Let's Make Robots!

RC RF directional beacon

I am still wondering if it is possible to create a simple, cheap, fairly accurate, directional beacon for outdoor use. And I keep thinking I should try to re-purpose an RC transmitter. So here it goes:

Would you deem it possible to create a directional antenna for an RC transmitter?

Would this antenna be able to rotate? Fully continous, or just 360 CW followed by 360 CCW?

Which RC system would you consider? There are so many frequencies in use. And modulation types.

Speaking of which, which freqs and mods are mostly used on the cheap RC toys anyway? Noone ever bothers to mention that when selling them second hand.


Here is the design (if I may use that word here) I have in the back of my head so far. Mind you, I never played with RC, let alone tinkered with these electronics.


Put a directional antenna on a rotation platform (maybe include the whole TX on there as well). Read the angular position of the platform (its "heading") using some sort of encoder. Use this position (0-359 degrees) to encode the signals like you would in servo communication. Zero degrees would come out as 1.0 ms wide pulses, 180 degrees as 1.5 ms wide and 359 degrees as 2.0 ms wide pulses.

Chris has demonstrated that the receiver can feed these pulses straight into a picaxe which could easily decode them back into degrees or rads.

Here's a thousand more words for ya:


The blue dot is my beacon. The yellow rectangle the receiving party (for arguments sake, lets call it a robot). The beacon in this position would transmit 1.11 ms wide pulses which the robot would translate back into 40 degrees. Thus the robot knows on which bearing from the beacon it is at that moment.

The beacon's transmitting beam would continue to rotate, further increasing the transmitted pulse width according to its heading. Every robot in the scanned area would be able to use this beacon passively.

Further calculating where the robot is exactly is beyond the scope of this post.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Allow me to bump an old thread into a, hopefully refreshed, new audience.

I am wondering if an Xbee receiver would be able to measure signal strength. And of course, report it to my micro controller.

I was thinking of using an 60MmW Xbee [$32USD]with a chip antenna mounted inside an old C-Band satellite horn that sweeps around and transmitts a variable that stores its' current heading. Then you could use a 1mW Xbee [$19USD] with a wire antenna to recieve the heading variable.
Thanks for thinking along. Could you link to a picture of a C-band sat dish. I have not idea what that means. How big are those? Smaller is better in this case: I need to rotate the thing at, dunno, 1 rev per 12 seconds?

Depending on where you live, you may have dozens of them like in this area where no one now uses them and people pay you to take them down and away. Here is what a Feed Horn looks like. But if they are local junk in your area then I suggest a different tactic. Soup cans. You can take one or for more narrower beam 2 cans and mount the Xbee in the bottom center of the first one. Cut the bottom out of the second one and attach it so that you have a long narrow tube. That is how I make DIY directional WiFi trancievers out of standard WiFi routers.


I think I prefer the DIY Cantenna. I read some about them all over the net for the past few years. It seems the dimensions are pretty tolerant and you could create "beam" of approx. 30 degrees wide. Never read about anyone using them on Xbees though.

My concern at the moment is with the receiving end. Would the beam swooshing by offer enough time for the RX to listen to the data stream and understand it too? Or does the Xbee require some time to pair up with its communication partner(s)?

For a directiontal antenna, something called a Yagi might work, sort of a beam, with short wavelength based toothpicks spaced along it perpendicularly. The log-periodic is similar. The idea of getting a robot location from 2 successive readings as the robot moves is a good possibility (similar to SLAM algorithms).

All I want the receiving bot to know is on which "Radial" from the beacon it is. Give or take 10 degrees. So let's hope that we can make the antenna sufficiently directional, that it will send a 20 degree wide beam (no wider). Sure, narrower is better.



The first robot is on Radial R40 (bearing from the beacon). The second in this new picture is on R120. The R40 bot knows on which Radial it is because it receives a 1.11 ms pulse. Maybe it also receives pulses that are a bit shorter and longer, but I am not worrying about that right now. Call it margin of error. When the beacon continues its sweep of the area, robot "R40" will no longer receive any signal until the sweep returns. Compare a light house that changes its signature as its swoops around.

The antenna on the mobile bot need not be directional. I don't give a hoot about its direction relative to North. Just about which Radial. For now, that seems complicated enough.

What if the signal changed depending on where it was pointing. When is is pointing north it transmits 0 or 360, when pointing south it transmits 180, etc. This way when the signal is received it knows roughly where it is at according to the signal received. Of course it would receive multiple signals since 359 and 360 (0) are very close together. It would need to average out the signals it sees to have a more exact idea of where it is at in relation to the sender. If you knew at what rate per foot the signal beacon widened you coudl detect how wide the beacon is being received and approximate distance to the source. In theory the closer you are the fewer signals you would see as the sender roates, the farther you are the more you can see...

I guess your RX would need to rotate as well. You would need a bearing reading at both points rx and tx. If the robot has a omni-directional antenna, it will do it no good to know which way the tx is pointing. However, if the tx is south and the rx is north and they hit eachother, then the robot knows the beacon is straight ahead. Same if the tx is shooting east and the rx catches it west. Shooting and reading pulses seems to be the easy part. I've never heard of a directional antenna though, at least not like you would need here. I have only seen the big ones on boats and the like.

Just to throw it out there: Does trianglulation help? Could you have 3 beacons way out to the corners of your yard each spitting out a code. At that point you would just have to sense the strength of each signal.

Sure, more beacons provide more info. And combining the info will give you absolute position.

Alternatively (but we are entering chapter three already), you could combine different readings from one beacon. Use dead reckoning in between readings. I am not sure what the English translation is. The Dutch call it "your observations verzeilen in the map".Something like "Sail your observations in your map", but transitive.