Cap Sensor

[bigkim100]

I have a small tank (about the size of a pop can) that is made out of plastic, and I need to monitor the water level inside. I CANNOT USE PROBES..no matter what, so does anyone have a circuit that can be used that uses capacitance in such a small container. I only need to know when the level gets low, not terribly accuratly, BUT the circuit does need to ignore short bursts of the tank "sloshing" the water around. .Again, the circuit only needs to trigger when there is about 1/2 to 1 inch of water in the container, after being full.
Is the "sloshing" sensor part of this something that only a microcontroller can do, or can a analogue way be designed.
Thanks a million

 

[dknguyen]

You can ignore sloshing by filtering the sensor's output signal. YOu can do this with an analog filter using discrete components (or ICs designed for the job) or you can use a MCU to digitally filter the signal using calculations and math. This will make it so that you can ignore certain frequency components in your signal (in this case you would make it ignore "high" frequencies because sloshing happens at a higher frequency than the level of the liquid, which would change at a lower frequency). I don't know if you know about Fourier Analysis and frequency spectrum...if you do this shouldn't be a problem for you to understand.

Why do you want to use capacitance? Why pick the method you want to sense something with when you don't even know how to sense it yet? That's working completely backwards. Capacitance requires charge moving around or something changing distance with respect to something else (like a vibrating membrane or two moving plates). Capacitance pretty much requires that you use probes of some kind, plus measuring capacitance requires some circuitry that is a bit unconventional. Without probes....I don't think you can use capacitance. I don't even know how you could use it even if you could use probes. We need more details about the setup.

I have 3 methods I can think of without sticking anything into the can:
-stick the tank on a scale and weigh it continuously. The can is small and this is the most straightforward way.
-shine a laser through the tank (must be transparent) and measure the light on the other side to figure out the height. may require lasers at different levels
-use a downward facing distance sensor to measure the height of the fluid in the tank. this requires an open top and something with a narrow beamwidth (most sonars will echo off the inside walls of the tank giving false readings because you just want to measure the fluid level).

I would go with a scale. We need more details to be more helpful. Again, why capacitance?

 

[bigkim100]

More info from original poster

As per usual...the devil is in the details. The sensor needs to be small, lightweight, battery powered, and most importantly, the tank must operate on its own (the tank will indicate that it is empty by illuminating a led on the exterior of the tank. The liquid may not always be water, and may not always be clear, so optics will not work.
The only reason that I went with capacitance is because of one thing that I didnt mention...it has to be kept small, and self-contained. I cant think of a way to weigh it, and keep it small, and self-contained. I tried a test with a pre-made, self contained capacitance proximity sensor, (THE ONES THAT LOOK LIKE A HUGE PLASTIC BOLT) on the bottom of the tank, and it worked....except for the sloshing. I cannot use a pre-made unit, as all the ones that I have found either require voltages that are so high, that batteries would be prohibitive, or the sensors are physically too big. I would also just prefer to make something from scratch, to try to learn something.
I found a sensor at:

https://www.electro-tech-online.com/custompdfs/2006/06/capgage.pdf

and was wondering if it would work by wrapping 2 pieces of foil around the exterior of the tank, leaving 2 insulating small gaps down either side. The 2 pieces of foil would then become my capacitance plates. Its really small, SIMPLE, and can be run off of a 9volt battery.
So now we must address proper filtering if Y'all think that a capacitance sensor will work. Unfortunatly Im rather old-school, and know nothig of stamps, pics, and microcontrollers, so is there a way to do this a good old-fashoned analogue way?
I REALLY appreciate the time that all of you have taken to supply feedback, and help that you have given me so far.

 

[BeeBop]

Hi,
I'm not sure how small of a container these work on, but take a look at this:

**broken link removed**

I kind of think this would work for you.
Regards,
Robert

 

[philba]

As per usual...the devil is in the details. The sensor needs to be small, lightweight, battery powered, and most importantly, the tank must operate on its own (the tank will indicate that it is empty by illuminating a led on the exterior of the tank. The liquid may not always be water, and may not always be clear, so optics will not work.
The only reason that I went with capacitance is because of one thing that I didnt mention...it has to be kept small, and self-contained. I cant think of a way to weigh it, and keep it small, and self-contained. I tried a test with a pre-made, self contained capacitance proximity sensor, (THE ONES THAT LOOK LIKE A HUGE PLASTIC BOLT) on the bottom of the tank, and it worked....except for the sloshing. I cannot use a pre-made unit, as all the ones that I have found either require voltages that are so high, that batteries would be prohibitive, or the sensors are physically too big. I would also just prefer to make something from scratch, to try to learn something.
I found a sensor at:

https://www.electro-tech-online.com/custompdfs/2006/06/capgage-1.pdf

and was wondering if it would work by wrapping 2 pieces of foil around the exterior of the tank, leaving 2 insulating small gaps down either side. The 2 pieces of foil would then become my capacitance plates. Its really small, SIMPLE, and can be run off of a 9volt battery.
So now we must address proper filtering if Y'all think that a capacitance sensor will work. Unfortunatly Im rather old-school, and know nothig of stamps, pics, and microcontrollers, so is there a way to do this a good old-fashoned analogue way?
I REALLY appreciate the time that all of you have taken to supply feedback, and help that you have given me so far.

That sounds a bit like what I suggested. Your problem will be the size of tank. Try it. use a dmm with cap setting. test empty and then full. if the difference is fairly noticable then you shold be ok.

the dave johnson circuit is over complicated. Use a 555 in asynch mode and the sensor as the capacitor. pulse period will vary with the value of the capacitive sensor. easy to measure with a uC. You could use a uC directly by measuring the RC time constant. It would be better if you could construct a sensor that fits inside the tank, though.

 

[dknguyen]

Isn't that just why optics might work? THe less clear the liquid is the better...as soon as you lose the laser beam (if the liquid is not clear) or it gets weaker or distorted (if the liquid is clear) you know the liquid level is high enough to be sitting between the laser.

Don't capacitve sensors depend on the dieelectric properties of the liquid? For the same fluid level, you will get different capacitance readings for different types of liquids because they have better/worse dielectric properties. Unless you can calibrate for it you may not be able to determine the true fluid level.

I still think the distance in this case is too large for capacitance sensors unless you use a high voltage, unless the capacitance changes involved are much larger than I imagine (or we can measure capacitance much more accurately and cheaply and I imagine).