I am trying to build a robot lawnmower(yes another guy trying this).. I need to understand Inductive guidance or RF technology a bit better to build a buried wire circuit for this.. I have read alot on the web and am starting to understand inductor/capacitor circuits.. I built a transmitter using a 555 and a 2n2222 to raise the current in the wire.. I then built a L/C circuit with a 339 comparator and I can tune it to pick up the signal from the wire(I can see this on my oscilliscope)..
My question is that I seem to have alot of noise as well and even if I touch the inductor I see alot of noise from that as well.. Is there an electronic way to filter this out? Did I make my circuit wrong? Is there a better way to accomplish what I am trying??(maybe using am radio signal in the wire instead of RF)?? Please any information on this topic would be greatly appreciated as I have invested ALOT of time and money trying to solve this issue..
Frequency
Inductive loop systems have been around for many years and they are pretty simple. For a long time they were used in churches for the hard of hearing. A simple coil, (just a few turns), went around the church. An audio amplifier, (5 watts is sufficient), is used to drive the inductive loop with the program material.
Small receivers were given out to people needing the “extra volume”. The receiver was nothing more than a small audio amplifier with an inductive loop as the input. Ideally, you need to have some active filtering, on the input to limit the frequency response to 500-5,000hz, otherwise, the 60hz all around us will make the units unusable.
For your application I would shoot for an oscillator at about 5-10Kc and about 5 watts of drive. Use some multiconductor wire for your loop and cross-connect unlike colors at the open ends so you have multiple turns. The receiver can be most any loop of wire, (experiment for best S/N ratio), feeding an op-amp for gain, an active notch filter, and then drive a PLL like the LM567 tuned for whatever frequency you’re using. You may want to make a “sensor deck” with 3 or more sets of sensors for proximity detection of the sense cable. You should be able to build the whole system for under $20.00.
Hope this has been of some help and good luck with your project…
This is not even my project but…
I have to say that was one of the best answers to any question I have ever seen around here. The information density in what you have written above is staggering.
Well done.
THANKS!
Chris,
That is high praise coming from you and I appreciate it.
I’ve written a number of articles for our on-line robotics club, http://robojrr.tripod.com/ you might want to check out. The information is somewhat dated now but still good stuff. Our series on basic electronics is especially good.
**Thanks… alot… **
Hey Salvage , I really appreciate the info you provided… I will be honest I am an extreme beginner when it comes to electronics but the good news is I can read, hehehe… So I am taking the info you posted and going to google to try and comprehend it all a bit better… If possible could you draw me a schematic putting it all together… Thanks again for the help so far…
**Questions… **
Salvage, sorry I posted my reply above yesterday instead of replying to your comments…(forum newbie). Anyway I have done some initial reading and I have a couple of questions…
1. For the transmitter you say I should use multiply loops of wire, the dog fence designs as well as all the other inductive guidance type systems that I have read about only use 1 wire… Is your suggestion to increase signal strenght or for some other reason?
2. The ‘recievers’ I have been reading about and trying to build all seem to use a inductor and cap, from your comments above I am interpreting them to say that I should just use a ‘loop of wire’ (or maybe a inductor?) as an antenna and then use a op-amp to raise the signals recieved from that, then I would use a LM567 and tune that around the frequency I am injecting into the transmitter wire?
3. I have only read breifly about a notch filter but I am not sure why or where I would use it in the design?
Again thanks for all the help so far… I really appreciate it…
Questions
The more turns you have in your loops, the greater the signal strength and the better the signal to noise, (S/N) ratio. If you use a multi-conductor wire as I suggested, it isn’t any more trouble than running a single wire. The reason the dog fence designs use a single wire is because they are RF and you said you were working on inductive loop. They are completely different.
You need a notch filter to attenuate the 60hz you are going to be picking up on your sense loop. You mentioned in your original post that as you got your hand close to your sense loop, (or touched it), you could see an increase in signal, (noise), on your scope. This is because you are seeing the 60hz AC being picked up by your loop. The more 60hz you have mixed in with the signal you really want, the harder it will be to seperate out the desired signal.
If you “Google” the LM567 data sheet, it will give you the complete schematic for a tone receiver. It also gives you sample values for different frequencies. The LM567 is simple, cheap, and easy to work with…it will tune over a 10 to 1 range with just a variable pot, (I suggest a multi-turn).
Also, remember that your transmitter and receiver circuits must be powered by a regulated power supply so the frequency will not drift. Good luck.
**RF vs inductive loop… **
Ok again bear with my rude determination of the terminology… Taking a step backward, should I be heading toward an inductive loop type of solution for this project. Or should I be trying to better understand RF as that is what most of these designs seem to be using for this purpose?
Questions
I guess the first question you need to answer is, “What exactly am I trying to accomplish with the system?”.
If you are only wanting to set boundarys around your yard so the mower does not take off running through your neighbor’s prized tulips, then either system would probably work. You indicated you are a novice, (this is a pretty ambitious project for a novice), and reliable RF design is much more difficult than inductive loops. It also requires more sophisticated test equipment and you have the pesky little issue of meeting FCC type acceptance guidelines.
Whatever method you decide to use, keep in mind that “good enough”, isn’t when you are talking about a heavy and dangerous motorized blade wandering around your yard by itself. Your failsafes must be 110% reliable so there is no possibility of injuring pets, children, or yourself. Your neighbor may be the nicest guy in the world until your mower runs over his cat.
**decisions decisions… **
First off thanks for taking the time to lead me through some of this stuff as I really have invested a lot of time and money trying to better understand some of this stuff… I actually have a fairly small fenced yard but with some flower beds, so I am in no hurry to add any spinners… hehehe , it is more for the fun of the project than the final result… However I don’t really want to encounter the wrath of my wife if I bulldoze over some flowers with spinners or not… hehe…
Now more seriously back to this… I am really just looking to make a perimeter wire fence much like for the dog… I actually bought one from ebay for this, but I can’t seem to figure out exactly how it works… Yes I probably could invest more in collars for sensors and adapt them but again it is more for learning… Below is a couple of links of what I had in mind. I know the one is more inductive guidance than perimeter but I believe it would be roughly the same technology… Anyway if you have more time please take a look at the links and that should hopefully explain the type of technology I am looking at…
Line following car - http://forum.allaboutcircuits.com/blog.php?b=328
DIY lawnmower(using a dog fence) - http://sites.google.com/site/diyrobotlawnmower/
Robocut - http://www.robotshop.se/micro/robocut_us.html -schematics farther down the page
As a footnote I have tried to construct some of the circuits from the links above but this is what prompted my post is that I seem to get alot of interference and I can’t acurately detect the wire…
Sorry
The third link with the schematics is a dead link.
**Updated link… **
Here is the main link…
http://www.robotshop.se/micro/wwwrc_us/indext.htm
For the schematics click on diary on the front page then scroll down to the midway point … It is not so much a schematic as a concept drawing though…
Not RF
In looking at the schematic and reading his description, it is not RF, it’s 38kc. So, basically, it’s still an inductive loop design with a range of only 50cm.
**it’s not RF… **
I thought you told me that for it to be an inductive loop it would have to have multiple turns?? Now I am confused…Does it have to be a certain frequency to be RF?
it’s not RF
I said the more turns you have, the stronger the field and the better the signal to noise ratio.
Normally, anything over 100kc is considered RF, though at that frequency the circuitry looks more like audio than RF.
**I see… **
First of all what do you think of that circuit/system? Will it work properly (in your opinion)?
Oh ok… that makes sense… So if it is running at 38khz and I were to use your other suggestion about multicolor wire that would increase the signal strength right?
I presume the capacitors that he has marked with an * are the same value, and I can calculate the values from an online L\C calculator correct?
One other thing can the signal strength on this circuit be increased by any other means? (ie incease voltage or current in buried wire)
