Hello,

I'm working on a project to detect whether or not water is flowing through a pipe. Since I only need to detect whether or not water is flowing through the pipe, I'm thinking I could just use a piezoelectric sensor, amplify the signal, send it through a comparator, and maybe add some hystersis in order to hopefully negate false positives. My concern is that due to the simplicity of the design, there may be quite a few false positives. Also, it's possible that at times the water supply will be low enough such that air will get into the pipes and the water flow will not be constant but will be maybe 1 sec on 1 sec off. In such a case, I would not want the circuit to trip (ie say that there is no wate). I don't have much experience with piezoelectric sensors but am thinking that the above mentioned scenario woulnd't occur because intermitant water flow would cause sufficient vibration. Can anyone please provide thoughts on whether or not my plan seems reasonable?

 

Is this a school assignment (paper) or project (working device) or a personal project that you plan to build?

In any case, are we limited to just rigid pipe? Is it copper, PVC or other?

An easy way would be to make a coil of flexible copper tubing. The coil would have one extreme when empty seed unpressurized, another geometry when pressurized but no flow, and it would have an intermediate and moving geometry when flowing. You could use some cheap ir sensors to check position.

You could use pressure sensors to measure the same three states as well. Please clarify the constraints and degree of flow accuracy you need.

 

There are better options, in my opinion. The one used in my water softener is a little rotating "propeller" with an embedded magnet. A hall sensor detects the rotating magnet. I bring this up because it's probably the cheapest way to measure flow. If it wasn't, they'd switch to something else to cheapen it still more.

A piezo sensor does not actually measure flow, only vibration. There could be other causes of vibration, and I think you could have flow with very little vibration.

A pressure differential across an orifice is another tried and true method of flow measurement.

 

This is for a personal project. I'm limited to rigid PVC pipe. There are really only 2 possible scenarios: water flowing and pipe empty. When water is not entering the tube, the water that is in the tube will drain into a tank and the tube will be empty. In saying that, I really only care if there is water in the tube. I don't care about how quickly the water is flowing through the tube although when water is flowing in the tube it should be flowing at a fairly constant velocity except for when air enters the tube due to the water supply being low.

 

A fun, non contact, no moving parts possibility would be a capacitive sensor. Here is an example of how cheap and simple the circuit can be.

This particular IC describes some automatic self-calibration behavior that could be bad for you - it gradually adjusts to whatever capacitance is present and re-zeros itself, meaning if the flow condition lasts too long, it will eventually recalibrate itself to call that no-flow.

But there are fixed-sensitivity ICs for this too. I just don't remember off hand where I've seen them.

 

I think ebeowulf17's suggestion is the preferred method but you could also use an IR LED emitter and sensor pair. Water is a pretty good absorber at certain wavelengths and would break the beam. Enough to trigger a circuit, anyway. You'd need a "window" in the PVC.

You could probably even get a simple conductivity sensor to work. A comparator could detect the small increase in conductivity as two nails, for instance, become immersed in water.

 

Deciding factors might be water quality and mineral content. If it's pretty clean water with reasonable tds, either of those would be great. If it's dirty I'd worry about residue drying on the window and interfering with IR sensing. If it's hard water, I'd worry about mineral deposits developing on the probes.

Either of those fears could be handled easily with occasional maintenance work, depending on how accessible pipe interior is.

 

The new clear PVC makes an optical method easy.

http://www.amazon.com/Pipe-Schedule...id=1420776228&sr=8-12&keywords=clear+pvc+pipe

 

....There are really only 2 possible scenarios: water flowing and pipe empty. When water is not entering the tube, the water that is in the tube will drain into a tank and the tube will be empty.

A simple vane at the pipe exit can act on a microswitch to tell if there is water flow.
A float sphere inserted in the pipe can also interact with probes to differentiate filled or empty. Like using a PVC 'T' coupling, with a ball in its side branch.

 

This is for a personal project. I'm limited to rigid PVC pipe. There are really only 2 possible scenarios: water flowing and pipe empty. When water is not entering the tube, the water that is in the tube will drain into a tank and the tube will be empty. In saying that, I really only care if there is water in the tube. I don't care about how quickly the water is flowing through the tube although when water is flowing in the tube it should be flowing at a fairly constant velocity except for when air enters the tube due to the water supply being low.

Since you mention that the pipe will either be empty of full, and you don't really care if, or how much water is flowing, then perhaps a sensor based on the conductivity of the water itself would be most appropriate. Two stainless steel probes placed close together such that the water completes a circuit when touched.
The resistivity of the water can vary depending on how much minerals are dissolved in it, so your circuit might need to be adjusted for your particular water. You will need the electronics experts here to help with the actual circuit, but I would think a resistor, a transistor, a little trim pot and a few DC volts would work to power a little LED or a little buzzer.

 

Water or not water? How about a float switch?

Still wondering if this is for drip irrigation or filling an oil tanker with ballast weight.

 

Water flow sensing is a well-developed technology with sensors readily available in the market. Perhaps studying how they work might give you some ideas. There are many types, ranging from mechanical, ultrasound, capacitive and magnetic-inductive, among other less common types such as optical and even laser. I suggest you study most possibilities before deciding on which path to follow.
EDIT: after more careful consideration of your particular requirements, I'd probably go for the mechanical type, plus a small timing circuit that would tell you if flow has stopped for a preset amount of time.

 

How about monitoring pipe sag with a micro switch.

 

A fun, non contact, no moving parts possibility would be a capacitive sensor. Here is an example of how cheap and simple the circuit can be.

This particular IC describes some automatic self-calibration behavior that could be bad for you - it gradually adjusts to whatever capacitance is present and re-zeros itself, meaning if the flow condition lasts too long, it will eventually recalibrate itself to call that no-flow.

But there are fixed-sensitivity ICs for this too. I just don't remember off hand where I've seen them.

Would a cqpacitive touch sensor work with a 5 mm thick plastic pipe that has a .75" diameter?

I lied, there is a water but no flow condition but i dont care about flow rate. To give some more background i work in an orphanage in S America. One problem we have is our water supply is unreliable. When our tanks arent filling we need to close our main pipe and open up a different pipe so that we are only supplying water to our essential items. Unfortunately this takes a lot of time because the pipes are in a fairly remote location such that we have to walk through the jungle to reach them. I'm looking into the possibility of using a circuit to automatically close and open the pipes when our main pipe doesnt have water flowing through it. I've figured out how to do everything except detect the water.

 

Would a cqpacitive touch sensor work with a 5 mm thick plastic pipe that has a .75" diameter?

I lied, there is a water but no flow condition but i dont care about flow rate. To give some more background i work in an orphanage in S America. One problem we have is our water supply is unreliable. When our tanks arent filling we need to close our main pipe and open up a different pipe so that we are only supplying water to our essential items. Unfortunately this takes a lot of time because the pipes are in a fairly remote location such that we have to walk through the jungle to reach them. I'm looking into the possibility of using a circuit to automatically close and open the pipes when our main pipe doesnt have water flowing through it. I've figured out how to do everything except detect the water.

Could you please clarify? Do you want to detect the presence of water in the pipe, or do you want to sense flow, however small? Does the pipe completely empty when this condition you're talking about takes place? I might have a few ideas that could help you.

 

There are commercially available flow sensor switches. I see them used on small simple water well systems, so when flow is detected, the chlorination system is activated somewhere upstream.

 

Would a cqpacitive touch sensor work with a 5 mm thick plastic pipe that has a .75" diameter?

I lied, there is a water but no flow condition but i dont care about flow rate. To give some more background i work in an orphanage in S America. One problem we have is our water supply is unreliable. When our tanks arent filling we need to close our main pipe and open up a different pipe so that we are only supplying water to our essential items. Unfortunately this takes a lot of time because the pipes are in a fairly remote location such that we have to walk through the jungle to reach them. I'm looking into the possibility of using a circuit to automatically close and open the pipes when our main pipe doesnt have water flowing through it. I've figured out how to do everything except detect the water.

First things first. You said there is sometimes a water present, no flow condition. Just to clarify, you want this system to treat water present as good, regardless of flow. Is that correct? So we need a water present detection, not a water flow detection? If it's flow, then my suggestion would be awful.

Assuming it's just the pipe being full, or mostly full, then I think it would probably work. You'd want to look for capacitive proximity as opposed to capacitive touch. I think they're conceptually the same, and mostly represent differences in sensitivity, but some chips are optimized for one or the other, and some can do both.

I think that with a sensitive chip and a good size sensor, 5mm pvc would be ok, but you'd have to read the datasheet for any chip in question, and maybe do a little experimenting to know for sure.

I've done some reading on them for a project we want to try at work, but haven't gotten hands-on with them yet, so I could be overestimating the flexibility of this technology. I like that it involves no modification of the pipe, no potential for causing leaks, and involves no parts that will be corroded or plated with deposits from the water. On the other hand, it is electrically more complicated than the two metal probe conductivity approach mentioned earlier, which would be the other contender in my book - you'd just have to check and clean the probes every once in a while.

 

Use a small can with a float switch on it. Poke a hole in the bottom of the can. Place the can in the flow from the main entry pipe. When the water flow from that pipe can not keep up with the intentional leak in the can, the float switch will stop floating.

 

Use a small can with a float switch on it. Poke a hole in the bottom of the can. Place the can in the flow from the main entry pipe. When the water flow from that pipe can not keep up with the intentional leak in the can, the float switch will stop floating.

All the elements of the KISS principle in one paragraph...

 

I once worked for a company that made "bubble detectors," very similar in concept to your problem. Ours were intended for small tubes (say 1/4") to detect bubbles; bubbles are a bad thing when injecting fluids into people. The same concepts would scale up to larger tubes, which we did play with for a while doing proof of concept experiments for a flow rate sensor.

Sensing is done right thru the wall of the tubing as the tube walls and the liquid both will pass the ulreasonic signal.

Our sensors worked in the 3MHz range, which may respond very differently then say the usual 40KHz sensors. The low frequency sensors pass thru air very nicely, air stopped our signals dead which was interpeted as the bubble. I do not have a source for such ulrasonic crystal material but you may find some sensors around.

I once did some experiments using "projected capacitance" sensors such as ebeowulf17 suggested. To get a reading thru a vessle wall required a large ground plate behind a smaller sense plate held against the vessle. I could sense water thru 1/2" walls easily, but by that I mean I could vary an oscillator frequency to read it on a scope. Making a practable feild effected sensor is a reasearch project in itself.

I am not saying either method is bad, either could work, I would seek a complete turnkey package system. Sorry I don't have a link for that.

</two cents worth.