Looking for liquid level switch

stolly · 21st July 2012, 7:33 AM

-First post, so not sure if this is in the right section or not, I am trying to source a simple, low cost, reliable liquid level sensor / switch. i want to build a rod / tube / pipe that will house up to 30 sensors that will mount inside a water tank and set it up so that as the water level rises / falls inside the tank, a corresponding LED on the exterior of the tank will light up so that you can easily know the water level inside the tank. - Needs to be 12V DC
If anyone has any ideas / advice would be much appreciated!!
Thanykou

mrbean_phillip · 21st July 2012, 8:27 AM

Hello Stolly,
Years ago, at a Water Treatment Plant at a Power Station I worked at, we used a similar principle to measure the level in some vessels.

What would work probably very well for your application, would be a simple setup of 'conductivity' probes, i.e design a solution that will have the tank as 'earth/negative' and you imagine these 32 probes mounted on a piece of non-conductive material.

Stand it vertically inside the tank, and as the water rise, it will close the 'switch'

On the other side, you can wire the switches in to a simple Digital to analog converter, there will be many circuits about, and use the analog voltage, derived from the inputs....

Ok, a bit elaborate, but will work well, and be stepped-based, not fully analog.

The best would be to source a level transmitter, which could be based on either differential pressure, or a single pressure measurement principle, and use the 0/4-20mA output as the indication, or going into a controller.

You can connect a 250 Ohm resistor in series with the circuit, and measure the 0/1-5V drop across if that makes matters easier.

If you can tell me what you will do with the signal, I can look at providing you with better feedback, and maybe link you to a few devices I think might work.

Last thing you realy want is a mechanical/electrical device, i.e floats inside a tube et al, as they always seem to get stuck somewhere near the top of the tube, as the level drops.

Enjoy, and let me know if you need further feedback on this.

Kind regards,
Phillip

zero_velocity · 21st July 2012, 8:33 AM

Jaycar sell a very cheap and easy circuit for this sort of application, it uses 2 wires, when both wires are connect, an LED will light up on the circuit. To use this circuit all you need is to have one wire slightly below the other in the tank and when the water covers both sections it will create a circuit and voila.

You could also go down the arduino solution - the code for it and everything you need for it is available online. It operates on the use of a pressure transducer to gauge the level of water in the tank, all you would need to do is install the unit in the bottom of the tank in a waterproof enclosure and use an ethernet rj45 patch board on the unit to allow communications to the unit regarding levels of the tank

Bastard Child · 21st July 2012, 8:38 AM

If the water's not treated then algae will fuck with your system, be it a float or conductance pads.

Go to siliconchip and search 'water level'.

http://www.jaycar.com.au/productView.asp?ID=KC5503

mrbean_phillip · 21st July 2012, 8:55 AM

Quote:

Originally Posted by Bastard Child

If the water's not treated then algae will fuck with your system, be it a float or conductance pads.

Go to siliconchip and search 'water level'.

http://www.jaycar.com.au/productView.asp?ID=KC5503

Hence my suggestion about the option of DP transducer, and, pending the application, a single-element pressure transducer might work very well too.

Not enough data as to the process liquid, et al, so a few options always better than just one. Simple single-pole stainless probes may work very well, and will not be to severly affected by algae, provided you are not measuring stagnant water.

In Power Stations, they use a similar method to measure the Drum-Level, and indicate it via 'Hydrastep Probes/Display", where there is a difference in conductivity between water, and steam, and they use a vertical array of green and red LEDS connected visa electronics to these Hydrastep Probes.

Of course, in this instance, the water <2 micro-Siemens conductivity, but it still works well.

Using the vessel as one of the 'points' or 'contac-elements' is probably the better way, if you go down that path, but my selection will always be a level/pressure transmitter.

It depends the end-use, criticality of the application, and budget. All relevant.

And as you have rightly mentioned, the condition of the process fluid.

edit: For the record, I see you have linked ultra-sonic - if you have any lateral/wavey motion of liquid on the surface, it will 'fuck' with your measurement, if I may use your vernacular. In that instance, you will need special filters and other crap, implement delays etc to get a useable reading. Ultra-sonic works pretty well measuring levels of 'solids; as in coal- or grain silos, but have sever drawbacks when measuring liquids.

Yeah, you get them expensive ones, of course....but simple pressure transducer imho the best device for this type of application. Take it through a simple DA converter, or, use the correct drop resistor, and feed it into a LM3914 bargraph driver, if the idea is to have a simple LED-based readout, based on level.

There will be many ideas around, a lot of which will work of course.

ninja-edit: Wrt your signature, no, you're not 'The One(s)'

RobRoySyd · 21st July 2012, 9:14 AM

One pretty safe / fool proof way to measure fluid levels in a tank is ultrasonic "radar".

Link to spec sheet for typical commercial unit: http://www.controlcomponents.com.au/...c_brochure.pdf

Probably costs big time but I doubt a ghetto build of something that uses the same principles would be difficult at all.

The other techique is a sealed pressure transducer that sits at the bottom of the tank. Not as accurate as ultrasonic.

Still the oldest and simplest way of seeing water level in a tank involves nothing more than a float, pulley, weight and a piece of string or rope. Better than a sight glass because if it breaks the tank doesn't drain and it is immune to gung growing in the fluid.

mrbean_phillip · 21st July 2012, 9:23 AM

To the OP: The simplest and arguably cheapest for you: go have a google at this: eTape Liquid Level Sensor

Might be what you're looking for, pretty cheap at around $40 for it, and you will need a small bit of electronics to get it to do exactly what you want. You can take the output signal, drive it through an Op-Amp, and use the earlier mentioned LM3914 Bargraph driver, which you can cascade by the way, to drive your LED's o the outside of the vessel.

Have a good one, and let us know how it goes.

edit: If you really want to rip the pond from under the duck's arse, here you go.

There you can have a good look at the various level-transmitter technologies in use by one if the most reputable companies in this Arena. We normally bundle their products with our Greenfields Projects. Very decent, and you can get anything from pure analog/ binary, HART, Profibus, and various others. But, the link is merely to give you an idea what is out there, and using the old google-ly-goo, you will find many ideas.

[SweN] · 21st July 2012, 11:12 AM

Quote:

Originally Posted by mrbean_phillip

To the OP: The simplest and arguably cheapest for you: go have a google at this: eTape Liquid Level Sensor

Might be what you're looking for, pretty cheap at around $40 for it, and you will need a small bit of electronics to get it to do exactly what you want. You can take the output signal, drive it through an Op-Amp, and use the earlier mentioned LM3914 Bargraph driver, which you can cascade by the way, to drive your LED's o the outside of the vessel.

Have a good one, and let us know how it goes.

edit: If you really want to rip the pond from under the duck's arse, here you go.

There you can have a good look at the various level-transmitter technologies in use by one if the most reputable companies in this Arena. We normally bundle their products with our Greenfields Projects. Very decent, and you can get anything from pure analog/ binary, HART, Profibus, and various others. But, the link is merely to give you an idea what is out there, and using the old google-ly-goo, you will find many ideas.

lol, i was going to suggest a P+F product too, but i doubt it comes in as cheap. They had a discontinued item which did not have to be submerged, would have been ideal i think.

IR camera would also work if you didnt need real time data, only need to check the level daily.

mtma · 21st July 2012, 12:01 PM

E+H must be only one of a few commodity instrument companies that charge extra for the privilege of being able to use their CAD models

so yeah I think they might not be the best ones to pick if $ is far more important than absolute reliability and performance.

But having said that, as long as your measurement height isn't more than about 5 metres then you are probably going to find that a linear sensor (probably ultrasonic, but float and capacitive might be ok if contamination and fluid compatibility is no issue) is going to be more economical than 30 point sensors if you choose to use industrial sensors.

Btw, if you're not after much, one of those reversing sensor with display kits may fit the bill. Testing one a while ago I found that they were pretty accurate. You'll have to mute the buzzer which isn't really a huge thing and the bar graph is probably going to be in reverse, but if the measurement height matches appropriately then you're probably good to go.

paulvk · 21st July 2012, 1:56 PM

I use the old way , a clear plastic tube on the outside of the tank self powered!

RussellK · 21st July 2012, 3:46 PM

I've had good results using a 50KPa pressure sensor connected to the bottom of the tank.

stolly · 21st July 2012, 5:33 PM

Thanks very much to everyone for the replies. To explain the application a little further, we manufacture broadacre boomsprays for use in agriculture - basically a trailer with a tank on it and a boom out each side of the trailer with a series of nozzles along it. The tank level needs to be monitored when filling so that the farmer can get the correct amount of water in. The two ways we do the sight level at the moment is a clear plastic tube with a flange through the tank wall at the top and the bottom (works well, but the big problem is that the chemicals that go into the tanks stain the tube very quickly and you can't see through them) or as mentioned in one of the posts, a tube extrenally with an indicator inside it on a rope / pulley arrangement with the float inside the tank - this works ok, but is prone to problems out in the paddock as it can be a pretty dusty, harsh environment and the pulleys stick / jam etc so the float indicator can jam inside the tube.
-So, what I am now thinking is just using a piece of black pvc pipe that will be mounted out the front of the tank, with a threaded inlet at the bottom with a hose to the sump of the tank, as the tank fills the water will obviously then rise up the inside of the PVC pipe. I would like to screw a series of sensors / switches through the wall of the PVC pipe and as the water covers each sensor they will light up an LED at the front of the trailer that will be mounted on a long piece of flat steel with Lt marks beside each LED.
My next question is does anyone know of a low cost, reliable sensor / switch? - if it works ok I will need 100's of them, it's not a one off project. My other question is, could I make it work using a bolt or any metallic object through the PVC pipe wall? - I'm wondering if the water will be a good enough conductor to act as a 'switch'?
-Also if I can source a sensor, it needs to be 12V DC
Again, thanks for the help and thank to anyone who bothered to read to the end of this post!!

Mickatroid · 21st July 2012, 7:31 PM

I just had a look at the Jaycar kit. My guess is that the transistors mean that a very small amount of current can switch on the LED (electronics and engineering in general is just a hobby for me, I can sometimes be wayyyy of the mark).

To put it another way, if the water as mixed is a poor conductor (and 100% H2O is a poor conductor) the current that it takes to turn on a LED may be such that it does not illuminate because of the drop in voltage though the water (which is equal to the current times the resistance).

One concern I have with this project if not done right is you could end up with the voltage you apply to sensors producing hydrogen and oxygen gases in an explosive mix.

For an altogether different idea, what if you housed the clear plastic tube inside some PVC pipe with some viewing portals or even long slots in it and lit the water in the tube with a bright LED? The LED or LEDs could be outside the tube shining into, up or through it or perhaps set up like fiber optics. Experimentation seems to be required.

Bastard Child · 21st July 2012, 8:04 PM

Never mind hydrogen gas - are the chemicals used flammable? Are they corrosive? Will they eat rubbers, glues, sealant, metal strips?

I'd be pretty careful immersing anything.

stolly · 21st July 2012, 8:18 PM

Nah, no worries with them being flammable. Some of them can attack rubbers, o-rings etc, but provided I stick with stainless steel, poly and any seals will be Viton then she'll be sweet.
Mickatroid, thanks for the idea of the clear pipe inside the PVC, might make a test rig up tomorrow and try.

21st July 2012, 7:33 AM	#1
stolly New Member Join Date: Jul 2012 Posts: 7	Looking for liquid level switch -First post, so not sure if this is in the right section or not, I am trying to source a simple, low cost, reliable liquid level sensor / switch. i want to build a rod / tube / pipe that will house up to 30 sensors that will mount inside a water tank and set it up so that as the water level rises / falls inside the tank, a corresponding LED on the exterior of the tank will light up so that you can easily know the water level inside the tank. - Needs to be 12V DC If anyone has any ideas / advice would be much appreciated!! Thanykou Last edited by stolly; 21st July 2012 at 7:55 AM.

21st July 2012, 8:27 AM	#2
mrbean_phillip Member Join Date: Jun 2002 Location: Doreen (3754) Vic Posts: 2,051	Hello Stolly, Years ago, at a Water Treatment Plant at a Power Station I worked at, we used a similar principle to measure the level in some vessels. What would work probably very well for your application, would be a simple setup of 'conductivity' probes, i.e design a solution that will have the tank as 'earth/negative' and you imagine these 32 probes mounted on a piece of non-conductive material. Stand it vertically inside the tank, and as the water rise, it will close the 'switch' On the other side, you can wire the switches in to a simple Digital to analog converter, there will be many circuits about, and use the analog voltage, derived from the inputs.... Ok, a bit elaborate, but will work well, and be stepped-based, not fully analog. The best would be to source a level transmitter, which could be based on either differential pressure, or a single pressure measurement principle, and use the 0/4-20mA output as the indication, or going into a controller. You can connect a 250 Ohm resistor in series with the circuit, and measure the 0/1-5V drop across if that makes matters easier. If you can tell me what you will do with the signal, I can look at providing you with better feedback, and maybe link you to a few devices I think might work. Last thing you realy want is a mechanical/electrical device, i.e floats inside a tube et al, as they always seem to get stuck somewhere near the top of the tube, as the level drops. Enjoy, and let me know if you need further feedback on this. Kind regards, Phillip __________________ HP Blackbird -> Dual-Quad Xeon ; Macho-Mini -> mini-ITX Xeon Server ; i-Prodigy -> Watercooled mini ITX ; iDuplex -> Prodigy SLI iRacing Button Box

21st July 2012, 8:38 AM	#4
Bastard Child Member Join Date: Jan 2006 Location: Crooklyn 3012 Posts: 2,063	If the water's not treated then algae will fuck with your system, be it a float or conductance pads. Go to siliconchip and search 'water level'. http://www.jaycar.com.au/productView.asp?ID=KC5503 __________________ I'm the reason we can't have nice things

21st July 2012, 9:14 AM	#6
RobRoySyd Member Join Date: Jan 2008 Location: Sydney Posts: 2,984	One pretty safe / fool proof way to measure fluid levels in a tank is ultrasonic "radar". Link to spec sheet for typical commercial unit: http://www.controlcomponents.com.au/...c_brochure.pdf Probably costs big time but I doubt a ghetto build of something that uses the same principles would be difficult at all. The other techique is a sealed pressure transducer that sits at the bottom of the tank. Not as accurate as ultrasonic. Still the oldest and simplest way of seeing water level in a tank involves nothing more than a float, pulley, weight and a piece of string or rope. Better than a sight glass because if it breaks the tank doesn't drain and it is immune to gung growing in the fluid. __________________ "I object to violence because when it appears to do good, the good is only temporary; the evil it does is permanent" Mahatma Gandhi

21st July 2012, 9:23 AM	#7
mrbean_phillip Member Join Date: Jun 2002 Location: Doreen (3754) Vic Posts: 2,051	To the OP: The simplest and arguably cheapest for you: go have a google at this: eTape Liquid Level Sensor Might be what you're looking for, pretty cheap at around $40 for it, and you will need a small bit of electronics to get it to do exactly what you want. You can take the output signal, drive it through an Op-Amp, and use the earlier mentioned LM3914 Bargraph driver, which you can cascade by the way, to drive your LED's o the outside of the vessel. Have a good one, and let us know how it goes. edit: If you really want to rip the pond from under the duck's arse, here you go. There you can have a good look at the various level-transmitter technologies in use by one if the most reputable companies in this Arena. We normally bundle their products with our Greenfields Projects. Very decent, and you can get anything from pure analog/ binary, HART, Profibus, and various others. But, the link is merely to give you an idea what is out there, and using the old google-ly-goo, you will find many ideas. __________________ HP Blackbird -> Dual-Quad Xeon ; Macho-Mini -> mini-ITX Xeon Server ; i-Prodigy -> Watercooled mini ITX ; iDuplex -> Prodigy SLI iRacing Button Box Last edited by mrbean_phillip; 21st July 2012 at 9:41 AM.

21st July 2012, 12:01 PM	#9
mtma Member Join Date: Aug 2009 Posts: 2,593	E+H must be only one of a few commodity instrument companies that charge extra for the privilege of being able to use their CAD models so yeah I think they might not be the best ones to pick if $ is far more important than absolute reliability and performance. But having said that, as long as your measurement height isn't more than about 5 metres then you are probably going to find that a linear sensor (probably ultrasonic, but float and capacitive might be ok if contamination and fluid compatibility is no issue) is going to be more economical than 30 point sensors if you choose to use industrial sensors. Btw, if you're not after much, one of those reversing sensor with display kits may fit the bill. Testing one a while ago I found that they were pretty accurate. You'll have to mute the buzzer which isn't really a huge thing and the bar graph is probably going to be in reverse, but if the measurement height matches appropriately then you're probably good to go. Last edited by mtma; 21st July 2012 at 12:11 PM.

21st July 2012, 1:56 PM	#10
paulvk Member Join Date: Mar 2010 Posts: 691	I use the old way , a clear plastic tube on the outside of the tank self powered!

21st July 2012, 3:46 PM	#11
RussellK Member Join Date: Jun 2003 Posts: 333	I've had good results using a 50KPa pressure sensor connected to the bottom of the tank.

21st July 2012, 5:33 PM	#12
stolly New Member Join Date: Jul 2012 Posts: 7	Thanks very much to everyone for the replies. To explain the application a little further, we manufacture broadacre boomsprays for use in agriculture - basically a trailer with a tank on it and a boom out each side of the trailer with a series of nozzles along it. The tank level needs to be monitored when filling so that the farmer can get the correct amount of water in. The two ways we do the sight level at the moment is a clear plastic tube with a flange through the tank wall at the top and the bottom (works well, but the big problem is that the chemicals that go into the tanks stain the tube very quickly and you can't see through them) or as mentioned in one of the posts, a tube extrenally with an indicator inside it on a rope / pulley arrangement with the float inside the tank - this works ok, but is prone to problems out in the paddock as it can be a pretty dusty, harsh environment and the pulleys stick / jam etc so the float indicator can jam inside the tube. -So, what I am now thinking is just using a piece of black pvc pipe that will be mounted out the front of the tank, with a threaded inlet at the bottom with a hose to the sump of the tank, as the tank fills the water will obviously then rise up the inside of the PVC pipe. I would like to screw a series of sensors / switches through the wall of the PVC pipe and as the water covers each sensor they will light up an LED at the front of the trailer that will be mounted on a long piece of flat steel with Lt marks beside each LED. My next question is does anyone know of a low cost, reliable sensor / switch? - if it works ok I will need 100's of them, it's not a one off project. My other question is, could I make it work using a bolt or any metallic object through the PVC pipe wall? - I'm wondering if the water will be a good enough conductor to act as a 'switch'? -Also if I can source a sensor, it needs to be 12V DC Again, thanks for the help and thank to anyone who bothered to read to the end of this post!! Last edited by stolly; 21st July 2012 at 5:35 PM.

21st July 2012, 7:31 PM	#13
Mickatroid Member Join Date: Jun 2002 Location: Canberra Posts: 882	I just had a look at the Jaycar kit. My guess is that the transistors mean that a very small amount of current can switch on the LED (electronics and engineering in general is just a hobby for me, I can sometimes be wayyyy of the mark). To put it another way, if the water as mixed is a poor conductor (and 100% H2O is a poor conductor) the current that it takes to turn on a LED may be such that it does not illuminate because of the drop in voltage though the water (which is equal to the current times the resistance). One concern I have with this project if not done right is you could end up with the voltage you apply to sensors producing hydrogen and oxygen gases in an explosive mix. For an altogether different idea, what if you housed the clear plastic tube inside some PVC pipe with some viewing portals or even long slots in it and lit the water in the tube with a bright LED? The LED or LEDs could be outside the tube shining into, up or through it or perhaps set up like fiber optics. Experimentation seems to be required.

21st July 2012, 8:04 PM	#14
Bastard Child Member Join Date: Jan 2006 Location: Crooklyn 3012 Posts: 2,063	Never mind hydrogen gas - are the chemicals used flammable? Are they corrosive? Will they eat rubbers, glues, sealant, metal strips? I'd be pretty careful immersing anything. __________________ I'm the reason we can't have nice things

21st July 2012, 8:18 PM	#15
stolly New Member Join Date: Jul 2012 Posts: 7	Nah, no worries with them being flammable. Some of them can attack rubbers, o-rings etc, but provided I stick with stainless steel, poly and any seals will be Viton then she'll be sweet. Mickatroid, thanks for the idea of the clear pipe inside the PVC, might make a test rig up tomorrow and try.