PDA

View Full Version : Advanced Water Level Indicator



MGear
05-24-2010, 02:26 PM
Hello everybody !

Do you have 'overhead tank' at your home , and have to check it every morning to see how full it is ? Well, I do. Being sick of this daily routine I'll be upgrading my analog water level indicator( WLI) into a digital one which will show me the water level 2 floors below in my room.

Now the point is, how is the transducer module up at the tank supposed to communicate with the display module down below?

Really need your suggestions and views over this ...

Thanks

Ulhas
05-24-2010, 04:07 PM
simple problem is expressed in very complicated manner

Just ask what you need form your post we need to scratch our head for what excatly you need.

OK, don't try to build it your self there are industry standard sensors with display are available (4-20mA) . go for it
and interlock with your uC for automation

pratheek
05-24-2010, 05:33 PM
A wired communication between your sensor and your display screen will be cheap and easy to implement.

Go for a wireless connection if you do not want wires running around your house. But you will have to devise a good power supply which runs off 240V mains for the sensor module(else you will still have to go up to the tank everyday to change the batteries of the sensor module!) .

lolman_returns
05-24-2010, 05:48 PM
Can you tell the model and build of your pump so that I can look it up to get an idea.

MGear
05-24-2010, 06:42 PM
@LOLman: I wont be using any Pumps... actually have'nt thought of automating it. Right now I just want to see the water level in the Tank.

@Ulhas: I dont want buy any devices from the market for this. Cuz WLI on its own will turn out to be a nice project on its own.

@Pratheek : Yes Wired com. was what I first thought, but considering 4 or more water level marks and the length of the wires required plus the attenuation the signal will face I am thinking of wireless one. But here too am not sure about its ability to transmit properly through so many walls.
And yes I want to keep the system as cheap as possible. I have known 433Mhz TxRx pair http://onlinetps.com/shop/index.php?main_page=product_info&cPath=126_127&products_id=454. But am not sure if they'll work here.

So what do you all say?

pratheek
05-24-2010, 08:57 PM
2 floors roughly equals 20 feet, add to that another 5 feet of the tank and you get total height of 25 feet. The commonly available ASK RF Tx Rx should be able to transmit for distances that are more than that, when the transmitter is powered at 12V and when proper antennas are used.

MGear
05-24-2010, 09:13 PM
All right,
But how am I supposed to power the Tx at 12v ? Lead Acid battery ? Solar Panel ( I have one with 6V 300mA rating ) ? Or by stepping up voltage from a low voltage source ?

pratheek
05-30-2010, 07:10 PM
Most ASK 433Mhz transmitters work between a wide range of input voltage of 3 - 12 V.
So you can use them even at 6V, but you will be getting a less range.

vikas
05-30-2010, 07:16 PM
Best is to keep it wired, Not high tech but fault proof and no batteries required.

Just put two wires with power and do power line communication over them for reading sensor data and displaying it below.

MGear
05-30-2010, 11:32 PM
I have worked on with wireless comm. but power-line comm. is new to me. Can you find me a tutorial which can explain how to implement it?
And how much voltage would I need to send signals over 30ft twisted pair wire?

MohitM
05-31-2010, 04:40 PM
Its basically injecting a high frequency carrier wave on the power lines in your home (the live and neutral of your electric mains) and then modulating it. Its complicated and more importantly, very risky.

MGear
05-31-2010, 08:34 PM
Thanks Mohit but if you've read earlier posts then you'd know I might not work on Home AC supply.

Aint there any technique to send signals over DC power supply ?

pratheek
06-02-2010, 06:09 PM
Aint there any technique to send signals over DC power supply ?

You can go for serial (USART) communication. You will need two wires for that though. One signal wire and one Gnd.

docel
06-04-2010, 02:42 PM
Hmmmm... looks like a Water level Detector Robot...

There are 2 practical ways of doin this : RF & Voice.
Everyones covered the RF part, so let me tell you about the Voice part.

The Tx part is a 555 timer Ic ( CMOS is better suited for its stability) as a multivibrator. The' R' is an LDR in a pipe. at the top end and an LED at the bottom. Water in the pipe will attenuate the light falling on the LDR and change the OP of the 555 at pin 3. If the RC is calculated for audio frequencies, the 555 will sing different notes at different water levels. Connect pin 3 to the cheapest 433mhz RF TX.
The RX can be connected to a Micro ( if you are a coding buff) or to a set of 567 tone decoder ICs.
The 567 IC ( Read the datasheet ) will respond to specific frequencies by a tunable Preset resistance. You can either get a digital o/p for specific levels with any number of 567s in parallel or get a continuous audio tone from the RX, depending on the level of water.
An Audio Freq to Voltage converter circuit can give you a LED bargraph visual output. There are many Simple discrete circuits to hi-funda DSP algos for this. The choice is yours.

ps: it can be a simple 2wires from the 555 to the 567 too. dont worry about cable power drop- it is good to a cool 1/2Kms, if you use a 12v at both ends.

MGear
06-04-2010, 04:19 PM
@Docel: Wow that was a twisted approach I never thought of ! :)
I appreciate your complex thinking but do you really think its necessary to use optical transducers ? Or use voice band frequency for analog comm. ?
I might be prompting for wrong reason, but still, enlighten me.

docel
06-05-2010, 01:21 AM
Thank you 2MGear , for your wonderful observations!

Well, I thought a little advanced technology may very well be utilized here.

1. All communication of ALL types is through Voice band.
2. Nothing wrong in using Optical transducers for the job- it guarantees Isolation- Keeps the water pure.
3. Nothing wrong in Voice band frequency. Infact, Voice band is the best as it automatically filters out the noise band.
4. The system doesn't bother whether it is Digital or Analog comm. All it bothers about is the best way of conveying the message from point A to point B.
5. The entire project is energy efficient and error free.

...I'm serious !

MGear
06-05-2010, 10:37 AM
Thanks Docel!
I got your point and will be working on this project the way you suggested ... This might take some time as i'll need to implement it my way!
Will be posting pics as soon as the work finishes! :)

docel
06-05-2010, 10:50 AM
:D ...never thought you'd swallow that !!!
I Just tried to make a very simple, practical project as complex as possible. The thing can be made in about 3hrs flat.

To further convince you: I have made 1000s of this very same thing as a commercial product, and it has never failed. It was used both for non-contact liquid level measurement as well as to test attenuation of 5Km long RF/AF Co-ax cables.

Let me know if there are any hiccups- will help you out.

Blacksteel
07-13-2010, 11:49 AM
If it's wireless that you're into, then have a look at this -> http://letsmakerobots.com/node/21596

I was equally frustrated with the tank overflow problem and made this. It uses a RF 434 MHz wireless link bought from an online shop in India. I'll be trying out the relay interfaces for the motor and let you know if it's worth in fully automating it.

If there's any other query pls feel free to ask.

vigyanabikshu
07-16-2010, 02:18 PM
I dont know powerline comm
but

To sense water level you can use many methods
Contact
1. Using a simple wire and a transistor
3. Using float sensor
4. Pressure sensor
Contactless
4. Ultrasonic Sensor

For communication you have below options
Wired
RS485 using a twisted pair wire
RS232/UART
Wireless
RF Module
XBee Module

If you are going wired, I suggest you RS485

Blacksteel
07-16-2010, 02:26 PM
Hmmmm... looks like a Water level Detector Robot...

There are 2 practical ways of doin this : RF & Voice.
Everyones covered the RF part, so let me tell you about the Voice part.

The Tx part is a 555 timer Ic ( CMOS is better suited for its stability) as a multivibrator. The' R' is an LDR in a pipe. at the top end and an LED at the bottom. Water in the pipe will attenuate the light falling on the LDR and change the OP of the 555 at pin 3. If the RC is calculated for audio frequencies, the 555 will sing different notes at different water levels. Connect pin 3 to the cheapest 433mhz RF TX.
The RX can be connected to a Micro ( if you are a coding buff) or to a set of 567 tone decoder ICs.
The 567 IC ( Read the datasheet ) will respond to specific frequencies by a tunable Preset resistance. You can either get a digital o/p for specific levels with any number of 567s in parallel or get a continuous audio tone from the RX, depending on the level of water.
An Audio Freq to Voltage converter circuit can give you a LED bargraph visual output. There are many Simple discrete circuits to hi-funda DSP algos for this. The choice is yours.

ps: it can be a simple 2wires from the 555 to the 567 too. dont worry about cable power drop- it is good to a cool 1/2Kms, if you use a 12v at both ends.

Man, thats one crazy idea!.. I'm gonna try that out on my existing waterlevel controller system. Dunno if my "cheap" RF modules with work at that baud, but if it works I'll post it on my LMR page!

vikas
07-16-2010, 04:43 PM
Wired we do not have problem of batteries , RF batteries i required.

What is the current usage of your system ? and which batteries are you using , maybe make it solar powered :) .

vikas
09-06-2011, 09:13 AM
Looks like i need to build it, Rather than tone decoder will use frequency counter and LCD and have actual water levels. The 3/4 point contact sensors are simply too unreliable and fail often, while giving no real indication of water level at any given point.

docel
09-06-2011, 09:48 AM
:confused: Vikas.....!
You mean you're REALLY gonna make it ?
Nice to hear that - was wondering how come no one took the bait for so many months ;)
Good luck!!

MGear
09-06-2011, 10:19 AM
I would really like to see your version Vikas.
What Docel had suggested didn't turn out to be my cup of tea. So I built the system my 'easy' way. Which turned out to be a costly affair. The recent reply made me look for my files in the backup again, so am attaching the schematics.
This design was very much to imitate the '1-wire' protocol. But was not efficient by any standards.
Also I wanted to run the Tx on parasite power mode, which I couldn't. Both the boards have separate power supply.

What do you think of it, Vikas and Docel ?

vikas
09-06-2011, 10:34 AM
Yes , our new house has 3 watertanks , i dont feel like running around and getting shouted at other than wasting water.

I thinking of the sensor like this -

LDR at top of PVC with 555 , Ultra Bright LED at bottom of PVC pipe , so when water fills up it will change the frequency. Though im thinking sediment might settle on the led even if i raise it a little and put some meshing in holes where water enters pvc.

Second I was also thinking of putting a buoy with a mirror on top inside the PVC , this way sediment will not settle and it will always be clean. In this case LDR and LED will be at top with a separator in b/w.

Is the frequency response linear ?

Which would be better ?

For uC I was thinking of 89v51rd2 , but it does not have any eeprom and IAP seems wagely documented and not sure if its worth the effort .

Second in choice is PIC16F1938 as i already have lots of tools for pics including ICD and pickits etc

Atmega's seems expensive ... even for 8 kb variant.

Arms ... seem overpowered :) .

vikas
09-06-2011, 10:35 AM
For power to sensor , i would use cheap 3 wire setup , 2 for power and 1 for frequency out from 555. Want to keep low cost and maybe make a open source water level meter :) .

Display will have a 16x2 lcd with 4 tank setup and control using 4 tact buttons.

docel
09-06-2011, 11:29 AM
Yes... I have done that too. Used one of my ADC Atmega8 LFR boards as sensor/ processor
The mirror on top gets fogged too.
Now I use a Acrylic Rod/tube with the LDR and LED embedded in a drilled hole. No Water fogging although regular wiping is needed once a month or so. Need some funny juggling of the code to get a linear output....

vikas
09-06-2011, 12:02 PM
Hmm , if things require maintainance ... im not upto the task :) , need something more reliable

Something similar

http://howmuchsnow.com/waterlevel/ , but using Sharp ranging sensors , similar can be done by us ?

Inplace of a mirror will a ping pong ball work ? So the selaed tube is not affected by water fogging ?

docel
09-06-2011, 12:41 PM
Anything needs some maintenance Vikas. I have used the sharp sensor with great results but even that gets fogged and gets spoilt. IF you need any accuracy, you have to clean it almost everyday too. Also, The plastic lens somehow reacts with water and becoms useless in time.
The simplest is what i'm using. IR LFR sensor/processor with ADC. This is a reflective, non contact method, You will have to keep the water condensation at a minimum. One way is to put a small DVD drive Fan which will evaporate the fogging moisture.
Mind you - All these were part of my school projects with kids and we never bothered about accuracy at all. just as a comparative indicator with a fair amount of accuracy say +/- 3" .

vikas
09-08-2011, 07:54 AM
So true, yep maintainance is required. Any other method , which provide more long term and lower maintenance approach ?

I also found some pressure sensors but available ones had limit of 10kpa , for now im thinking of putting a float switch to get started and a quick fix.