Water Level Measurement Sensor

[Stuart]

Hi,
Heres yet another approach...

http://www.electronic-circuits-diagr...armsckt6.shtml

and this one...

http://www.siliconchip.com.au/cms/A_30607/article.html

Stuart

[YoYo]

Good find Stuart. This is a great alternative to the more expensive stuff and doesn't require any type of actual water sensor, plus no moving parts.

I'm going to post the actual text of the link along with circuit diagram below so we don't have to worry about it going bye bye...

From Stuart's link above:

This circuit not only indicates the amount of water present in the overhead tank but also gives an alarm when the tank is full.
The circuit uses the widely available CD4066, bilateral switch CMOS IC to indicate the water level through LEDs.
When the water is empty the wires in the tank are open circuited and the 180K resistors pulls the switch low hence opening the switch and LEDs are OFF. As the water starts filling up, first the wire in the tank connected to S1 and the + supply are shorted by water. This closes the switch S1 and turns the LED1 ON. As the water continues to fill the tank, the LEDs2 , 3 and 4 light up gradually.
The no. of levels of indication can be increased to 8 if 2 CD4066 ICs are used in a similar fashion.

When the water is full, the base of the transistor BC148 is pulled high by the water and this saturates the transistor, turning the buzzer ON. The SPST switch has to be opened to turn the buzzer OFF.
Remember to turn the switch ON while pumping water otherwise the buzzer will not sound!

[dmohr]

I see in your profile that you use SECUs in your HA setup. I have an idea for getting a level from a voltage range input on the SECU.

Have two wires going down a pipe. One a solid bare conductor and the other broken every inch or two by resistors.

What will happen is that when the water is at the bottom of the pipe, the 5V will go through all resistors and register close to 0 volts to the SECU and as the water level rises, the water shorts out the lowest resistor creating a higher voltage to the SECU. As the water gets to the top, the water will be bypassing all the resistors and will transfer 5V to the SECU.

No moving parts, thin design and as long as the resistance of the water stays constant which I assume it would, your readings should remain accurate.

Worst part will be calculating the resistance of the water and choosing the number of resistors, than calculating the value for each resistor.

Doug

[YoYo]

Hi Doug,

That's a great suggestion. This idea is actually very similar in concept to what Stuart's links provide. Instead of using LEDs you could connect them to an analog input on the Secu-16 and let the voltage level indicate the water level as well.

The nice thing is you'd only be using one input to get the job done!

[dmohr]

Hi Doug,

That's a great suggestion. This idea is actually very similar in concept to what Stuart's links provide. Instead of using LEDs you could connect them to an analog input on the Secu-16 and let the voltage level indicate the water level as well.

The nice thing is you'd only be using one input to get the job done!

Something like this:

[Jim Doolittle]

Due to resolution, I think you may only be able to do up to five resistors. I use a series of three resistors connected to Secu input for garage door position and was thinking of adding more and seem to recall a discussion about the practical limit on the appdigsupport forum.

[dmohr]

Due to resolution, I think you may only be able to do up to five resistors. I use a series of three resistors connected to Secu input for garage door position and was thinking of adding more and seem to recall a discussion about the practical limit on the appdigsupport forum.

I thought there were 50 resolution points on the analog input setting? Of course that is from my memory only, I haven't picked up the manual in years (if I even have one)

One thing I left out of the picture is that you would have to put a resistor on the two bottom contacts or else when the water goes down below the sensor, it will trigger an "open" 253 parameter, the same as a 5V signal would.

Am I right or is my memory failing me already?

[Michael McSharry]

I believe when you add up tolerances and environmental variations such as conductivity of the water your resolution will be nowhere near the resolution of the A/D in the SECU16.

[Jim Doolittle]

Michael is correct about environmental conditions and tolerance of the resistor itself and an added source that doesn't apply in my garage door application: water condition. You may want to spend a little more on better resistors to at least reduce that source of variability.

[YoYo]

Yet another inexpensive option... Remember the BASIC stamp? They've got an ultrasonic sensor kit for it for about $25 that I think might also work for this. Measures distances from 1" to 10'. Could also use it to measure the salt level in the softener, etc.

It's called the Ping))) Ultrasonic Sensor. Looks like it's backordered for now though, and I'm not sure what would need to be done to make it work long-term in a high-humidity environment.

And you know, I think I've got a BASIC stamp laying around here somewhere...

[zoomkat]

In commercial applications a "bubbler" level setup is used. If you can get a cheap analog pressure transducer you probably could make one. In the bubbler setup, a pressure transducer/transmitter is connected to the top of a piece of tubing and the tubing is then immersed into the liquid, which increases the pressure in the tube as the liquid tries to move up the tube. The bubbling part is a small constant amount of air is added to the tube such that it bubbles out of the bottom of the tube. This ensures the level instrument isn't affected by air in the tube disolving into the liquid and lowering the pressure over time. I think the pressure transducers are cheap now, as they are what is used in the $5.88 digital tire pressure gizmos at Wal-Mart.

[zoomkat]

If you are interested in sonar, the below gizmo might be of interest. I think it could actually be directly connected to the computer serial port to get the range data.

http://www.wrighthobbies.net/catalog...products_id=71

[YoYo]

Wow Zoomkat, that's a great find. If it actually works, the price is amazing at $25 and nothing else is needed to make it work, except a 5v supply and only uses something like 3ma.

What's really interesting is that it supports serial, analog voltage output (10mv/inch), and PWM, apparently all at the same time.

I think I'm going to give one of these a try...

Btw, the Hexamite folks got back to me about that unit, but the web price is wrong and the real price is USD $285, so I put that one out of the running.

Thanks,

[smoothtlk]

I assume sonar has no problem bouncing back off of water. If so, put a flat floating plastic board in the tank.

But since we have fun with jury rigging, how bout a 10 turn potentiometer with a pulley on the shaft. Then, run some fishing line one wrap around the pulley. One free end goes to a float in the water, the other free end with a weight to keep the line taught with friction on the pulley.

when the float sinks, the pot will turn. The pot can be measured by PC connected voltage measuring device.

[viclick]

Yo Yo,

The April 2006 issue of Circuit Cellar magazine has a very interesting way of measuring water depth. The author uses the system to monitor his sump pump tank level. His implementation is a rather complex device, but he uses a simple method of measuring water level. Basically, he attaches a gauge pressure sensor to one end of a copper pipe. The open end is submerged into his sump pump tank. As the water level rises it compresses the air in the copper pipe, which is registered by the pressure sensor. The sensor outputs an analog voltage that can used for anything you need. The output voltage is low (25 mV full scale) so amplifiers are needed to make the voltage uesable by everyday electronics. It is an interesting and robust way of measuring the level in an open tank.

The ultrasonic system will be a great way to do it. I've used the PING module and have had great luck with it. I also explored the device Zoomcat pointed out but it didn't have the resolution I needed for my application. It too looks like a very nice device. If you use ultrasonics and need accuracy, don't forget to compensate for temperature changes at the transducers. For example, if the surface of the water is measured at 72" at 72 degrees, it will measure 75" at 32 degrees because of the change in the speed of sound of air due to temperature changes. Three inches may not matter but if it does you'll have to correct for temperature.

Good luck with your project!

Vic