Buried wire fence
Hey all, well I am revisiting my lawnmower project, well not really concerned with mowing lawn but more of an outdoor autonomous robot for my backyard..
Anyway I spent ALOT of time with circuits over the past winter and with no real success. So to cut rapidly to the chase what I am trying to design/figure out is a buried wire fence type solution that I can put in my back yard with 2 sensors (left/right) on the robot and when it gets close to the wire it can either avoid it or follow it home depending on programming..
So I have read and tried endless circuits from the web and from my very limited electrical knowledge I was able to build some cool circuits that would detect a RF signal in the range of about 20Khz. The problem I kept having with all of them was I could detect a short loop of wire say 5 feet no problem at a distance of about 2 feet. However when I make the wire loop a bit bigger say 20feet then I practically have to touch the inductor coil to the wire to detect the signal..
Sorry this seems to be running on but now getting back to my point, I have used cable tracers at work alot and they work on even 1000ft of wire, so is there any reason why I can't use this type of solution? Are there any drawbacks? Also on a final note the production systems like pet fences use this continuous wire which has to make a loop back to the transmitter and a wire tracer does not require the other end to be connected, why is this? Would this cause me problems? Is there a better way that someone can help me with? Thanks in advance for any help..
@ Thu, 2010-09-23 04:05
Buried wire fence
When you posted this thread last time, I provided all the information you should need to get it working. And, as I pointed out to you last time, 20Khz is NOT RF, and a 555 has insufficient drive by itself to power an inductive loop. You need several watts for an effective system and you can't get that from a 555, plus the fact, 555's are not designed to drive an inductive load.
@ Thu, 2010-09-23 04:24
Salvage...
While I really do appreciate your help from the last time I posted I was unable to get that working.. I have found and provided information regarding the use of a 555 timer to drive a 2003 which I did build as described above.. After being unable to get a proper signal (strength) I have been experimenting with other solutions.. Which brings me to this post where I was asking the possible pitfalls of using a cable tracer type design with detecting a tone in the wire instead of RF or whatever acronym you refer to a 20Khz signal..
Again while I am thankful for your help I was unable with my limited knowledge of electronics to complete a working circuit. So it is my understanding that rather than continue down a road that I can't seem to grasp I was attempting to start another thread that would possibly give me a working solution to the problem.. I appologize if this seems to affend you but
@ Wed, 2010-09-22 08:05
so, you are driving the wire
so, you are driving the wire loop with a signal? and longer loops don't work? It sounds like your driver isn't powerful enough.
@ Wed, 2010-09-22 17:57
I agree..
Yes I am driving the wire with a square wave signal from a 555, and yes when I make the loop larger I lose the signal strength. Sorry I am adding this at the top as I didn't read your response correctly.. So with that said.
I agree that my driver isn't powerful enough. What I don't really understand is do I continually need to increase the power of the driver if I make the loop bigger? With that see below for a link to a simple schematic that shows the driver I built.. Those guys mention adjusting the power and resistance for your system, but I am confused with how much? If 12v and nearly 1amp only works for a 5 foot loop when testing then how much would I have to go up for a 500 foot loop.. Also is it amps that would make the difference or volts?... Anyway if you have a chance take a look at the schematic and let me know what you think..
http://www.robotshop.se/micro/robocut_us.html