ESP8266 Community Forum

- Wed Sep 16, 2015 8:40 am #29153 Someone posted this one on Hackaday, but it is using magnetometer sensor :

https://hackaday.io/project/4648-analog ... onsumption

- Fri Sep 18, 2015 9:35 am #29352 According to the theory document you linked to, the ultrasonic range finder you listed won't do the trick, at least not without modification. The range finder just gives you the round-trip time of an ultrasonic ping. To figure out the flow rate, you need the frequency of the returned pulse. I also imagine that there is a narrow range of frequencies that will work through the copper pipe and with moving water.

There is almost definitely an off-the-shelf part that would work, not that I have any idea where to find it. I'll bet it isn't cheap (ie. > $10). Certainly, I'm attracted to the convenience of just clamping a sensor onto a pipe. The alternative is to do some plumbing (which isn't all that hard) and install a cheap flow meter. I've got a couple of these sitting right here on my desk.

To me, the interesting part of the problem is the software that recognizes the consumption pattern of various fixtures and pulls them out of the aggregate data. I wonder what field of knowledge that is covered by. Wavelet theory?

[...]

Can this ultrasonic range finder for the Arduino do the trick? $1.80
http://www.aliexpress.com/item/HC-SR04- ... 4,201409_1

[...]

Theory on how they work.
http://www.engineeringtoolbox.com/ultra ... d_495.html

So when can I read about this amazing ESP hack on hack-a-day?

- Fri Sep 18, 2015 8:46 pm #29401

According to the theory document you linked to, the ultrasonic range finder you listed won't do the trick, at least not without modification. The range finder just gives you the round-trip time of an ultrasonic ping. To figure out the flow rate, you need the frequency of the returned pulse. I also imagine that there is a narrow range of frequencies that will work through the copper pipe and with moving water.

There is almost definitely an off-the-shelf part that would work, not that I have any idea where to find it. I'll bet it isn't cheap (ie. > $10). Certainly, I'm attracted to the convenience of just clamping a sensor onto a pipe. The alternative is to do some plumbing (which isn't all that hard) and install a cheap flow meter. I've got a couple of these sitting right here on my desk.

To me, the interesting part of the problem is the software that recognizes the consumption pattern of various fixtures and pulls them out of the aggregate data. I wonder what field of knowledge that is covered by. Wavelet theory?

Did you see my post? The app notes say 1mhz is good for water. You can use time of flight method. If you can get a free sample from maxim or ti you can probably do this project for under $20. But it would require testing and calibration per pipe size and material.

- Mon Sep 21, 2015 12:22 pm #29595 I saw it, but didn't click through to anything. That has been remedied. All those different approaches are really interesting. To succeed in the consumer market, if it requires plumbing, it will have limited success. That limits the options.

The magnetometer on the meter idea is actually a pretty elegant solution. At least your numbers would match the utility company's.

[...]
Did you see my post? The app notes say 1mhz is good for water. You can use time of flight method. If you can get a free sample from maxim or ti you can probably do this project for under $20. But it would require testing and calibration per pipe size and material.