Water flow sensor that isn't cheap but doesn't cost the earth.

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
I have installed a cheapy water flow sensor just after the meter at home. The sensor gives a 5V square wave that cycles with water flow. It's hooked to a wemos which counts the pulses and sends the data via wifi to influxdb/grafana. The sensor is from Banggood and was $9 with postage. It works amazingly well for $9 but it has 3 major problems. First is accuracy, it can be a good 25% out. Second problem is it is not linear so 1 pulse doesn't directly equate to a certain water volume. They supply a formulae but it doesn't work that well and is open to a bunch of timing issues. The third and biggest issue is it has "start flow" which is 1L/minute or 2L/min in different parts of the "documentation". The whole reason I put it in was for leak detection, we have a larger property with lots of taps. At 2L/minute we could have a 3000 litre per day leak and not know.

So I looked for better water sensors and found some that fit the bill but they were $380 and only did one pulse per half litre, so the resolution wasn't that great. So my question is, does anyone know of a water flow sensor that isn't cheap but doesn't cost the earth? Something that is 3/4BSP that I can install at home?
 

Attachments

Last edited:

MisterBill2

Joined Jan 23, 2018
18,167
I am guessing that the sensor uses a magnetic pickup and that the magnetic force is what keeps it from sensing below some flow level. Depending on your mechanical skills it may be possible to increase the gap between the moving part and the pickup, which will reduce the force and allow the detection of lower flow levels. Or, depending on how it is built, you may be able to remove the magnet completely and modify the electronics to have it serve as a variable reluctance pickup, with no magnetic drag at all.The variable reluctance pickup uses the change in air gap to change the inductance of a coil, and that change then varies the amplitude of a high frequency signal, which can then be used to drive a detector. But with no magnetic drag it should be a lot more sensitive.
 

oz93666

Joined Sep 7, 2010
739
You really need a detector that can detect extremely low flows 1cc/min at night with no activity there should not even be that much flow .. I think you have to make your own , many ways , an almost neutrally buoyant plastic ball in a vertical tube with water flowing upwards ....

The easiest way to check for leaks is to turn off supply 3am , turn on after some time if you can hear an inrush of water , then some is escaping .
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
I am guessing that the sensor uses a magnetic pickup and that the magnetic force is what keeps it from sensing below some flow level. Depending on your mechanical skills it may be possible to increase the gap between the moving part and the pickup, which will reduce the force and allow the detection of lower flow levels. Or, depending on how it is built, you may be able to remove the magnet completely and modify the electronics to have it serve as a variable reluctance pickup, with no magnetic drag at all.The variable reluctance pickup uses the change in air gap to change the inductance of a coil, and that change then varies the amplitude of a high frequency signal, which can then be used to drive a detector. But with no magnetic drag it should be a lot more sensitive.
This is the sensor. It has an impeller and hall effect sensor. I could muck around with it but end of day it will never be perfect.
https://au.banggood.com/DN20-G34-Co...-all-1129&ad_id=400330096020&cur_warehouse=CN

The one I want is the second one down on this page but it's too expensive for my budget
http://www.manuelectronics.com.au/water_studies.html
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
You really need a detector that can detect extremely low flows 1cc/min at night with no activity there should not even be that much flow .. I think you have to make your own , many ways , an almost neutrally buoyant plastic ball in a vertical tube with water flowing upwards ....

The easiest way to check for leaks is to turn off supply 3am , turn on after some time if you can hear an inrush of water , then some is escaping .
That would be an option, have something cut supply and then measure the pressure drop. The pipes certainly take several litres due to expansion when I turn the water on and off.
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
You really need a detector that can detect extremely low flows 1cc/min at night
Actually it doesn't have to be that accurate as that would only equate to 1.4 litres per day and I can live with that. I also don't want to be chasing every tiny little drip. If I set a limit at, say, 14 litres per day then that's 10cc/min, still pretty low. It really just needs to be as good as the existing water meter.
 

MisterBill2

Joined Jan 23, 2018
18,167
The existing water meter mention gives me an idea. How about a sensitive contact microphone attached to that meter to detect the vibration from it's operation? Then an amplifier and a means to trigger some sort of alert device if it detects some rate of flow above whatever limit you choose. The clear benefit is no invasion of the piping and so no chance of causing a leak. For experimenting you could just connect the contact mike to an amplifier and listen. Cheap, easy, simple, and safe!
 

pmd34

Joined Feb 22, 2014
527
Hi, do you need to be able to measure an absolute flow, or simply be able to spot a leak? The old "rota-meter" type flow meters that use a clear plastic tapered pipe and float are pretty good and usually come with a magnetic float so you can put in a magnetic sensor and alarm.
 

MisterBill2

Joined Jan 23, 2018
18,167
Revising the plumbing to include such a device is really a fairly big deal. Not at all like connecting things with clip leads. That is why the contact mike to sense vibration seems like a good choice, at least to me.
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
Revising the plumbing to include such a device is really a fairly big deal. Not at all like connecting things with clip leads. That is why the contact mike to sense vibration seems like a good choice, at least to me.
I've already inserted the cheapo BangGood flow meter into the plumbing and I actually need to dig up all the plumbing in the yard and replace it all. In 4 years we've had 3 leaks and there is evidence of at least 7 repairs by the previous owner. I found 7 joiners in 3 metres of straight pipe. They've done the entire yard, probably 200+ metres of poly pipe, in the wrong pipe. They've used rural poly for mains pressure. As I'll be replacing it all, inserting a flow meter isn't a big deal.
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
Hi, do you need to be able to measure an absolute flow, or simply be able to spot a leak? The old "rota-meter" type flow meters that use a clear plastic tapered pipe and float are pretty good and usually come with a magnetic float so you can put in a magnetic sensor and alarm.
Absolute flow isn't essential but while I'm doing it I may as well have that ability.
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
I'm fairly keen just to buy a decent flow sensor and put that in. They are available but just looking for one that isn't too expensive. I think if I found one for $200 I'd be happy. Thanks for the link to the thread, I found this site from there which looks exactly what I need, although possibly still expensive https://www.gemssensors.com/flow/electronic-flow-sensors/rotor-flow

Currently I have everything pretty much done except for the flow sensor. I've installed a solar panel, cabinet, battery and wemos at the meter and this is sending data back to my influxdb and graphs are available via grafana. I just need an accurate sensor. If I can get a cheap enough sensor then I'd get multiples and monitor multiple points, eg garden, dishwasher, shower etc.
 

MisterBill2

Joined Jan 23, 2018
18,167
It may still be that a pickup on the meter can give you an adequate measure of flow.
One other way to detect small flows is to monitor the differential pressure across a spring-loaded check valve in the main line. That check valve will need to have a forward cracking point of possibly 2 to 5 PSI. The concept is that at in a no-flow condition any downstream leakage will create a pressure difference that can be measured. This exact concept has been used in a line off leak detector models and it works very well
 

Thread Starter

MikeKulls

Joined Apr 4, 2016
60
It may still be that a pickup on the meter can give you an adequate measure of flow.
One other way to detect small flows is to monitor the differential pressure across a spring-loaded check valve in the main line. That check valve will need to have a forward cracking point of possibly 2 to 5 PSI. The concept is that at in a no-flow condition any downstream leakage will create a pressure difference that can be measured. This exact concept has been used in a line off leak detector models and it works very well
That sounds like the venturi flow meters where they have a short section of smaller pipe diameter. The idea being there will be a pressure drop with flow at the smaller diameter. It could be a good option if I can find something. Someone suggested cutting the water flow with a solenoid and measuring the rate of pressure drop. That would actually give a very good indication of leaks although wouldn't give me the added bonus of flow rate detection.
 

Reloadron

Joined Jan 15, 2015
7,501
I'm fairly keen just to buy a decent flow sensor and put that in. They are available but just looking for one that isn't too expensive. I think if I found one for $200 I'd be happy. Thanks for the link to the thread, I found this site from there which looks exactly what I need, although possibly still expensive https://www.gemssensors.com/flow/electronic-flow-sensors/rotor-flow

Currently I have everything pretty much done except for the flow sensor. I've installed a solar panel, cabinet, battery and wemos at the meter and this is sending data back to my influxdb and graphs are available via grafana. I just need an accurate sensor. If I can get a cheap enough sensor then I'd get multiples and monitor multiple points, eg garden, dishwasher, shower etc.
I have used the Gem Sensors you linked to in the past, GEM has been around for decades. I have also used similar sensors from Omega Engineering. My experience has been in cases like yours pay attention to the data sheet as most good quality sensors do point out mimimum flow you can expect good results at. Also important is the pipe diameters. When using different types of sensors manufacturers often will say in their data to the effect of 5 pipe diameters upstream and 10 pipe diameters downstream so if I am using 1.0" pipe I want a straight 5" section before my sensor and a straight 10" downstream. I like the units available now with the Plexiglas or acrylic allowing the end user to actually see a paddle wheel so you know if it is moving (rotating) even at very low flow rates.

Ron
 

MisterBill2

Joined Jan 23, 2018
18,167
That sounds like the venturi flow meters where they have a short section of smaller pipe diameter. The idea being there will be a pressure drop with flow at the smaller diameter. It could be a good option if I can find something. Someone suggested cutting the water flow with a solenoid and measuring the rate of pressure drop. That would actually give a very good indication of leaks although wouldn't give me the added bonus of flow rate detection.
The difference from a venturi setup is that if a low pressure high flow check valve is used the greatest sensitivity is at very low flows, which if you are looking to detect small leaks in a big system is what you need. The second advantage is that with the right type of check valve there is very little pressure drop at higher flows.
 

Phil-S

Joined Dec 4, 2015
238
There are all sorts of methods to measure flow depending on whether it's open channel or closed pipe. Ultrasonics are used for open channel and actually rely on the change of height of the water or whatever in a channel when the flow is restricted by a weir or similar.
Closed pipe systems in industry use some fancy Doppler systems, but for small diameter pipes with clean water, the domestic displacement meter is hard to beat. In the UK, the utility company supplies the charging meter, but there's nothing to stop you buying a secondary meter from plumbing trade counters. They aren't as expensive as you think and secondhand ones or refurbished could be available.
The big advantage is that all the hard work, calibration etc. is done for you, they are accurate and very sensitive to low flows such as minor leaks.
My meter already has a small magnet set into one of the dial pointers, plus optical reflectors. I just use a low power Hall sensor, or a reed switch to pick up the pulses. This wakes up an XBee radio set up to sleep, and it transmits every litre. Certainly good enough to detect leaks. Gas meter works the same. Battery life is about a year on a 1/2 AA 3.6-V lithium. Battery life can be monitored by using one of the XBee analog inputs, as well as temperature and even an external moisture sensor
 

MisterBill2

Joined Jan 23, 2018
18,167
There are all sorts of methods to measure flow depending on whether it's open channel or closed pipe. Ultrasonics are used for open channel and actually rely on the change of height of the water or whatever in a channel when the flow is restricted by a weir or similar.
Closed pipe systems in industry use some fancy Doppler systems, but for small diameter pipes with clean water, the domestic displacement meter is hard to beat. In the UK, the utility company supplies the charging meter, but there's nothing to stop you buying a secondary meter from plumbing trade counters. They aren't as expensive as you think and secondhand ones or refurbished could be available.
The big advantage is that all the hard work, calibration etc. is done for you, they are accurate and very sensitive to low flows such as minor leaks.
My meter already has a small magnet set into one of the dial pointers, plus optical reflectors. I just use a low power Hall sensor, or a reed switch to pick up the pulses. This wakes up an XBee radio set up to sleep, and it transmits every litre. Certainly good enough to detect leaks. Gas meter works the same. Battery life is about a year on a 1/2 AA 3.6-V lithium. Battery life can be monitored by using one of the XBee analog inputs, as well as temperature and even an external moisture sensor
In post #1 the TS mentioned "not expensive", and while there are indeed a whole lot of arrangements available they lack that quality. A spring loaded low pressure relief valve plus a differential pressure gage, transducer, or even a switch, will serve the purpose and be both simple and cheaper, and also less complicated.
 

PhilTilson

Joined Nov 29, 2009
131
You mentioned "the existing water meter" in post #8. I don't know what kind of meter this is, but all the ones I have seen have a small wheel in the centre with segments marked on it, which rotates when flow is present. As far as I am aware, this can be VERY sensitive. Would it not be possible to use some form of opto-electronic sensor placed on the face of the meter to detect the movement of these segments - they tend to be very high-contrast. I am sure that, as long as you kept the figures and any dial unobstructed, whoever has to read the meter would not be unduly bothered, and such a sensor could be very small.
 
Top