Flow Sensor for Hot Water (50 degrees C)

"For pulse output, one of the dials is replaced by a gear, which turns a magnet that is detected by an encapsulated sensor threaded into the outside of the lens. Pulse rate is determined by the gear and the dial on which the gear is placed. Changing the pulse rate requires no special tools and can be done in the field."

The above water meter claims 1.5% of reading.

Sorry, but it has a 3/4" NPT thread, so you'll need adapters.

The FTB8007HW-PT model, with 20 pulses/gallon, costs $240 from Omega, which might be a show stopper. If you can trace back to the original manufacturer, my guess is the price will be about half.

On the other end of the price scale is this thing for about $10. The specs are rather sparce and the English abysmal. It will do 30 liters/min at 2% accuracy. The picture and the "G1/2" name sorta hints a 1/2" NPT fitting, but I can't be sure.

Pulse rate is: F = 7.5 x Q (L / Min) without the slightest clue as to the value of Q. If Q is somewhere near 1, you'll get 7.5 pulses per liter, which should be sufficient. Buy and try, I guess.

That means pulse frequency is 7.5 x Q where Q is L/min...

I don't believe this one:

I *think* this datasheet:

Oh. That sorta makes sense.

Well, here's a smaller unit that's even cheaper ($2.34). It has a clear case so that you can see what's jamming the mechanism. Works to 80c.

In the description: "Can use for Water heaters, credit card machines, water vending machine, flow measurement device." A water powered credit card machine?

Looks likely.

Yup. I've been down that road and still haven't found an acceptable solution (though I want finer-grained measurement than you and a wide range of flow rates (the low end is just as hard to address as the high end!)

Are you looking to use it for *process* water or *potable* water? Indoor/outdoor? etc. How "clean" is the water?

It would all be easy if it werent for the 5%. Just 2 bent paperclips to count drops.

NT

Potable. Clean water.

I found one solution for $110 each. 7% accuracy but you can get 5% if you calibrate it. NPT threads.

Nearly every flow sensor is BSP threaded and requires adapters. The adapters are not cheap. It was explained to me by one company that these all come out of China and they're all BSP threaded and that adapters are used.

Many of the sensors are only +/- 10% which is not quite good enough. Some put out only one pulse per gallon, using a reed switch to pick up a spinning dial on the totalizing meter. We have no need to totalize.

Rainbird makes some flow sensors for industrial irrigation but they are extremely expensive, over $500 each, and the flow rate is too high.

Wide ranges of flow rates are not really do-able on these paddle wheel sensors, you have to choose a pretty narrow range.

Neptune makes a T-10 model but it's a bit pricey. 1.5% accuracy in modest flow rates (0.2 to 20GpM). For me, the low end of resolution isn't quite low enough (0.1G). But, max water temperature is ~30C (I'd, ideally, like 2 for "cold" water and one for hot)

Ditto. I'm only interested in "rate". I can "accumulate" just as easily as a electronic/mechanical meter can! :>

it's a nutating disc meter, a type of displacement meter.

looks like another nutating disc meter, I would expect one pulse per cycle.

those guys claim "outer diameter 2cm, inner diameter 1cm" (which esentially means 3/8" pipe)

"G1/2" is a type of BSP (do a search on G1/2 pipe thread)

and the outer dimension of G1/2 is close to 20mm which agrees with their claimed dimensions.

yeah, aliexpress has brass G1/2 to 1/2"NPT adaptors for about 5 bucks each, doubling the cost of that device

Have you looked at gemsensors? They meet the temperature have no idea about the % accuracy.

Cheers