103 replies to this topic

[Terry]

Hi All
I've been thinking about building a tracking system for my downlink as my asistant keeps loosing track of the plane.
I know it can be done using the gps data sent to the ground and into a PC, the PC then moves a stepper motor with the aerial on. Is it possable to do this with out the PC ? Maybe with a micro-controller to make a neat all in one unit, or is it to much for a micro-controller to handle ?
I it possable to track the transmitter/plane with out the gps and using the RF ?
If you used 2 patch aerials next to each other pointing say 20deg apart and fed the signal to 2 recievers, could you use the AGC signal to drive the patch aerials right or left and balance in the middle ?

Terry
UK

[Kilrah]

Well...

Of course it is possible to do it with a microcontroller. I don't know how you transfer the GPS signal to the ground, but I suppose that you do it via the audio channel. I've never done this so I don't know what signal the receiver gives out. According to this it can load the controller quite a bit to decode the GPS info, but I don't think it should cause problems. Your tracker would have to know its position on the ground, and could calculate the necessary angle for the antenna to point to the plane.

About the system with 2 antennas, I would think it could work, but would cause problems. Imagine moving yourself and coming too close from one antenna, or in front of it. This would reduce the signal level on one receiver (or both, but of course not in the same manner) and cause errors easily. And this means 2 extra antennas and 1 extra receiver...

Regards,
Kilrah

[Mr.RC-Cam]

Cyber-flyer's (www.cyber-flyer.com) antenna tracking system uses a PC and GPS data. He offers some info on his site, including the PC software. A picture of the antenna can be seen in his 2003 flight log.

He also posted the schematic of his other antenna project, a diversity antenna system. The RSSI signal (buffered/conditioned AGC) from a pair of video Rx's is used to switch to the best video source.

[Bosse]

Hi

I also belive your idea could work, with some limitations however. Could be me brain-farting, but I think you would need three recivers/antennas. One of them would have a rotatable antenna, the others fixed, and you would be limited to fly within range of all three recievers. It would perhaps provide a better signal but it would not be a perfect sullotion.

I have been thinking of a second reciever, working more or less as a field-strength meter, with a sweeping or rotating directional antenna (In case of a rotating antenna, I wonder how a swivel coaxial connector should be designed ), in conjunction with a computer (or microcontroller). It would work like those flight instruments (VOR's ?) used when navigating a full sized aircraft with help of radio-beacons.

My Idea surely got a lot of limitations. Interference from external sources could be one. Another thing is that some vertical sweeping could be needed as well, making it even more complex for a hobbyist like myself.

I really wish I knew more about both analog and digital circuit design and other knowledge I need so I could test some of my ideas !

Like Mr RC-cam says, I guess GPS-data sent from the airplane is the best approach.

Best Regards,
Bosse

[Terry]

Im not sure why you think I would need 2 fixed aerials but anyway I like idea of using a sweeping or rotating aerial just to find the strongest signal then tell the main receiving aerial where to point. I think with suitable control software it would be more reliable as any odd blips could be filtered out. I am not interested in any vertical tracking as this angle can be set before the flight as I know the expected range I will be flying at and all but the highest gain aerials have a wide enough beam. I like this idea more than using GPS data as it can be used with any camera plane without the need for GPS,data link in the plane or a PC on the ground and little setting up. All in all much less to go wrong I think, if a signal is there it will find it ! No computer to crash, no data link to fail.

Terry
UK

[yb2normal]

Wouldn't using a sweeping aerial mean that you might loose your signal for prolonged periods as the antenna searches for the plane? If you are using the video to control the plane, this could be a bit nerve-wracking

[Terry]

No....The sweeping aerial only finds the direction of the signal then it tells the aerial with the screen connected where the strongest signal is.

Terry
UK

[Bosse]

The sweeping or rotating would be performed quite rapidly by another reciever/antenna and the information about wich bearing gives strongest signal, would be used for pointing the antenna of the actual radio-link by means of a servo motor and a computer/ microcontroller. In its simpliest form, the sweeping or rotating reciever/antenna would not even need to de-modulate the signal.

Best Regards,
Bosse

[Terry]

Ive been thinking about it and I think its worth a try !
We are looking at having 2 units here, 1 finding the bearing and the other pointing at it. The unit finding the bearing would sweep 360deg until it found a signal then would sweep right and left detecting the edges of the signal which may only be a few deg. The bearing of this would then be used by the other unit to pinpoint the origin of the signal. If the signal was lost the the first unit would go back to sweeping 360deg again but the other unit would stay pointing at the last known signal origin until the first unit found it again. Dose anyone have any ideas how I might send the position info from unit 1 to 2. The easy way is to just use a pot geared for 360deg use and outputting a 0-5v so 0v=0deg and 5v=359deg but its a bit iffy at the ends. Maybe stepper motors and just count the steps ?

Terry
UK

[cyber-flyer]

Please see this link for previous discussion of the same problem:

http://groups.msn.co...406320995506800

Regards,
cyber-flyer.

[Terry]

Thanks for that !
You said :
>thinking of scanning the ant using the RSSI for pointing
I used this technique on the field with manual antenna scanning. If plane is out of sight I move antenna back and forth and it will give idea about plane location.
Its pretty robust but takes time (10-20sec) to get the lock.One problem is that you may loose signal as antena moves away from the target. Another problem is that the pointing error is as large as beamwidth of you antenna. Yet another problem is that the radiation pattern from heli/plane is never uniform. When the plane turns in the air, the signal can increase or decrease without plane actually changing its position. One needs to do averaging to get correct idea of the plane location. If your antenna can sweep the sky pretty fast - let's say several cycles per second, than you can average faster and get lock in a second or two - something to try.

Its good to see it has been tryed, but I think I am right in saying that you gave up on the idea quite quickly ?
The problems you highlighted I think can be overcome, I remember seeing a demonstration of RADAR tracking at an airshow where a 10ghz dish tracked a metal target as it moved round a table. It followed it with a wobbling motion as it found the edges of the object. I thought then it must be possable for me to do the same with my plane as I have a good signal to track and not just a weak reflection.
I think the key to this is the two separate parts,the 'locator' system and the other the 'receiving' system.
You say it takes 10-20sec to get a lock, I dont see this as a problem as it is only the 'locator' system, the 'receiver' system will still be pointing at the last known position.
You also say that the beamwidth of your aerial causes a pointig error, this is not a problem as long as the 'locator aerial' beamwidth is smaller than the 'receiver aerial' beamwidth.
The problem of dropouts as the plane turns will not effect the position of the 'receiver aerial' until the 'locator aerial' finds the signal again and updates the 'receiver' system with the new possition.
Dose this make sense ?

Terry
UK

[Bosse]

If I held an antenna in my hand, and goofed around with it, I might get intermittent feedback but I would never be able to effectively track or as you say "lock" on a relatively fast moving object beyond visual range.

Radars VOR's and some other hardware is used in a similair fashion as I describe. The principle is very old and is not my invention at all. A combat radar can track another airplane and get a "lock" in less than one second.

Only the antenna movement speed limits performance in that respect. As a side note, I wonder how a synthetic aperture radar works.

I could be wrong but I do not think you need to first rotate 360° and then wiggle sideways. I think I would either use continious rotation or continious +/- sweeping.

Sweeping simplifies some design issues.

The voids (or interminent abcence of data if you like), could be filled out by means of (hopefully smart) prediction code if needed.

Since that is mostly about software design, you should try all methods you can think of. Making such tests when you have a well functioning hardware, will most likely be so rewarding and fun so it will be well worth the effort.

What is most important is making a smart hardware setup, at least as I think of it.
Software tweaking is so much more tolerable and fun than re-building a mechanical construction.

Best Regards,
Bosse

PS

Remarks about spelling errors are only accepted if written in Norse or Norse language deritatives.

DS

[cyber-flyer]

Dose this make sense ?

It is all sounds plausible and I'd love to see such system at work.
The major difficulty with such approach is that two moving antennas (and receivers) will double the cost and size of the project.
The reason I mention the size is that I kind of miss the days when I had single receiver with double quad antenna attached to it. I didn't even had a case to carry all the video downlink equipment to the field. All I needed is that somebody pointed the antenna at the plane. Nowdays I am howling big case full of electronics to the field and it takes me half and hour just to set it up. And I still rather have my buddy standing by and watching me fly (just in case something goes wrong). To tell you the truth I keep asking myself - was it worth is?
Regards,
cyber-flyer

[Bosse]

The extra hardware required would not cost that much. I spent 50 bucks on a DC-DC converter from TracoPower the other day, just to get rid of that annoying 10 cell battery pack. (and it works superbly, thank god).

My ordinary R/C transmitter is, (but not reciver, wich is a Schultze reciever and strangely enough, not very expensive but anyways almost famous for its extremely reliable operation) a cheap Hitec (not the 7-channel one wich is posessed by evil spirits on 35 MHz) and it works perfectly well. I never had a single jerk on any channel except when powering up, wich is normal

Before I used this reciever, was like a pinball table sometimes.

Sometimes I notice the reciever indicates one single glitch after 30 minutes of flying but since it masks it very effectively during operation, it is just a lazy blink on a diode when the airplane stands safely on the tarmac.

I would gladely spend some 75-300 bucks for another reciever and some other hardware and components, to get an advacend antenna pointing system, if I had the knowledge to design the circuitry or had drawings from other sources.

Best Regards,
Bosse

[Bosse]

I agree a buddy pointing the reciver antenna is the best method generally.

At least within visual range, and that he or she is required to be serious and do not goof around too much, making the video annoying to look at with those drop-outs you should expect unless you have a high-powered transmitter.

(Both my young sons points antenna much much better than the buddies at the club )

Best Regards,
Bosse

[Terry]

The cost is not an issue as I dont think it will be much different to a GPS based system. As I make my own aerials the only big cost is for another receiver, I also have stepper motors and a lathe/milling machine so time is the main problem.
If I am to use the agc voltage then it must be above the 'no signal' voltage even at long range, IS IT ? IS THERE A PARTICULAR RECEIVER I SHOULD USE ?
Also I need to choose a bearing reporting system that lends itself well to software manipulation of the bearing data so that I can fine tune the performance as has been said. ANY IDEAS WHICH WAY TO DO IT ?

Terry
UK

[Bosse]

The reciever could be any reciever designed for the frequency in question, as long as it has an RSSI voltage output somewhere (inside probably). RSSI voltage could be digitized and fed into the computer by readily available components/ devices. Since the RSSI voltage throw might me small, it could be good to either use a differential-amplifier or a high resolution A/D converter.

There are also power-meter chips from MAXIM but that is a long-shot.

Apart from the reciever, I think I would use one of those robotic interface cards, one or two standard servos and an old PC-computer to point the antenna. The rest would be handled by sofware wich guess I would have to write myself (with great pleasure).

One advantage in using an ordinary computer to point the antenna is that you might want to integrate other cool functions in the future. I think it is good to have as many of the functions as possible concentrated into one box.

Maybe a PC-computer is already in your field equipment.

You might for instance want to view downlinked video on the computer screen (requires graphics card with video-in), use a vehicle tracking program, view telemetry data, even use it for virtual flight instruments on the screen, like a virtual cockpit, digitize video in real time, all in one box that also controls antenna.

Perhaps it would be fun to control the airplane from the computer, using a joystick and an interface-cable connected to the transmitter.

What I write is only ideas, thoughts and suggestions so be very critical before you spend time or money on it.

Best Regards,
Bosse

Edited by Bosse, 15 September 2003 - 10:30 AM.

[Bosse]

About RSSI vs GPS approach.

I think I might want to have both.

Best regards,
Bosse

[Terry]

What you say makes sense for you Bosse, but I dont like computers much !
I can not see the screen on my laptop in sunlight and I dont know how to write
the software for it so the ideal system for me would be PIC based as I can write software for those !
I also do the opposite to you in that I keep all important systems as simple as possible so hopefully they wont fail !!!?
Im not sure about the use of servos, Im thinking about stepper motors still as they should be more reliable for constant use.
I need this system to be 100% reliable as I will not have both systems like you, I dont see the point ?
I just need to find the right receiver then I can make a start.

Terry
UK

[Bosse]

Terry,

Microcontrollers have advantages over PC-computers. They are cheap and very small size. If you are comfortable in programming them, then the choice is easy if you do not need the power of a PC computer for other stuff.

Reason I would use both methods is that I think it is interesting and fun.

The RSSI method, does not exclude the GPS method hardware wise in case I use a PC-computer wich I will. (the hardware for pointing the antenna will be the same). The GPS will be onboard anyway and the extra reciever not needed for antenna pointing could be used as an auxiliary video reciever instead, possibly equipped with a different antenna.

Stepper motors are far more long-lasting and provides much better precision than ordinary servos and should be preferred if reliability is of outmost importance.

However I find it easier to start experimenting with ordinary servos. They already have a gearbox and various types of servo arms.

Edited by Bosse, 16 September 2003 - 01:02 AM.