Automatic adjusting FPV receiver antenna


the project I'm trying to realise is an automatic adjustment system for a patch antenna for my FPV drone.

There are similar systems working with a GPS tracking of the drone, which do work but you need a good (expensive) GPS module to realise that.

So my idea is based on the NDB communication by using the FPV transmitting waves to adjust a loop antenna like the ADF does. Only here is the transmitter the moving one and the receiver on the groundstation.

Now comes the tricky part: The NDB works through MF anf LF, but my transmitter sends in UHF (5.8GHz). Placing a second NDB transmitter on the drone is not an option due to the long antenna system that it has to be. The thing is with UHF the antenna has to be very short and I don't know how to realise making the loopantannae to receive a proper signal from the transmitter.

Is there a chance to make this work with UHF? Could be a high resolution be achieved?

The big advantege I see by doing it with UHF is the faster setting time, but I can imagine that there will be complications with the adjusting and/or receiving through the loop antennae.

Is it posible to build such an antenna?

Since I've never seen this type of direction finding other then in NDB's I don't know if it works with higher frequencies.

I hope someone can help me.

Best regards MajorProb

The higher frequencies are blocked by smaller objects. But they reflect off pretty much anything. That fills in the ‘holes’ behind objects. So a directional antenna will see reflections coming from everywhere.

Look closer at the commercial systems. If you find one that doesn’t use GPS, then the concept is practical and you can see what kind of antenna they are using. If there isn’t one, then it suggests that DF isn’t practical at that frequency. Nothing’s impossible, but it might take several thousand dollars worth of antenna and signal processing to do it.

I haven't found anything like it... Everyone is just talking about either GPS or optical tracking with stereocameras, which isn't practical for me either since when I'm flying behind a tree I don't have a visual...

But your point with the reflection is very helpfull since I haven't thought of that yet. This makes it a lot harder. Thank you for that advice.

I might have to fall back to plan C: building a NDB that is very small and doesn't use much power, which I don't think is realisable, is it? (has to fit on a 250 quadcopter frame...)

Do you have another idea to realise my plan to keep a patch antenna focusing my drone?

Thank you for your help I really appreciate it.

Best regards MajorProb

Can it follow the noise? If there are no other noise sources in the vicinity, a couple of microphones might provide the required pointing.

I don't think so the distance get pretty far, plus the surrounding is too noisy.

I was thinking about dead reckoning tracking with an inertial sensor but that might be to inaccurate...

Why not use a simple 1/4 wave as an omnidirectional antenna?

dangle the receive antenna below the aircraft so it isn't obstructed.

It has no directional gain, so you'll need higher power for long range.


Hi, Google RDF systems UHF

RDF= Radio Direction Finding

It uses an array of aerials, no moving parts.

There have been homebrew versions. Check Amateur/Ham Radio

Tom..... :)

At the moment I'm using a dipol antenna as transmitter (on tlhe aircraft) and a helix antenna on ground. I am limited in power with 25mW... (my country doesn't allow more)

So my plan is to concentrate the patch antenna on ground to receive a better signal of the dipol antenna. In order to adjust the patch antenna (due to its small income angle), I want to make it automatic. That is the only way I see atm to improve the connection and therefore the distance. Placeing a DF on the drone is not possible sadly. But my problem is the ADF.

@Tom Good advise I will follow that lead. The problem with that ist, that my FPV transmitting works with 5.8GHz ant the controlls with 2.4GHz (Icould use both for ADF) so that is still a higher frequency then the RDF u suggested, but the working principle sounds interesting. Will follow taht lead tomorrow ;)

Best regards MajorProb

Can you use two patch antennas? Does the video system give you any output related to signal strength? Just set the antennas at a small angle from each other and rotate the assembly in the direction of the antenna with the strongest signal. It will weave back and forth, but the signal should be good on both antennas.

Actually, there may be even a simpler way to do it. Use goggles with a separate feed to left and right eyes. Hang two antennas around your neck with the left one feeding the left eye, right to right. Then rotate your body until both eyes are clear. If the drone moves, you will know which way to turn.

Hahaha Morgan funny idea but the goal is to sit and focus on racing instead of turning the head all the time.

The two antenna system would be only two dimensional, but I need it to be 3D... But having something like a watchdog, that indicades signal strengths from different directions could be a way of thinking.

Therefore I need a way to analyse multiple signal strengths. Any ideason how?

You have to track up and down too? That’s an extra level of difficulty. For that I would recommend a dedicated emitter on the copter that can be tracked. Maybe a powerful IR LED pulsed with a specific pattern that can be recognised by IR remote control receivers? You may still have a reflection problem and you may even need more than one set of receiver elements to account for near and far - if the copter passes directly overhead, you need something different to tracking it 100m away.

The human eye is an amazing thing. How about an assistant to point the dish?

I recently came across a system intended to do this for full-size aircraft. The radio units were $10k-$30k for each end but the required tracking antenna cost $300k and shipped in a 20-ft container. If your antenna only cost ten times as much as the rest of your rig, I think you got it cheap.

Hahaha we propably came across the same thing :P

I don't think this project is realizable with the methods I can effort...

But I thank you guys for your great Input (which brought me to new projects already) and the help.

I'll keep this project in the back of my head and once I'll find any type of solution I'll try again :)