industrial flow meter

[ertansuluagac]

looking for a flow meter to use in an industrial environment. Is there a product that is inexpensive and you know?
_________________
Es

[Ttelmah]

You need to give a lot more data. Rates of flow. Liquids involved.
Interface required. Pipe connection involved.
It's really a bit 'off topic', but I'm sure some of the posters here may be
able to recommend something.

[ertansuluagac]

Yeah, I know, and I'm asking for the opinion of the people here. As a result, someone has measured the flow. I want to have the event max 6lt \ mn.
_________________
Es

[temtronic]

re: 'flow meter'
OK, WHAT kind of 'flow' do you need to measure ? air, water, helium,liquid concrete ?? NEED to know material as well as expected flow rates(min,max, mps, GPH ) ! Also what kind of communications interface ? RS232, RS485, IR, US, wireless( RF), DW( direct wire). Comm format.. bitbanged, ASCII, ?

re: 'industrial environment'
indoors, ourdoors, Alaska or Mexico, summer, winter,HAZMAT area or a 'clean lab' ?

not trying to be hard on you BUT you need to supply specific requirements.
I've got inexpensive remote control devices in hazardous locations communicating 15 miles on solid copper wire so information lke 'fault detection', 'loss of signa' is kinda important to have.

Jay

[Ttelmah]

Basic turbine meters are quite easy to interface to a PIC (pulse signal, count
with a CCP). PCT Technologies do the most accurate examples of these
that I know of:
<https://www.pctflow.com/our-products/flow-meters/turbines/dm-series-turbine-flowmeter/?gclid=Cj0KCQjwoKzsBRC5ARIsAITcwXH3DQ7yCb7cTK3pUKU6l26RJ_pJVPrMc6AR-wCSGR7NB3g9e7CSGUcaAorOEALw_wcB>

Piston type units generally are more accurate for lower flow rates.
I've done basic liquid detection using the PIC's capacitive sensing
abilities, combined with flow using a simple inductive coupled rotor, for
relatively low accuracy.

[ertansuluagac]

I already found 2 products. One is proteus brand and the other is yf-s401. I calculated the frequency of yf-s401 by pulse. It worked perfectly. But because the product was cheap, it broke down after a while. Proteus is a very good brand but the room is very expensive. I want a middle class product.
_________________
Es

[newguy]

What fluid are you attempting to measure? If water, I've had good experiences with the seeedstudio sensors.

[ertansuluagac]

I measure water.
_________________
Es

[Ttelmah]

RS components sell worldwide, and their 'own brand' unit has worked
quite well for me in some systems:

<https://uk.rs-online.com/web/p/flow-sensors-indicators/0257149/>

Just look up the part number in your local RS webpage.

All sensors of this sort will have issues if there is much dirt in the
water.

[temtronic]

hmm... 'water' ...
nice and clean or like Mr. T says 'dirty' ?
also temperture ? cool, warm or boiling hot ??

I see Banggood has a lot of 'water flow sensors' and I suspect they're all versions of 'paddlewheel and hall effect'.

Providing the water is clean, low pressure and cool I don't see HOW the sensors can fail.

Jay

[Ttelmah]

Common failures:

Limescale.
Algae.
Even small amounts of grit will kill the bearings.
Chemical attack on the bearings.

Clean water 'no problem', but it is amazing just how much muck is in water
in general, and (in fact) if you have very clean water (DI water for example),
then this will also attack bearings and even the casings. Send DI water
through a normal plastic pipe and you will be amazed at how much the
purity has dropped at the end of the pipe. You have to use materials like
PTFE to carry this.....

[ertansuluagac]

The data sheet of the RS model that you throw at Ttelmah says rise / fall 60 / 6ns. Can I read it so fast? I've done reading at 400ns before with ds33ep256mc502 processor.
model
RS PRO Radial Flow Turbine Flow Meter, 0.25 L/min → 6.5 L/min

https://uk.rs-online.com/web/p/flow-sensors-indicators/0257149/
_________________
Es

[Ttelmah]

The rise and fall is not the reading time.
It's just how fast the signal changes between levels.
It gives 4800pulses/litre. All you need to measure is either the time between
pulses, and then the reciprocal of this gives the frequency (allowing you
to have a measurement for every pulse), or simply count pulses over a
period.

At 5l/min, will give 400 pulses/second.

So if you measured the time between pulses (t) as 2mSec, then:

f =1/t = 1/0.002 = 500Hz

Flow in litres/min = (f*60)/4800 = 6.25l/min

Neat thing then is if you cancel out the maths, you get:

Flow = 0.0125/t

So measure 2mSec, and flow is 0.0125/0.002 = 6.25l/min. Measure 10mSec
and flow is 0.0125/0.01 = 1.25l/min

So maths is really easy.

Even better multiply everything by 1 million, and read the time between
pulses in uSec, rather than calculating 'seconds', and you can use:

Flow (ul/min) = 12500000000/t(usec)

Which can be done in int32 arithmetic. So with the 2000uSec measurement
you get:

flow = 12500000000/2000 = 6250000ul/min