This is an instructible on how to crudely detect water flow in a pipe and drive an ambient display. I am using a piezo transducer, some LED's and an arduino. The device is a rough prototype of what will eventually become a persuasive technology that motivates sustainable behavior and raises awareness about water use.
This is a project by Stacey Kuznetsov and Eric Paulos at the Living Environments Lab, at Carnegie Mellon University Human Computer Interaction Institute.
Produced by
Stacey Kuznetsov
stace@cs.cmu.edu
http://staceyk.org
Eric Paulos
eric@paulos.net
http://www.paulos.net/
Living Environments Lab
http://www.living-environments.net
The video below illustrates a previous version of this project, where a microphone is used instead of a piezo element to detect water flow. You will achieve better performance when using a piezo transducer, so this instructible details the piezo approach.
Special thanks to Briam Lim, Bryan Pendleton, Chris Harrison and Stuart Anderson for help with ideas and design of this project!
Remove these ads by
Signing UpStep 1Gather Materials
- Breadboard
- Microcontroller (I used an Arduino)
- Mastic
- Piezo Transducer (http://www.radioshack.com/product/index.jsp?productId=2062402)
- A few LED's (I used 2 yellow, 2 red, 2 green)
- Candle holder or similar-sized container
- Wire
- 1 Mohm (or other large value) resistor
- 4.7K Resistors (3)
- 1K Resistors (1)
- Low-value Resistors (for the LED's)
- Clipping Wires
- Jumper Wires
- Mastic
- op amp (LM613)
| « Previous Step | Download PDFView All Steps | Next Step » |











































I've been thinking for some time now to create something like that but I didn't know the way, I didn't think about piezo. I will try to make something that measures how much water passes the tap and displays that on a 2 digit display. I think to actually see the number of litres that are being wasted is more shocking. But I still haven't figured how to measure the quantity of water passing through...
Anyway,
great implementation! Well done!
Dimitris
After this you have data relating the vibration signal level to flow rate (volume/time). If it is linear relationship, you can interpolate/extrapolate a function to create the readout. If not linear, but is still structured and not random, you can quantize the vibration level depending on how accurate you want the flowrate readout.
could be an interesting way around the usual methods of flowrate sensing, which aren't typically very suitable for small-scale unobtrusive projects.
let us know if you do anything like this, it would be interesting to see the results, I myself would use something like that.
Regardless, awesome instructable!
-Dane
http://electronics.union.rpi.edu/
http://www.refresheverything.com/
Given that the labels of v+ and v- are reversed in that picture than these are the values:
R1 = 1M
R2 = 4.7k
R3 = 4.7k
R4 = 1k
R5 = 4.7k
Q1 = piezo transducer
Not to get nitpicky on you but you might change the schematic so there's a dot between R1, R4 and Q1. The way it is now it looks like the wires cross over each other but to not connect.