarkably well to improve the signal.
Normal receiver = almost complete snow
Receiver with preamp = almost perfect picture.
Sounds like an LNA. I have one I got from ebay but never got around to trying it out. Do you have any specs for noise figure to get those good results?
I'd also say either the receiver is rubbish or there's an awful lot of loss in the cable between the antenna and receiver for the difference to be that much.
-
#43 Reply
Posted by
hendorog
on 01 Dec, 2015 08:26
-
I am not sure of the legislation outside the UK, but my understanding is that here the a/c must be within line of sight vision of the operator, so under these regulations what would be the need for low latency video?
I am sure that the cost of digital will reduce, possibly a problem as well as latency is reducing power consumption and therefore battery mass on a light weight design. If extremely low latency is a requirement for remote RC flying, I'm not sure how realistically a digital system could improve over analogue in that scenario: digital systems are dependent on FEC to combat bit loss in both burst and gaussian scenarios, and work by sending redundant data with time separation, hence the latency. In analogue there's no attempt at error correction, which is why you get the fading and fuzziness, and your brain is left to interpolate the errors and see through the noise. In digital, well as we all know, it either works or it doesn't, there's nothing in between for us to interpret.
FPV flyers wear goggles or stare at the screen and so can't see the model directly, the legislation usually requires an observer.
Therefore the LOS is just a safety thing. The rules are really not that well thought through however, as you should be looking out and listening for full size heading into your area and not observing the model. Full size are much easier to see and hear, its obvious thats what you should look for.
That fixation on the model is what caused the RC guy to fly his model in front of me when I was taking off.
Not many people actually understand the dynamics of what happens when RC and full size are in the same airspace. Things like parallax are a big issue for the observer.
The power consumption of the video system isn't significant when compared with the consumption of the model motors. The nature of digital dropping out suddenly isn't ideal. A snowy picture is better than no picture at all
-
#44 Reply
Posted by
hendorog
on 01 Dec, 2015 08:28
-
There's new super sensitive receiver preamps you can get that work remarkably well to improve the signal.
Normal receiver = almost complete snow
Receiver with preamp = almost perfect picture.
Sounds like an LNA. I have one I got from ebay but never got around to trying it out. Do you have any specs for noise figure to get those good results?
I'd also say either the receiver is rubbish or there's an awful lot of loss in the cable between the antenna and receiver for the difference to be that much.
Yep I'd say that's fair comment, I suspect they are rubbish, but they are also cheap. The 5.8 ones are based on a single chip from Richwave from memory.
Edit: I am surprised the difference was that much though.
Edit2: This is how cheap they are
http://www.ebay.com/itm/High-Performance-5-8Ghz-Video-AV-Receiver-Module-for-FPV-/111827864924?hash=item1a0975cd5c:g:3vYAAOSwsFpWTZAz
-
#45 Reply
Posted by
Howardlong
on 01 Dec, 2015 08:53
-
That fixation on the model is what caused the RC guy to fly his model in front of me when I was taking off.
Well, nobody died. I was the pilot of a PA28 in an airprox earlier this year, this was roughly the equivalent to having oncoming traffic on the wrong side of the road. We had to take evasive action for fear of a collision as it was not at all clear what the other pilot's intentions were, he was certainly not following pilot 101 procedure inside an airport ATZ: although it doesn't state it in the report, I can only conclude he was lost. Very scary!
http://www.airproxboard.org.uk/docs/423/2015039.pdfThe power consumption of the video system isn't significant when compared with the consumption of the model motors. The nature of digital dropping out suddenly isn't ideal. A snowy picture is better than no picture at all
I'd agree with that.
-
#46 Reply
Posted by
hendorog
on 01 Dec, 2015 09:25
-
That fixation on the model is what caused the RC guy to fly his model in front of me when I was taking off.
Well, nobody died.
Heh, yeah thankfully. There are limited options when you are low and slow and something large is whizzing around in front of you. Not sure if many pilots have worked out what they would do.
I was the pilot of a PA28 in an airprox earlier this year, this was roughly the equivalent to having oncoming traffic on the wrong side of the road. We had to take evasive action for fear of a collision as it was not at all clear what the other pilot's intentions were, he was certainly not following pilot 101 procedure inside an airport ATZ: although it doesn't state it in the report, I can only conclude he was lost. Very scary! http://www.airproxboard.org.uk/docs/423/2015039.pdf
Not nice. Steaming in on base like that is just not British