Digital Current meter

Thread Starter

andhy

Joined Nov 17, 2007
1
Hello guys....

I'm an Instrumentation & Control engineering student
My final project is to make a "digital current meter" that can measure the speed of water flow.
I need your help !!
What kind of sensor that i can use ??
What about the water??
How to disconnect the circuit and the water ??
Any opinion??
And the most important is the circuit itself.....
I plan to make this project with a display
Should I do this with microcontroller???
I need Your help!!

Best regards....
(sorry, I'm little speak english)
 

GS3

Joined Sep 21, 2007
408
OK, so we are talking measuring "water current" (flow) here. The first thing that you need is a water flow transducer which will convert water flow to some measurable electrical magnitude (current, voltage, frequency, etc).

The first thing that you need to specify is the characteristics of the transducer like max & min flow, linearity, output signal, price, etc. Do you intend to buy one or to design and build it yourself? Industrial transducers will be expensive while home built can be cheap and simple but the performance will not come anywhere near. There are many types of transducers and just discussing them and designing one would fill pages.

Then you need to design and build the interface and display. This will probably be simpler as you can probably just use a standard voltage meter and feed it the right voltage so that it reads whatever units you want (l/min, GPM, kilobuckets/fortnight, whatever).

here's what you are looking for: http://www.globalw.com/products/MicroFlo.html You can buy it off the shelf. Note that paddlewheel transducers are not the most precise.
 

studiot

Joined Nov 9, 2007
4,998
If you want 'do it yourself' fix a small magnet to the blade(s) of a plastic impeller ( a model shop propellor will do). Mount this inline in a plastic pipe and mount a hall effect sensor on the outside. Then all ya gotta do is count pulses as the impeller speed is proportional to the flow rate. You can calibrate this by passing a measured quantity of water in known time. If you use a transparent pipe you can see the impeller going round.
 

GS3

Joined Sep 21, 2007
408
If you want 'do it yourself' fix a small magnet to the blade(s) of a plastic impeller ( a model shop propellor will do). Mount this inline in a plastic pipe and mount a hall effect sensor on the outside. Then all ya gotta do is count pulses as the impeller speed is proportional to the flow rate.
A turbine type transducer is notoriously inaccurate and non-linear. See http://en.wikipedia.org/wiki/Water_Meter
and http://en.wikipedia.org/wiki/Flow_measurement.

It is difficult to design a meter which has a linear output over a wide range. If all you need is 5% accuracy over a range of 5:1 then you can build something quite easily. If you want 0.5% over 50:1 then you've got your work cut out for you. I worked with machines for the manufacture of carbonated beverages and the flow meters were mass flow meters which were more accurate but plenty complicated.

Come to think about it you could probably start out with an old household water meter and attach an electric sensor to it. That should simplify the mechanical part.
 

beenthere

Joined Apr 20, 2004
15,819
I think that by "current meter" you mean a flowmeter and not an ammeter. As the poster above have suggested, there are several ways to measure water flow. If the sensor you decide to use has a voltage output, you may be able to get an inexpensive digital panel meter that you can arrange to show the flow in engineering units. No computer necessary.
 

mrmeval

Joined Jun 30, 2006
833
If you want to actually measure current then you'll have to convert a proportion of the flow to current. An impeller can do that with a small motor as a generator, water would turn the

Ok, I'm deliberately taking you literally but you can measure water current as electrical current.
;)
 

studiot

Joined Nov 9, 2007
4,998
A turbine type transducer is notoriously inaccurate and non-linear.
An impeller type indicator is perfectly 'accurate' once it is calibrated. It is also quite precise in that rotation speed versus flow is reliably reproducible. However, I don't recall saying it would be linear. I said you could calibrate it with a measuring bucket! Obviosly calibration would involve measuring different flow rates and rotation speeds. A suitably programmed pic could automatically apply the correction. This could be an interesting part of the exercise.
Don't forget this is a learning design project not a factory design project. As such there is plenty of scope for examining the possible effect of impeller shape, size and fit in the pipe, counting circuits, optical versus magnetic transducers, pulse technology and many other variables. If this is a hydraulics project then the difference between laminar and turbulent flow becomes important and could be investigated.

carbonated beverages
Carbonated water is a compressible fluid which behaves very differently from ordinary water for flow measurement purposes.
 

Ron H

Joined Apr 14, 2005
7,063
I believe a Doppler shift measurement on an ultrasonic transducer applied to the outside of the pipe would be very accurate.
 

GS3

Joined Sep 21, 2007
408
Carbonated water is a compressible fluid which behaves very differently from ordinary water for flow measurement purposes.
The machine does not measure "carbonated water" flow anywhere. It measures the flow of treated water and the flow of syrup concentrate to produce the uncarbonated beverage and then again it measures the flow of the uncarbonated beverage in the carbonation process. Nowhere is the flow of a carbonated fluid measured. Mechanical impeller and turbine type flow meters were not precise or accurate enough and mass flow meters were used instead.

As I said, "accuracy" and "precission" are not absolute, yes/no, terms. Rather they are measurable quantities and that is why I asked what accuracy is needed. As long as we do not know the range, accuracy and linearity required we are whistling in the dark.
An impeller type indicator is perfectly 'accurate' once it is calibrated. It is also quite precise in that rotation speed versus flow is reliably reproducible.
The assertion is meaningless if it is not quantified. Can you quantify it and specify the range? The Wikipedia article I mentioned especifically says these types of meters are less accurate than other types. In fact, they are probably among the least precise of all flow meters.
Wikipedia: Turbine meters are less accurate than displacement and jet meters at low flow rates, but the measuring element does not occupy or severely restrict the entire path of flow. The flow direction is generally straight through the meter, allowing for higher flow rates, and less pressure loss than displacement type meters. They are the meter of choice for large commercial users, fire protection, and as master meters for the water distribution system.
In other words, they are used for large flows where other types are impractical but they are not as precise as other types. Do you have cites that contradict this? Because that has always been my understanding too. Can you show me impeller or turbine flow meters which have great accuracy (defined as greater than 1%) over a wide range (say 50:1 or 100:1)?
However, I don't recall saying it would be linear.
I'll remind you. In post #3. Here you go:
If you want 'do it yourself' fix a small magnet to the blade(s) of a plastic impeller ( a model shop propellor will do). Mount this inline in a plastic pipe and mount a hall effect sensor on the outside. Then all ya gotta do is count pulses as the impeller speed is proportional to the flow rate.
(My bolding.) In my book that is the very definition of "linear".

So, I stand by what I said in that we need to know the performance required before we can go any further. If it is meant only as a learning experiment with no regard for any actual precission that is one thing but if it is meant as a real life experiment where accuracy and precission over a given range count then we need to know those figures. And, in any case, he should know the limitations of the different systems or he might end up making the (common) mistake of giving results with many significant figures when the process does not support it.
 
Top