I need to build an underground water leak detector. Can anyone offer any suggestions for a home workshop project?
Assuming that a sensitive microphone listening for hissing/gurgling sounds is the easiest type of project, what are some guidelines for microphone sensitivity, frequencies of interest, filter parameters, etc?
What kind of microphone would best fit the task? How to mate it to a metal probe? How much total audio gain is appropriate?
Actually, that sounds like an excellent bit of research for you to do! You may be able to try out different, existing microphones and methods of mating them to the ground (I assume a standard copper grounding rod might work?) and trying out different situations to collect data. But that seems a bit more like putting the cart before the horse, to me.
You might start out, if you want to avoid building anything at first, with some quality sound recordings of various hissing and gurgling sounds you feel are appropriate and just analyze those as a start. Let that take you where it does and only then, after you think you've got that nailed down okay, start asking yourself what kinds of transducer and detector algorithms make better sense. Then go out and test, once you have your own idea about what you expect to gather and observe.
Looking at the windowed data via FFT would seem reasonable, to see what you could observe as common factors. For some vague reason, though, I think LPC might be more useful in the end. Anyway, it might be complex, or quite simple. But hissing/gurgling seems recognizable to people, so I imagine there are some characteristics you should find and algorithsm that will work reasonably okay.. though never entirely 100%.
Also, what's more important? Avoiding false positives? Or avoiding false negatives? You will have both, but one is often worse than the other for applications.
First thing to do is DEFINE/SPECIFY the problem. How far underground? What is ground? rock? sand? other? How big is the water leak? Volume? Pressure? An "ooze" might take a different detector than a "high pressure" leak. Most "ground" is wet even without a leak! What else is in the ground? Clear ground might take different tools than if in a tunnel with dozens of sources. Access?
Simplest solution that satisfies your (lack of) specification is a shovel. Most exotic might be a dowsing rod. No electronics required for either.
Most important part of any project is the specification.
How deep? You can get moisture probes that check to see if your potted plants need watering, and ones that can tie into your sprinkler system so it doesn't turn on when it isn't needed.
In the 80's I designed one of these for a plumber with his input suggestion of using a ultrasonic microphone based receiver. At the time I thought it would not work but he insisted. As I recall it was just an audio amplifier I used, nothing special. When I first tried it I put the microphone on the ground and used headphones. I heard some thumping and thought I had some oscillation or power supply problem. What I was actually hearing was my dog walking on the lawn 20 feet away! By turning on a faucet in the house we could trace the water pipe under ground by listening. The plumber often later told me how well this device worked with real leaks.
Realizing that many leaks start out small, seeping and don't make an audible signal at all... Only after you have a major failure would you have that type of signature, and even then it may not be present.
Soil conductivity measurement may be more like a workable solution. A microphone will pick up everything other than what you want to hear, until you have gallons per hour leaking...
I forget who posted the top link, but this is interesting:
formatting link
Dig down and you can find specs. Because the pickups are spec'd in V/g I suspect that they are some sort of direct contact device (actually a high bandwidth accelerometer) rather than an open air microphone. The closest thing easily available might be a phono pickup with the needle resting on the ground contact plate.
This kind of design will eliminate attenuation due to the boundary between the ground and air that you'd have to deal with using a microphone.
--
Paul Hovnanian mailto:Paul@Hovnanian.com
------------------------------------------------------------------
In theory, practice and theory are the same, but in practice they
are different -- Larry McVoy
No, not at all.... I said that I assumed that a mike that listens for the sounds is the easiest to design and build, especially for a quick, fairly easy project. Certainly, if there's a better, easier type of circuit, I'd be more than willing to consider it.
Pipes are around 12"-16" under the surface. Garden moisture types aren't suitable... especially this time of year. The rains have the soil pretty wet already. The area of scrutiny is in an open field with a pipe run of about 800 feet.
My plumber brought in a locating company that thermal-mapped my floor to locate a slab leak. ...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
Remember: Once you go over the hill, you pick up speed
I have an ultrasonic leak detector. It's an ultrasonic oscillator and mixer that mixes ultrasonic down to audio frequencies you can hear. Very sensitive on high pressure leaks that produce ultrasonic frequencies. Useless on a drip.
As I said in a previous post, the soil is really wet from winter snow and rains, so conductivity tests, to me, wouldn.t give useful results (unless you know something I don't?). I've looked at several sites selling water leak detection equipment, and the acoustical solution seems to be the standard. There are several that have a microphone mated to a rod, which is inserted into the ground to listen to the sounds. Some have selectable filters. Some have headphones, some have analog meters, some have have an LCD (graphical?) display to present the results. Of course, there's ground penetrating radar, but that would be a bit over my budget.
I'm going to try an electret mike fastened to a short piece of copper ground rod as the sensor, connected to a high gain audio amp with headphones. I have a few opamp-based audio filter modules that I might try just for fun. If all that fails to give results, I might look into an ultrasonic converter like Herman suggested. Something tells me that might ultimately be the best solution. The source is a municipal water supply, so it would seem to me that some sort of high-pitched hiss or gurgle could be heard.
Conductivity might still work, if you filter the longer term changes due to the weather, which would produce a long time constant bias. Short term changes may be detectable as a spike against this.
You could also detect pressure drop across the length of the pipe...
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.