52 replies to this topic

[Mr.RC-Cam]

Other than Old Man Mike's cool looking CPOD that he discussed a few months ago, there's been little talk about improving our FPV antennas. I'm surprised by that since there are an endless number of interesting designs that can be built. And it is possible for mortals to do this without expensive test equipment (your Rx's RSSI signal is a cheap but useful measuring tool).

But I will admit I am lucky to have access to some nice RF test equipment (the more complicated antenna configurations really deserve precision measurements). Because a lot of what I do is for commercial use, I'm reluctant to post about it since I can't really divulge all the technical details. However, to help spur on you DiY'ers, I will at least show some pretty pictures of two antenna designs that are working well. With luck, these will show up at my favorite FPV shop before the end of the year.

The first design is a 900MHz Turnstile antenna for use on the Video Rx. Turnstiles have circular polarization which helps to eliminate cross-polarization fading (like when the model aircraft banks in a turn). I have to give CyberFlyer credit on this one since his success with his DiY 2.4GHz Turnstile several years ago has always intrigued me. Unfortunately, he never published plans. Starting from scratch, through a couple iterations I ended up with a decently performing configuration. The required precision phase splitter is a quadrature hybrid component that is mounted on a custom PCB with microstrip feed lines. I'm currently using brass rods for the elements, but if all goes well the final design will be pushed onto a single PCB with the elements incorporated in the copper art. Not shown is the reflector, which is simply a copper ground plane that mounts behind the Turnstile. This one has been a tricky design, so it will be awhile before it is released. Also, I expect to have 1.2GHz and 1.3GHz versions too.

Here's the Turnstile:



The second antenna is a much simpler design. It is a classic 900MHz V-Dipole for use on the Video Tx. To help ensure consistent builds it uses PCB construction. It has the same performance as a full size dipole, but the V shape reduces the overall height. It is very lightweight too, so it is a nice choice for smaller models using 900MHz Tx's. If it becomes popular then it will be offered in 1.2GHz and 1.3GHz configurations too.

Here's what the V-Dipole looks like:


Sorry for the teasers, but that's about all I can show. But, hopefully some other forum members have ventured into DiY antennas and will discuss their project beyond pretty pictures.

[Helix1]

Hi RC-Cam

Well, for what it worth I have some ideas regards antennas which I will be sharing with RC- CAM forum members sometime in the near future, to include a directional antenna idea for mounting on a UAV/FPV platform – an electronicly steerable idea I worked on in the 1990’s, but which was then shelved by the "suites" when priorities moved on. Nothing too fancy to it and it suprizes me that I haven’t come across any efforts to implement a similar idea in the amateur model flying community yet – the skill sets are certainly present on this and other forums.

Since retiring earlier this year I’ve spent a good few hours perusing RC/UAV/FPV websites and forums, and they are all full of great model/camera/control/software and autopilot ideas, but this all important component in comm’s links seldom gets anywhere near eqiuvilant attention.

Antenna design offers enormous development opportunity to address the range limits imposed through power output restrictions placed on the comm’s hardware used by the unlicensed model UAV & FPV flying community.

Other than for the relationship between model size/frequency used, and from the ground station perspective the size of antenna one is willing to lug around from location to location, the antenna is the one hardware component on the comm’s equipment list that is free from legislated performance restriction.

If as much time & effort is put into antenna improvement & development as is given to other aspects of UAV/FPV/RC flying, we could realise very significant & worthwhile increases to the average comm's range flyers currently achieve.

Helix1

[Devonian]

Thomas, how odd you should start a topic on antennas.
In the last couple of days, I came across cyber-flyers website whilst googling turnstiles
http://www.cyber-flyer.com/ and click the link on left side for 'Video, Telemetry....'
There are some of his musings on turnstiles, but the links are now dead (MSN pages).

I've made and used turnstiles for weather satellite decoding but the best performer was a QFH.
I've tried to stimulate some interest in the QFH, but everyone just poo-pooh's it.
Another alternative is the Lindenblad, but not really suited for the video transmitter.


Helix, welcome to the forum and maybe you have some useful input/ideas on antenna design you might like to share?

Nigel.

[Terry]

I have made some aerials in the past for 2.4ghz and lower but now I use 5.8Ghz and have no test gear for it so just use off the shelf. The only slight exception is to add a plain reflector behind the dipole on my 4 way system, surprised nobody else bothers as its an easy way to double range over plain dipole.

I look forward to seeing any more aerial ideas here , anyone made a 5.8Ghz Helical yet ?

Terry

[Mr.RC-Cam]

Since retiring earlier this year I've spent a good few hours perusing RC/UAV/FPV websites and forums, and they are all full of great model/camera/control/software and autopilot ideas, but this all important component in comm's links seldom gets anywhere near equivalent attention.

I agree. Nearly all of the attention nowadays is on OSD's and RTH systems. Almost nothing is being done to help the video system's RF link. Improving junk Tx's and Rx's is difficult, but making a good antenna is within the skills of us mortals.

In regards to FPV antennas, overall I am amazed at the manufacturers' poor attention to their recipe. For example, in the last few months I have personally retuned a couple hundred factory made dipoles because the far east suppliers have no respect for the needed build accuracy. Every antenna shipment is like 52 card pick up, and it usually ends up with me on a network analyzer fixing their incorrect tuning. So, I can say that there is plenty of room for performance improvement and good opportunity for DiY solutions.

In the last couple of days, I came across cyber-flyers website whilst googling turnstiles
http://www.cyber-flyer.com/ and click the link on left side for 'Video, Telemetry....'
There are some of his musings on turnstiles, but the links are now dead (MSN pages).

Cyber-Flyer's links went to the old MSN hosted rc-cam forum. MSN shut down a couple years ago and all of our old technical discussions were lost (interesting R/C video history, from the pioneers, is gone). But I think Kilrah is hosting archives from the old forum, so maybe he will chime in and post the link to his special site.

[Devonian]

I'll agree on the apparent focus on everything but the antenna system.
On almost all other forums, antenna discussions are few and far between, with most prefering to simply buy rather than explore.
For me, half the enjoyment/fun is in the build and not only in the flying.

Terry,
So far, I've only made a couple of antenna's for 5.8GHz - a simple quarter wave and a threequarter wave (on my homebrewed 5.8 VTx & VRx).
Another simple and high gain Rx ant to try would be a corner reflector - I've made one for VHF and it works well, but obviously it's directional.
A helical Rx is only going to be of any use if the Tx is transmitting the same polarity.
Maybe if the weather is good over the weekend we'll have a chat (with my brother, Steve )

Thomas,
I'd like to be able to read some of the 'pioneers work' and if Kilrah does have it archived, it'd be a bonus.
I don't have the luxury of any test gear for such high frequencies (5.8GHz), so any testing is of the 'suck it and see' variety.

Looking forward to this thread's development...

Nigel.

[Kilrah]

http://kilrah.dynalias.net/rccam/

It's a rather crude archive, not searchable, and all long discussions only show the last page, but still quite a lot of info there. You can browse with the links on the top/bottom, that should work.

Back 1 Month Back 1 Week Forward 1 Week Forward 1 Month
Prev 50 Next 50

[Helix1]

Devonian

"A helical Rx is only going to be of any use if the Tx is transmitting the same polarity...."?

I must respectfully disagree with you – the helix is one of the few antenna’s ideally suited to recieving Tx from dynamic platforms. But I think it important that we put your statement in context – namely, against the background of how antennas are put to use in the context of this forum i.e. as either antenna’s for use on flying models, or as ground-station antennas for those flying models – and in that latter role, they display characteristics few other antennas offer.

Most important of those characteristics, in the context of this forum at least (and I think its important we keep the points relivant to the role i.e. how antennas are put to use by members of the forum), is the axial mode helical antenna’s linearity across changes to received signal polarization.

Granted, it will demonstrate roughly -3dB* used as a receiver for any signal not circular/elipticaly polarized, as you may well be aware of, but considering the variation in and regularity with which flying models change there oreintation, there is a strong argument supporting helical type antennas(?)in that they eliminate firstly the drop-out, and secondly, the drop in recieved signal strength that users of other antenna types experience when their model is flying at any orientation other than straight & level.

Axial mode helical have other advantages as well (compared to linear type antennas), but how relivant they are in the context of this forum is very much dependant on defining the problem or issue one is trying to address first, but I think it worth noting that designed specificaly as a receieve only antenna (compared to most of the other antenna types used by FPV model flyers as receive only antennas), axial mode helicals can be optimised for extremly low side lobe "leakage", and consequently, excellent resitance to off-axis on-band interference, offering the user an SNR ** [figure] many can only dream of.

Were it not for the fact that parabolic dishes exist, and more recently the invention of SAR (synthetic-aperture radar) arrays, the axial mode helical would be the primary antenna in search & intercept ground-to-air and air-to-ground applications.

* A lessor known point regards this 3dB loss: 80% plus can be recoverd through optimisation of the reflector size & shape (especialy shape). Another few % is recoverable through carefull tapering of the helix winding. Size/shape optimaisation of the axial mode helical antennas' reflector is argubly as important as is optimisation of the helical coil. In short - the 3dB loss is a mute point.

** Another issue altogether, but I'll mention it: SNR is not given anywhere near the attention it deserves when undertaking efforts to optimise system performance - it is as important as Gain, and in certain circumstances, argubly even more important!

Helix1

Edited by Helix1, 15 September 2010 - 11:15 PM.

[Terry]

Thanks Helix1. I was just going to say to Nigel that many years ago it was making a helical for my 2.4Ghz system that made it even usable. I wont drag on about it too much as I have told the story many times but back in 2000 I got my first 10mW system that gave about 300ft range and cost me £400 I think. It was making a helical that boosted this to around 1500ft then later a very low noise amp made from a kit boosted it again to over 3000ft. The picture I posted of the helical is on the old forum

Terry

[Old Man Mike]

Other than Old Man Mike's cool looking CPOD that he discussed a few months ago, there's been little talk about improving our FPV antennas. ....
The first design is a 900MHz Turnstile antenna for use on the Video Rx. Turnstiles have circular polarization which helps to eliminate cross-polarization fading (like when the model aircraft banks in a turn). ...

Just a few quick comments here:

1) Turnstiles with the elements in the X and Y planes (assumed parallel to ground) will have circular polarization only in the Z axis (straight up and straight down).

2) The primarly reason to have circular polarization is to reduce multipath. Multipath will occur most often at low radiation angles (close to parallel with ground) and/or when the signal is traveling thru reflection objects such as trees. For this, the turnstile antenna will not be effective.

3) You can make the turnstile antenna radiate circular polarization more effectively (at the horizon) if you place a 30 degree sloping ground plane underneath it with narrow slots cut in the ground plane just beneath the elements. Certainly more complicated than desired for an fpv antenna.

4) No matter what type of circular polarized antenna you develop, I strongly recommend that you construct a version of the antenna with the opposite polariation (RHCP vs LHCP) so that you can test the axial ratio. Multipath rejection is all about the axial ratio. Jump for joy if you can get 15 dB or more over most of the angles used between the RX and TX during flights. For quadcopter flying I find the critical angles are -10 degrees to +30 degrees.

5) Circular polarization is often confused with trying to achieve an omni AZ pattern. Of course you want your airborne FPV antenna to have an omni AZ pattern regardless if it is linear polarization or circular polarization.

6) QFH antennas suffer from the same problem as the turnstile antenna: maximum axial ratio is off the ends which will be pointed to the ground if you want an omni directional pattern.

OMM

[Mr.RC-Cam]

Thanks OMM for the comments. My primary focus for the turnstile is to reduce cross polarization fading. In this type of installation the Tx uses a common dipole and the Rx's turnstile antenna is used like a flat panel patch (that is, positioned so it faces towards the model). I'm measuring good gain symmetry through steep bank angles, so it is working well. With this configuration Cyber-flyer demonstrated remarkable improvements to his system (I believe it's what he used for his 20K altitude flight). I was quite impressed by his work back in the dawn of FPV and I'm simply riding off his coat tails on the implementation.

Here's the only photo I can find of his 2.4GHz Rx antenna setup (photo is linked is from his site). His turnstile efforts are basically what I have done, but redesigned for 900MHz.

The only slight exception is to add a plain reflector behind the dipole on my 4 way system, surprised nobody else bothers as its an easy way to double range over plain dipole.

Totally agree about the use of a simple reflector. Just a simple kitchen pie plate, roughly 1/2 wavelength behind the Rx's dipole, will give +3dB more gain and will help reduce multipathing. Something like the windsurfer reflector hack will do even better.

Edited by Mr.RC-Cam, 16 September 2010 - 07:51 AM.

[Terry]

I agree I dont think circular polarisation is chosen in most cases for its resistance to multipath but to combat fading in turns. I chose to make a helical as it was the most forgiving design and therefor had the most chance of success, which turned out to be true.
As I fly level with low angles of bank I tend not to worry now and just use vertical aerials for video unless I want to fly overhead.KISS !
I dont even bother with vertical tracking as once your a mile or more out the angle required is very low and dose not change much, left and right is all I need. I did have vertical tracking but it was not worth the trouble for the type of flying I do.

Terry

[Old Man Mike]

Thanks OMM for the comments. My primary focus for the turnstile is to reduce cross polarization fading ....

I agree I dont think circular polarisation is chosen in most cases for its resistance to multipath but to combat fading in turns......
Terry

Ah, sorry guys. I keep forgetting that most folks here are flying planes instead of quadcopters. Since you guys are almost always flying line of sight and high then multipath is not a big issue. As a quadcopter flyer I like to hover in low places far away chasing deer, etc. Multipath rejection is key in that application.

Of course using circular polarisation on the ground side works to combat the deep fades that can happen when both ends use linear polarization as long as you don't mind the 3 dB loss. If you are looking for an even simpler approach that provides more gain than the turnstile, you might consider just modifying the standard patch antenna by doing this:



This approach only works over a narrow frequency range compared to feeding adjancent sides with a 90 degree phase shift. However it is a simple way to eliminate deep fades while retaining most of the gain of the patch.

OMM

[Mr.RC-Cam]

Thanks OMM for the info. Multipath rejection is something I like too, but that is not the focus of this particular antenna design. Basically, I'm waiting for you to mass produce your CPOD.

I've tried the dog ear'd patch trickery (back when the GP Patch was being developed) and it never gave me much benefit. Cyberflyer tried too, with a couple of patch designs, and did not find any of the advertised advantages from the DiY "circular" patch ham radio projects floating around. It seems that without carefully fed phasing, at best they are still linear beasts with perhaps a slight tease of circular RF currents.

For sure, I can see a remarkably uniform gain from the Turnstile (versus the traditional signal reduction during the cross-polarization angles), so it is definitely a capable design. But I agree that it is not the simplest antenna to create. But my goal is to make it easier to build (still working on that) so that it can be affordable to produce.

[Old Man Mike]

Thanks OMM for the info. Multipath rejection is something I like too, but that is not the focus of this particular antenna design. Basically, I'm waiting for you to mass produce your CPOD.

Rusty's CNC machine is currently not up to producing the two special PWBs required for the CPOD without a lot of cutting time which is making it quite unattractive for cost effective production. Not sure what we will eventually do but it is possible that we will release all the details into the public domain to see if anyone else can make a low cost version of it. With all the work Rusty did iterating the cutting pattern of my design, I consider it a co-developed product. If we decide to release it, it will include the model details which can be run in 4necX as well as the CNC fabrication details. That decision probably will not be made until much later this year.

I've tried the dog ear'd patch trickery (back when the GP Patch was being developed) and it never gave me much benefit. Cyberflyer tried too, with a couple of patch designs, and did not find any of the advertised advantages from the DiY "circular" patch ham radio projects floating around. It seems that without carefully fed phasing, at best they are still linear beasts with perhaps a slight tease of circular RF currents.

Well that is interesting. I had not tried it myself and wondered how well it would work, especially since the lowest SWR would NOT be at the optimum frequency for circular polarization. Probably the separate 90 degree phase shifted feeds would have worked much better but then that is much more work.

For sure, I can see a remarkably uniform gain from the Turnstile (versus the traditional signal reduction during the cross-polarization angles), so it is definitely a capable design. But I agree that it is not the simplest antenna to create. But my goal is to make it easier to build (still working on that) so that it can be affordable to produce.

Good luck with it. As others have noted, a helix antenna would also work well for the ground side antenna. Some years ago I designed a helix which used a single loop instead of the normal ground screen and just recently I made one to test with 1280 Mhz Video:



It has about 2dB more gain than a two stack of the CPOD antennas but of course it is not omni directional like the CPODs.

OMM

[Terry]

The single loop is an interesting idea, how dose it compare to the normal larger ground screen?

Terry

[Old Man Mike]

The single loop is an interesting idea, how dose it compare to the normal larger ground screen?

Terry

Forward gain for helix length is at least equal to the a typical helix. As you would imagine, there is a little more radiation behind the loop reflector helix but it is still more than 20 dB down. Here is the plot for a 23 turn design which you can compare to other helix designs:



This design also has a three step taper and the model results were confirmed by independent testing. Additional details as well as the NEC model file is available here:

http://www.af9y.com/helix.htm

I probably should not be posting this info here since I believe it has already been discussed in another thread on this forum and it is getting off the subject a bit - sorry.

OMM

[Mr.RC-Cam]

I probably should not be posting this info here since I believe it has already been discussed in another thread on this forum and it is getting off the subject a bit

All FPV type antenna project discussions are welcome here, so not off topic at all. More details, the better; so bring it on!

[Old Man Mike]

.....Here's what the V-Dipole looks like:


Sorry for the teasers, but that's about all I can show. But, hopefully some other forum members have ventured into DiY antennas and will discuss their project beyond pretty pictures.

Hey Mr RC-Cam,

Since you are now selling the V-Dipole, you might want to correct the "Balum" to "Balun" in the following PDF file:

http://www.dpcav.com...s/antvd-900.pdf

Also, since you have included a Balun, I'm a little surprised that you would advocate an up or down position for mounting since the balun should isolate the coax and produce balanced radiation on both elements. A side mounted V pattern should look like this:



Of course without the balun the coax becomes part of the antenna resulting in an unbalanced radiation pattern similar to this:



This pattern varies depending on the length of the coax and everything in the ground side of the TX wiring.

I applaud your inclusion of the balun and I wonder if you know how effective it is. It is not easy to find ferrite material that works well at 1 Ghz and it is often a struggle to get even a few hundred ohms of impedance for isolation.

OMM

[Mr.RC-Cam]

you might want to correct the "Balum" to "Balun" in the following PDF file:

Oops, need to have a talk to the proof reader. Thanks for the heads up.

I'm a little surprised that you would advocate an up or down position for mounting since the balun should isolate the coax and produce balanced radiation on both elements.

I agree that it should not matter. But there were a couple reports that the recommended orientation seemed to have a slight effect. Since I can not see any difference in my testing, I suspect it was just false observations, but sometimes unexpected results have a reason. Just for good measure, I added the comment to the instructions just in case there is some unforeseen interaction with some installations. Even if it is unnecessary, it does no real harm.