Water Level Detection

[Boncuk]

Hi tariq,

you need only one switch if you use two floats connected to a cord (rope). Make them adjustable along the cord according to the desired ON/OFF water level. The lower float will determine "pump on" position pulling the switch ON if the water level drops beyond a certain point. The upper float will release the switch as soon as maximum water level is reached.

For the function refer to the attached sketch.

A) water level well below minimum. Both floats (filled with sufficient water to overcome the spring force of the switch) pull the cord. (switch on)

B) water level above minimum, but upper float still pulling the switch. (on)

C) water level at maximum, no more force on the switch, cord released - switch off

Additionally this method doesn't bear the possibility of electric shock or electrocution since the switch won't get in contact with water.

Boncuk

 

[tunedwolf]

There is another method that nobody has mentioned yet, MIR or Microwave Impulse Radar.

This was a spin off technology from a related LLC project a few years ago, which had not quite worked out as planned. LLC also wanted a gazillion dollars plus a shot of your wife whenever they felt like it just to get involved in the licensing talks!

Eventually there were a couple of sensor companies that took up the license, one of which was Bindicator if I remember correctly.
However as I also remember, LLC used the term "Prototype development idea" to escape any comeback from it not working properly as described, despite it's costs. I also remember hearing of various law suits being mentioned about a year or two after I walked away from it. I think Bindicator did eventually get the idea working properly though, but I have no idea how or what they did to achieve it.

It's operating principal was basically that a narrowband, low power signal was bounced off a launchplate at the top of a stainless steel or titanium rod, or weighted cable, and a TDR reading taken, this gave a starting time, or top of tank reading. The next largest reflection seen would be that of the liquid level, the next largest would be the end of the sensing element, or bottom of the tank. The time taken between the first two readings or last two readings was proportional to the height of the liquid in the tank. There were other claims too, in that each device used a random gating signal so that each "chirp" sounded sufficiently differently to allow more than one device to be used within the same location. There was a very small dead band at the bottom of the sensing element, but that didn't matter to my application.

All of the methods I had looked at the time had drawbacks that made none of them suitable to my application. I had varying temperature of and aggressively corrosive liquid, build up of scale, high turbidity etc. MIR seemed like a perfect solution as it was allegedly, impervious to these problems.


rgds

 

[premkumar9]

Basically there is no difference, except with the water column inside a tank and the hose reaching to the buttom you will measure pressure, with a lifted hose you will measure vacuum.

My doubt is
take two cases where we use flexible hose of 25 Ms length.
In case one after filling it with water I close one end air tight and leave the other end in water in the tank. Then lift the closed end to 5ms height.

In the second case everything is same but I lift the closed end to 20Ms height.

My question is whether the tube will remain filled with water in both cases.

What I thought was in case1 the tube will remain filled and in case 2 water will be filled up to 10Ms height (as the water barometer) and remaining height of the tube will be vacuum.

 

[Boncuk]

In both cases the trapped water won't escape through the end of the hose.

But why make measurement that complicated?

Stick your hose into the tank all the way to the buttom after the top has been closed (by the sensor port) leaving the negative pressure port open for atmospheric pressure compensation.

It doesn't really matter if you measure vacuum or pressure as long as they are equal, but the latter is easier to do and is always accurate.

Moving a water filled hose up loosing it's reference point the measurement will be true only once - the level when you filled the tube to the actual water level in the tank. There won't be any change with changing water level. You might just dunk the open end of the tube in a glass of water and lift the entire assembly 100m above tank level. As long as the water is trapped there won't be any change in pressure (vacuum).

Get that out of your mind. You're much overcomplicating things!

Boncuk

 

[premkumar9]

In both cases the trapped water won't escape through the end of the hose.

But why make measurement that complicated?

just for academic interest I am asking. Just wanted to know whether my statement on case 1 and 2 is true ( in case1 tube will be filled with water and in case 2 , beyond 10Ms height it will be vacuum)
Please don't relate this to sensor or pressure measurement.

 

[kinarfi]

Just a thought, if you put a closed end tube in water, over time the gasses in the tube may be absorbed by the water and displaced by the water and eventually, the water level in the tube will be the same as outside the tube. Same for vacuum, the the water will give off a gas and the water level difference will go to zero.
Again, I suggest the bubble tube, it will always have a pressure equal to the the depth of the water and the pressure gauge can be almost anywhere
kinarfi

 

[Mr RB]

What about a capacitive sensing system? No moving parts, and it should work ok with water. Two flat plates in the tank vertically, running top/bottom, say like 2" x 1/8' alloy flat bar you can get an any hardware store with a thin plastic separator. Then coat the whole thing in any waterproof varnish, polyurethane etc they use on floors is thick and cheap.

Then just connect it to an oscillator, (or PIC as in the Microchip cap sense datasheets) the output period should be almost linearly based on the amount of water in the tank.

 

[premkumar9]

What about a capacitive sensing system? No moving parts, and it should work ok with water. Two flat plates in the tank vertically, running top/bottom, say like 2" x 1/8' alloy flat bar you can get an any hardware store with a thin plastic separator. Then coat the whole thing in any waterproof varnish, polyurethane etc they use on floors is thick and cheap.

Then just connect it to an oscillator, (or PIC as in the Microchip cap sense datasheets) the output period should be almost linearly based on the amount of water in the tank.

Will look in to this possibility also. Thanks.

 

[premkumar9]

just for academic interest I am asking. Just wanted to know whether my statement on case 1 and 2 is true ( in case1 tube will be filled with water and in case 2 , beyond 10Ms height it will be vacuum)
Please don't relate this to sensor or pressure measurement.

Boncuk may please give his comment on this as he is an expert in this area. I would like to clear my doubts before going for pressure sensors mounted above the water level.. It is difficult for me to conceive that in the case of a water filled tube with open end kept in a water tank, the water will remain fully filled in the tube even if the closed end is lifted up to any height. If that is the case, what is the difference in the case of a water barometer?

 

[premkumar9]

Boncuk may please give his comment on this as he is an expert in this area. I would like to clear my doubts before going for pressure sensors mounted above the water level.. It is difficult for me to conceive that in the case of a water filled tube with open end kept in a water tank, the water will remain fully filled in the tube even if the closed end is lifted up to any height. If that is the case, what is the difference in the case of a water barometer?

I have been waiting for the comments. If you say that water will reach to any height, pls tell me the difference between this case and a water barometer. Otherwise, I may have to conclude that to measure the pressure, the pressure sensor has to be at the same height where we have to measure the pressure (it can can be fixed outside the tank also with a connection as mentioned).

 

[Mr RB]

If the tube is completely filled with water and you lift the top end, there will be a negative pressure at the top. Once it get to about 20 feet (about 0.6 bar or so?) it's gonna boil even at room temperature so the top part will turn into water vapor. It's the same reason that water pumps have to pump water uphill, they can't suck it uphill more than a few feet.

 

[colin55]

Get a sensor out of an old washing machine from the tip.
It will cost you nothing.

 

[premkumar9]

If the tube is completely filled with water and you lift the top end, there will be a negative pressure at the top. Once it get to about 20 feet (about 0.6 bar or so?) it's gonna boil even at room temperature so the top part will turn into water vapor. It's the same reason that water pumps have to pump water uphill, they can't suck it uphill more than a few feet.

So what is the answer to my question. Can I keep the pressure sensor at any height to measure the water pressure at the bottom of the tank (by connecting the sensor and bottom of the tank with a water filled air tight tube)

 

[premkumar9]

Get a sensor out of an old washing machine from the tip.
It will cost you nothing.

This is not for a hobby project. We have to supply so many numbers to a customer.

 

[Boncuk]

Hi premkumar,

I guess measuring water level applying capacitive sensing will be maintenance free, but I doubt it will be more accurate than pressure sensing, which might require maintenance from time to time. (as every technical device does.)

Capacitive measuring burries the uncertainty that modifications around the water tank might lead to erroneous readings of the true water height.

Differential pressure sensing is always true as long as the measuring tube is sealed air tight on the +P side of the sensor which applies using PU-hosing.

To make it absolutely safe and maintenance friendly you might use a manually shifted three-way valve to blow compressed air down to the bottom of the water tank and then switch back to normal operation. There is no danger involved for the sensor.

The MPX2050 can stand a pressure of 200KPa, which won't be exceeded even blowing at 10 bar (10,000KPa) pressure through the hose. At a pressure inside the hose corresponding to water pressure (49KPa at 5m) the air will vent through the open end of the hose at the tank buttom.

Water won't normally take up oxygen (no chemical reaction) out of the atmoshere, so the pressurized +P port should always represent the actual water level compared with local air pressure.

Differential pressure sensors are reliable and relatively cheap. What makes the pressure transducers expensive is the conversion to electric output, which you can do yourself using an appropriate instrumentation amplifier.

The MPX family already puts out a linear and temperature compensated voltage. You just have to take care of proper zeroing and amplification.

Attached is the principle of pressure measuring in a water tank.

Boncuk

EDIT: My recommendation: Use an MPX2050 (max water height 5m) and an INA111. If you correct for amplifier offset (max 500µV) in the software you might omit an extra OpAmp for offset compensation.

 

[premkumar9]

This attached fig makes sense to me. The green colored area is compressed air. So this air column will transfer the pressure at the bottom of the air column to the sensor diaphragm. Isn't it?

 

[Boncuk]

If the tube is completely filled with water and you lift the top end, there will be a negative pressure at the top. Once it get to about 20 feet (about 0.6 bar or so?) it's gonna boil even at room temperature so the top part will turn into water vapor.

Where did you learn that?

Water normally boils at 100deg/C at sea level (1013.2mb) and 25deg/C ambient temperature. It starts "boiling" at altitudes of minimum 25,000feet (8,000m), where the ambient air pressure is about 1/2 of that at sea level.

There are people living in Chile at elevations of 15,000feet. If you want to go there you better use an oxygen diluter equipment. Those people were born there and can stand breathing "thin" air.

If you want to learn about "boiling" water at high altitudes visit your AF-medical center. They should have a high altitude simulation chamber there.

20 feet is as much as nothing, since it's still positive pressure compared with the earth atmosphere - just slighty decraesed.

Have you ever heard about people living in a skyscraper having starved of boiling blood?

Boncuk

P.S. My water pump "sucks" the water from 80m depth (~266feet). The trick: Fill the suction pipe with water and use a check valve at the bottom to always have the water column up to pump intake level.

 

[Boncuk]

This attached fig makes sense to me. The green colored area is compressed air. So this air column will transfer the pressure at the bottom of the air column to the sensor diaphragm. Isn't it?

Exactly that, and it's accurate! No matter how high you lift the upper end of the "pressure" hose the result won't change.

Remember, I posted you might put the sensor on the moon and connect the positive pressure port to the tank and the negative port to earth atmoshpere.

It will result in equivalent water pressure in your tank.

The only problem which might arise is leakage.

If properly sealed nothing can go wrong.

Boncuk

 

[premkumar9]

Do you expect any problem if I use the conduction property of water to sense water level ? I mean, when water comes between two conducting plates connected to a DC supply through a resistor, it will cause a voltage drop across the resistor.

 

[Boncuk]

You could of course make a cheap circuit using conductivity of water to measure the tank level, but you'll not be able to measure any level like it can be done using a pressure sensor.

To measure at five different levels you must have 5 probes and a return (common) probe.

Electric current through water is also involved with electrolytic effects on the sensor wires. Copper will slowly disappear caused by it. You might minimize electrolytic effects by applying AC to the probes.

This means rectification and amplification to get a valid water contact indication.

You might use stainless steel rods but possible lime deposits make the steel rod an isolator (sooner or later).

Anyway measuring water level using voltage and current is not long lasting.

Unless you intend to upset customers you better forget about that idea.

A pressure sensor is approximately US$15, a TL074N is 15Cents. To make the circuit work 100% accurately you will need a symmetric power supply of ± 6 to 12V and not more than 8mA.

This solution seems the best and worth putting US$10 more into the single unit than sending repair men to the customer every once in a while.

Controlling water pumps using water conductivity is not a safe way at all from my point of view.

Living in Germany I'd reject doing a project like that (wires in the water), since I have to guarantee 10 years (by law) of proper function (without maintenance)

Going one further step, I guess the device would be a good seller for rich people in India who can afford to buy the sensor + a digital display of water height with the accuracy of 4.88mm (total water height 5m) using any MCU with 10bit A/D converter.

Boncuk