Author Topic: Contact-Less Flowmeter Idea?  (Read 885 times)


  • Newbie
  • *
  • Posts: 21
Contact-Less Flowmeter Idea?
« on: May 02, 2018, 11:57:33 am »
The Problem:
I've always wanted to know the actual gallons per hour of my aquarium, without having to break the tubing or doing a bucket measurement. (My canister filter is "rated" for 383 US Gallons (1,449.81 Liters) Per Hour)

So I've been staring at the pile of Digispark Pros in my garage. The key idea here is to use a U shaped tube to submit a signal, and then measure the delay with another transducer.

Doppler Ultrasonic Flow Meters - These are fantastic for thicker fluids that reflect the ultrasound between two transducers. Several types exist, typically it's a V shape, or W shaped pattern where the liquid medium bounces differently based on the speed of the flow.
Transit Time Ultrasonic Flow Meters - These are fantastic for thinner fluids where your measuring the time displacement as the flow increases, the delay increases or decreases. (I'm more interested in this approach.)
Vortex Flow Meters - These seem to be designed with a impediment inside the tubing that causes swirls on the liquid, the swirls are measured, to me this looks too complicated.. ?

I found this page from omega engineering that discusses the formula in detail:

I found another page from texas instruments:

Essentially they jury rig a ~50 pico second trigger called "TDC1000-Q1" to figure out the time between a transmitting transducer vs. a receiving transducer. I've also read that water will transmit frequencies differently based on temperature, maybe this is loosely involved with NIST measurements?;

Some things I'm not clear on is all the math required to determine the ideal transducer set, the ideal frequency, the ideal length, how accurate the transit time needs to be, the effective flow measurement range, creating a table of time delay variables based on the temperature vs. transit time. Any bubbles in the flow will degrade the flow measurement, I know the TI demo device uses oversampling to try to remedy inaccuracies, and slower flow rates are going to be more difficult to measure.

I have seen a few different air flow measurement devices written in PIC, but I'm much more interested in arduino-based code. There doesn't seem to be any DIY explanations of how to do this. Any suggestions on what might explain this better? Typically I find ultrasounic flow meters that costs thousands of dollars for this concept :P