Water flow detection but not for leaks
- Thread starter Handywill
- Start date
rockinarmadillo said:
I would suggest just using a 1/2", 3/4" or 1" Hall effect sensor (whatever size matches the current plumbing) rather than the 1/4" version just to keep things as simple as possible. There are other sizes listed further down on the web page linked to above.
kwilcox
Active Member
Very interested in this thread since I have a similar need to detect flow in 3/4 copper pipe for my bathroom vent fan that runs only when the shower is on. Today I use PIR in the shower. How does a hall effect sensor work? Current when flow detected? I'd have to convert to a contact closure but I imagine a relay would work.
How would I do acoustic sensing?
How would I do acoustic sensing?
pct88
Member
My control solution for hot water boosters has been much simpler- time of day. Turn on the circulator pump before the typical morning/evening shower and bathroom activities so the water is already hot when needed. Turn it off over night and when the house status is unoccupied. Add override controls as discussed above for that off-schedule usage.
kwilcox said:
The hall effect sensor has an impeller in the pipe chamber and the sensor detects that it is spinning as water flows. You get pulses out, making it easy to detect the flow. The sensors are quite accurate, making it a good device to use if you need to measure the actual water consumed.
Here's a simple example of how to use one.
http://www.seeedstudio.com/wiki/G3/4_Water_Flow_sensor
wkearney99
Senior Member
What about sensing the temperature instead? Presumably the water pipe's going to get warmer when the shower is used. That and I'd venture sensing temperature is probably easier and considerably less expensive.
kwilcox
Active Member
hmmm... that's another good idea. The only issue would be off triggering since the pipes would retain residual heat for a period of time. This isn't necessarily a bad thing since its important to run the fan for a while after the shower stops to keep the bathroom mirrors defogged. I could probably attach a probe to the pipe and graph temperature change over time to see if its feasible.
wkearney99
Senior Member
If you're going to use a controller of some sort then some kind of timer in the programming might be useful to help determine needs. I'd think you'd have to calibrate based on temp rise time and perhaps factor ambient temp. As in, pipes in the summer/winter might have different rise/fall times, perhaps also by time of day.
Whatever you do make SURE the method you use to control power to the fan has an override at the switch. You do not want to set up some embedded gizmo that applies power directly to a device without having a manually operable interrupt in a place where it would be OBVIOUS to anyone that might be servicing it. You don't want someone coming along later thinking just turning off the wall switch would be enough...
Likewise, be sure to factor power loss/interruptions. Resetting clocks on a bunch of stuff can be tedious if/when power to them cycle off/on. Having just had to deal with that for the steam shower, the floor heat, the towel warmer, the microwave, coffee maker and a handful of clocks...
This is where the 'internet of things' that gets bandied about stands to get annoying tedious... Don't get me wrong, I do like the idea of monitoring/controlling devices, especially if/when standards (which won't be) are developed.
The same thing could be said for the temperature detection logic. Presumably you'd have a baseline temp and a rise time. If your device came on and the temp was suddenly above the baseline, what should it do? There your logic would need to know the max temp and factor the fall time. Those would be values that'd have to be kept in some sort of non-volatile storage. They could be part of the programming code but I'd think it would be more effective to have them adjustable values. For a PC-like environment with an easy programming interface (aka, network connected) that'd be relatively easy. But for something that had to be 'manually' loaded via a programming cable, not so much...
I'm reminded of the 'boiled frog' tale. Though while untrue, this is one of those areas where programming can fail to achieve desired results when ineffective logic is applied. It's all about parameters and their times. How cold was it, for how long, how quickly did it heat up, how long did it stay that way, how quickly did it cool, etc. So it's not just about 'is the pipe warm or not'...
Whatever you do make SURE the method you use to control power to the fan has an override at the switch. You do not want to set up some embedded gizmo that applies power directly to a device without having a manually operable interrupt in a place where it would be OBVIOUS to anyone that might be servicing it. You don't want someone coming along later thinking just turning off the wall switch would be enough...
Likewise, be sure to factor power loss/interruptions. Resetting clocks on a bunch of stuff can be tedious if/when power to them cycle off/on. Having just had to deal with that for the steam shower, the floor heat, the towel warmer, the microwave, coffee maker and a handful of clocks...
This is where the 'internet of things' that gets bandied about stands to get annoying tedious... Don't get me wrong, I do like the idea of monitoring/controlling devices, especially if/when standards (which won't be) are developed.
The same thing could be said for the temperature detection logic. Presumably you'd have a baseline temp and a rise time. If your device came on and the temp was suddenly above the baseline, what should it do? There your logic would need to know the max temp and factor the fall time. Those would be values that'd have to be kept in some sort of non-volatile storage. They could be part of the programming code but I'd think it would be more effective to have them adjustable values. For a PC-like environment with an easy programming interface (aka, network connected) that'd be relatively easy. But for something that had to be 'manually' loaded via a programming cable, not so much...
I'm reminded of the 'boiled frog' tale. Though while untrue, this is one of those areas where programming can fail to achieve desired results when ineffective logic is applied. It's all about parameters and their times. How cold was it, for how long, how quickly did it heat up, how long did it stay that way, how quickly did it cool, etc. So it's not just about 'is the pipe warm or not'...
wkearney99
Senior Member
kwilcox said:
The downside being putting something in the flow results in a maintenance item, along with potential problems with building code limits depending on the locality. Using a flow meter doesn't escape the need to use sensible programming logic. I'd think it would still make sense to do something smarter than just a dumb on/off, as intermittent use (say, for cleaning) wouldn't require the fan. I don't supposed it'd hurt, just waste power. Likewise you'd want to avoid nonsense like a child turning things on/off quickly and potentially wrecking the fan motor (to say nothing of the switch). So a bit of 'we have a detectable situation, and it's gone on long enough to warrant using the fan' would make some sense.
pete_c
Guru
Here much of the world of water stuff is on the 1-wire network.
A while back installed (copper pipes here) two water meters and these two are connected to one 1-wire dual counter. The meters sat in a box for some 2 years before I decided to install them. I have also added analog water pressure guages which I would like to change over to digital.
That said temperatures are measured here and there with 1-wire sensors.
I did a while ago install a "Davy" pump which appears to work via flow more than water pressure. There is a sort of sensing flow plate on it; its all mechanical mostly.
It is though not a recirculating pump. It appears to be built to last forever and I have done maintenance on it (once in 10 years now) getting talked through said maintenance with the folks in Australia (nice bunch there and great customer service). While this is in place; I use it for monitoring more than automation though.
I have over the years though learned a little bit about copper sweating of pipes. I did recently upgrade the main ingress water meter here going to a newer "plastic" meter with plastic compression fittings from an older brass meter; plastic cuz of some new EPA requirements and such.
A while back installed (copper pipes here) two water meters and these two are connected to one 1-wire dual counter. The meters sat in a box for some 2 years before I decided to install them. I have also added analog water pressure guages which I would like to change over to digital.
That said temperatures are measured here and there with 1-wire sensors.
I did a while ago install a "Davy" pump which appears to work via flow more than water pressure. There is a sort of sensing flow plate on it; its all mechanical mostly.
It is though not a recirculating pump. It appears to be built to last forever and I have done maintenance on it (once in 10 years now) getting talked through said maintenance with the folks in Australia (nice bunch there and great customer service). While this is in place; I use it for monitoring more than automation though.
I have over the years though learned a little bit about copper sweating of pipes. I did recently upgrade the main ingress water meter here going to a newer "plastic" meter with plastic compression fittings from an older brass meter; plastic cuz of some new EPA requirements and such.
dpilati said:
3/4" pipe contains 0.023 gallons per foot of pipe. So 150 feet holds about 3.5 gallons. It doesn't sound like 30 seconds is unreasonable to move that amount of water around the loop, depending on the size of the circulator pump. 7 minutes seems very long, but I'll take your word that's what you see.
Similar threads
- Question
