SparkFun Forums

[Shawnee6d]

Hi all, First, I'm just starting out. "Discovered" Arduino about a month and a half ago and have become so addicted to physical computing it makes a heroin addiction seems like a mild craving for something salty in comparison.

I'm a self taught software engineer with a 2 year pre-engineering degree, all of my electronics knowledge is self-taught as well and is scattered all over the place. One cool aspect of Sparkfun's site and Adafruit's is how their tutorials end up tying together all of the various bits and and pieces I've picked up over the years.

Onward. Having gone through some of the basic Arduino tutorials, I've set my sights on what seems to me a fairly ambitious project. I picked up one of Spakfun's pan and tilt brackets, got the facial reco tutorial working more or less and no want to go wireless so that I can basically mount a few camera's around my property to keep an eye on things for me and make it possible for me to focus on refining my OpenCV code to to help the camera's do a better job of focusing on looking for humans and not deer, stray dogs, racoons, etc.

I've done a fair bit of browsing and forum reading and it looks like most wireless work is currently focused on small RF Transceiver modules that have a fairly limited range. The front of my driveway is about a quarter mile from my home. Ideally, I'd like to be able to breadboard together an RF transceiver or transmitter and receiver separately from discrete components of sufficient power to reasonably transmit and receive data out to one mile (so that I can use the same method to control the cameras, transmit and receive video info, and handle gate open/close functionality, etc). I figure if I can put together the hardware that transmits and receives reliably enough, I can then shift the problem more into the software domain where I feel much more confident.

I have found several sites that have all manner of schematics and plans for various built up transceivers, transmitters, receivers, etc but honestly, I'm not sure about some of the more fundamental topics to be able to confidently select one over another. ie, is there an FM or AM hetero-dyne frequency that is better suited to transmitting data at this distance? I haven't been able to find a project site or forum post that would indicate anyone else had tried this, but I'm sure someone somewhere must have. To be sure, there are commercial units I could buy that would likely do the trick, but I wouldn't learn near as much and I'd like to be able to tackle this from start to finish without resorting to one of those packages. AS I understand it, there are frequencies I can stick to that keep me out of harms way in terms of FCC interference and such, too I live in the middle of cascades and I'm reasonably sure that if I keep my transmission strength down to just what I need to hit the half mile/mile mark I'm very unlikely to bother anyone, regardless.

In any case, it'd be great to read what some of you folks have to suggest, perhaps a different forum specifically about this sort of thing, any gotchas I should be looking for, etc. Has anyone ever tried to create a built of transceiver pair for transmitting this kind of data volume with? Video ends up being quite a bit of data, even compressed. Looking forward to reading you folk's thoughts on the matter,
John West

[Joeisi]

How do you propose transmitting video?

[stevech]

I'd like to be able to breadboard together an RF transceiver or transmitter and receiver separately from discrete components of sufficient power to reasonably transmit and receive data out to one mile (so that I can use the same method to control the cameras, transmit and receive video info, and handle gate open/close functionality, etc).

Way too difficult to do this starting with discrete components. Start with modules.
Send video using cheap analog sender/receiver - or a baby monitor.
send data one mile - do-able in the 902-928MHz band (No. America) with near line of sight and the right antennas, and something like Digi XBee 900Mhz band and the right gain antennas. Don't attempt 1 mile with 2.4GHz stuff.

[Duane Degn]

I agree with what stevech said.

If you want to combine video and data you really need a computer at the sending end.

I'd suggest two separate signals, video and data.

For data, a pair of XBee XSC modules would likely work at a mile if there are too many obstructions.

https://www.sparkfun.com/products/9085

I have a pair of XBee XSC modules but they wont communicate a block away from each other if the space between them is full of houses. They work fine a block away with just a few trees in between the two units.

SuperCurcuits.com has some video transmitter/receiver sets. I couldn't find the set of 900 mHz units I purchased from them.

Here's a 900 MHz transmitter:
http://www.supercircuits.com/Wireless-S ... 900-MINI-A

It doesn't look like it would work a mile away though.

I think they do still sell the receiver I have.

http://www.supercircuits.com/Wireless-S ... s/AVX900R1

Another source from wireless video is:

http://www.securitycamerasdirect.com/br ... deo-links/

Again, I didn't see a video link that would likely work that far away.

DealExtreme also sells some video transmitters/receivers but I don't know if they would work at the ranges you're after. I think you might need some sort of directional antenna on your video transmitter.

I haven't check Amazon's selection of video transmitters/receivers but it seems like they sell everything these days.

You might need to check out a licenced (ham) solution. There was some discusion lately on the Parallax forum about quadcpter cameras that included wireless links.

http://forums.parallax.com/showthread.p ... ost1111935

There are probably better video link sources in the thread I just linked to than those I listed above.

[Shawnee6d]

How do you propose transmitting video?

Well, I want to do it digitally, but it may make more sense to transmit the video as an analog signal and then deal with converting the analog video info into a digital image at the PC side of things.

The more I think about it, the more it seems to me that trying to send the image as digital data might be impractical.

I want to at least try and build up a much less ambitious transmitter/ reciever and play a bit with trying to communicate data in a serial fashion so that I can understand more about what's involved. I'd guess that once I have some kind of understanding of what's involved I'll most likely go along with what seems to be the mass suggestion here and use some rf module and likely a separate module to transmit video some how to a PC.
-john

[Joeisi]

Well I had thought about doing a live video feed over X-BEE on a quadcopter awhile back. It is possible, but the resolution,color, and range will be sub-par. You will probably need a dedicated uC(pretty fast) and dedicated X-Bee. Even with that, the quality may not be good enough to be used for recognition. You would probably want to go with either a hard wired solution or a built for video transmission transmitter and receiver pair.

[stevech]

XBee (2.4GHz) is IEEE 802.15.4 which by definition is a 250Kbps burst bit rate and a net yield of about 80Kbps with overhead.
Wrong choice for streaming video.
Best done in analog form. Far simpler. Even a venerable old X10 cam w/sender/receiver - dirt cheap.

[Joeisi]

XBee (2.4GHz) is IEEE 802.15.4 which by definition is a 250Kbps burst bit rate and a net yield of about 80Kbps with overhead.
Wrong choice for streaming video.
Best done in analog form. Far simpler. Even a venerable old X10 cam w/sender/receiver - dirt cheap.

It is, but possible.

[Joeisi]

In addition to what was said earlier with the 900 MHz radios: Use directional antennas at that distance. Since both positions are fixed, you can also fix the antennas. If not in line of sight of each other, you may want to use some sort of repeater. Then again it depends on your radios power if you even need a repeater or even a directional antenna.

[Shawnee6d]

I hadn't checked back in a short while. Just wanted to say thanks for all of the replies. I had thought to put together a descrete transmitter reciever to get a basic sense of how data transmission might work, then form there move to more mainstream methods. I might still do that, but in the interests of getting this project done in my lifetime xciever modules it is, most likely xbee's, with likely one or two repeaters mounted at opposite ends of my property so line of site isn't as much of an issue. As for video, that'll likely need to be a sepereate signal transmitted directly from the camera, possibly with a bought unit, not sure there yet, but it appears transmitting analog and then worrying about converting to video at my pc end is much more do-able than converting to digital on my remote units.

I do have the serial control portion worked out. That is, my proto-type unit can now scan or, if I want to take control from my PC, I can do that as well to position it to something more interesting. Right now I'm doing all of my video/reco work on my PC which is ok, but not really suitable for the final version so I may need to move to a more powerful processor/more memory, etc I want the remote units to be able to do their job autonomoisly, but also be able to be coordinated via PC as well. Load sof fun work to do, but as things get seperated into small projects, it's becomeing quite do-able:)