Water Level using Capacitnace

[odsk]

Hi,
I am working on a sensor to measure water level in a metallic container and for this is I am using capacitance measurement. I used an oscillator 555 and I am measuring the frequency compared to the level.
The issue that I am facing right now, is the change to the frequency that the sensor give and I am suspecting:
- Drift of the oscillator vs temperature (the 555 or accompanying resistors).
- changes in water dielectric due to temperature.

Anyone worked on such sensor? what would be the best route to solve this issue.
Thanks,
ODSK

[th_sak]

I thing it's time to try a different approach. See my suggestion here. I have made many water tank level meters with this design.

[odsk]

Thanks,
My original design contained 24PCDFA6D and 24PCBFA6D wet/wet sensors and they worked OK for some time before the water Lime build up damaged the sensors.
I agree with you that the pressure sensors are the easiest one to use but due to the buildups I had to change to another type of measurement  where I have only a wire immersed in the water.

ODSK

[solutions]

Hi,
I am working on a sensor to measure water level in a metallic container and for this is I am using capacitance measurement. I used an oscillator 555 and I am measuring the frequency compared to the level.
The issue that I am facing right now, is the change to the frequency that the sensor give and I am suspecting:
- Drift of the oscillator vs temperature (the 555 or accompanying resistors).
- changes in water dielectric due to temperature.

Anyone worked on such sensor? what would be the best route to solve this issue.
Thanks,
ODSK

A 555 is one of the worst frequency sources to use for long term frequency stability. You are relying on absolutes, which are not.

That said, if you switch in a known capacitance to do a calibrate-before-measure, you should be able to compensate some of that "drift" in the 555 circuit.

Same principle applies to your dielectric shift with temperature - an immersed (at the bottom of the tank) electrode pair should allow you to determine the change. Ideally this would be that reference cap you are calibrating to. Yo could even conceptually heterodyne the ref cap oscillator to the measurement cap oscillator - use a 556 (dual) which should match better than two discretes.

[th_sak]

...they worked OK for some time before the water Lime build up damaged the sensors...

I had the same problems with water and humidity until I put some drops of oil inside the tube of the sensor. Anyway it's your project, you decide.

[chandra2sekhar2000]

hai,i used a Ripple counter IC to generate the frequency,using capacitance variation.in that design also temperature is a very big problem ,i am struggling. i have some idea ODSK,use a Lm35 TO measure the temperature and then compansate the capacitace change in the Raw count.i did not had time to work on it.if it works ,pls let me know

[kukumar]

Ad  (via rele or similar...) capacitor (fixed value) to capacitor probe - for calibrating...
Its simple...

[odsk]

I had the same problems with water and humidity until I put some drops of oil inside the tube of the sensor. Anyway it's your project, you decide.

What is the rate of success when using the drop of oil? Also which sensors you used? My other concern is the price of the sendsor package + op. Amplifhier used in conjuntion?

[Unhappy]

What is the rate of success when using the drop of oil? Also which sensors you used? My other concern is the price of the sendsor package + op. Amplifhier used in conjuntion?

Please go through April , May & June issues of EPE. Also Siliconchip Nov., Dec.,2007 and Jan2008 issues
24PC series sensor in India can be had for INR950=USD19/20 (1USD=50INR)

[sharning]

If you want to work with any type of liquid, including dirty ones, I suggest using an ultrasonic sensor ( same as used in car parking bumper sensors ) this would need no contact with the water. I think many robot kit suppliers as well as auto accessories would have the parts you need. ( of course the result would need to be subtracted from the height of the tank! ) Mount the sensor inside the roof of the tank facing the liquid.

[odsk]

If you want to work with any type of liquid, including dirty ones, I suggest using an ultrasonic sensor ( same as used in car parking bumper sensors ) this would need no contact with the water. I think many robot kit suppliers as well as auto accessories would have the parts you need. ( of course the result would need to be subtracted from the height of the tank! ) Mount the sensor inside the roof of the tank facing the liquid.

How good the sensors sustainability in humid environment? as it is going to be mounted inside the tank so I assume a lot of humidity inside.

[Elmer]

How good the sensors sustainability in humid environment? as it is going to be mounted inside the tank so I assume a lot of humidity inside.

Ultrasound is a very common method for level metering, so acoustic transducers designed especially for this purpose exists.
http://en.wikipedia.org/wiki/Level_sensor#Ultrasonic

[dotm]

there was an article on elektor 4/09 describing exactly such a sensor , with mathematical background, schematics and bascom code.
i hope you know german

[titi]

Hi,

an other article about Capacitive Liquid Level Sensor

http://www.best-microcontroller-projects.com/capacitive-liquid-level-sensor.html

Best regards.

[Mworks]

Another way I have done liquid level sensors is to bounce an IR light source, led, IR laser,  off the surface then measure the amount reflected off the surface. Works great and is non contact , if you are working with things like oils. The further away the surface of the liquid the less light returned.

[betosmax500]

Cypress Solution here...

http://www.youtube.com/watch?v=0JL85O1BM8g

[pushycat]

One of my friend in India had developed a Capacitance based waterlevel controller which uses PIC microcontrollers. The master will have 28 Pin PIC micro (Number erased ) connected to a 18 pin PIC (Number erased) which drives 4 nos of seven segment Display and two wires from the master board same will carry power as well will communicate with the slave unit which is installed at the over head tank will read the level in the tank and will communicate with the master unit too. Level will be displayed on the led linearly in 0 to 100% steps. One needs to calibrate the unit everytiime it is used in new tank.

[Langley]

Looks like the R.M. Young Model 50202 capacitance sensing rain gauge uses a 556 (dual 555) as the oscillator and gate for a PWM that is then converted to 0-5VDC out for 0-50mm of precipitation via Op Amp  low-pass filter. Very simple circuit for a high precision $1200 device spec’d to operate from -20C to +50C.

Clearly, it's possible to use 555 family timers in such applications with good performance and stability.

See schematic on page 7 of manual:
http://www.fondriest.com/pdf/rm_young_50202_manual.pdf

http://www.youngusa.com/products/3/17.html

[mare69]

ACAM has excellent chip for measuring capacitance. I have good experience with picocap. Power consu mption is low, so you can run from small battery.

[solutions]

If you've never heard of a company, the first place you look to see how they are doing is http://www.acam.de/company/jobs/