toilet tank water level sensor/timer/alarm

Thread Starter

AnalogKid

Joined Aug 1, 2013
9,395
For $6 on ebay you can get a battery powered sensor/alarm that beeps if the water level in a toilet tank gets too high, indicating a clog or stuck valve or something. Great; not what I need. Last night a toilet ran all night because of a tangled chain, so I need a too-low system, not a too-high system.

My thought is about a sensor/timer/beeper that tests if the water level in the tank is too *low* for too long a time. This would indicate if the flush valve were stuck open (tangled chain, etc.) because the water level stayed below x inches for over 1 minute, or something like that. Small plastic case stuck on the back or side of the tank, two wires up under the lid and wire-tied to the fill stem at an appropriate sensing height. The question is about optimizing the design so a standard 9 V battery lasts years. Hopefully it actually beeps very rarely; I'm looking at the static current through the sensor conditioning. Something very low power detects no water between the sensor tips and releases the reset input of a CD4060. I've got everything after that.

This project is in the schematic puttering stage. My first thought is a 2N7000 MOSFET with 10 M ohms from the gate to the 9 V, and the sensor tips between the gate and GND. The FET is in series with the rest of the circuit GND net and the battery negative terminal. Yeah, it is DC, and I've seen enough water sensing threads to know what y'all think of that. But what about a couple of heavy gold plated connector pins for the tips?

ak
 

crutschow

Joined Mar 14, 2008
27,940
A low-frequency CMOS astable multivibrator to power the sensor tips through a capacitor would take little power and minimize the problem of tip erosion.
 

#12

Joined Nov 30, 2010
18,223
If you want to replace the probes every 5 years, that seems reasonable to me. We're not making a retail product here.
Even if we were, the marketing department would demand that it fail every 5 years.:D

Considerations: The bulk of the metal is more important than the coating. ANY metal you energize with DC will un-electroplate one of the probes. Better to waste something cheap, like zinc.
You should use a p-mos for the power driver because every voltage except the noise maker ground will float to, "high" most of the time that the noise should be "off". You might ground the noise maker to the N-mos transistor if its current is low enough.
You can mess around with the spacing between the probes or you can adjust the gate resistor...depending on your water conductivity.

I think you made good choices for active components. The CD will operate on a wide range of voltages and leak 1 ua/V.
Something like this?
 

Attachments

LesJones

Joined Jan 8, 2017
3,520
Two suggestions.
One To reduce the electrode erosion problem you could just apply power to the sensor for say 100 mS every 10 seconds. (This would also reduce the average current.
Two You could us e similar system that used in washing machines to sense the fill level. They have a tube dipping into the water with the end just above the empty level. The top of the tube is connected to a sensitive pressure switch. (On a washing machine this is a diaphragm about 3" diameter which presses on a micro switch.) When the water level gets above the bottom of the tube it tries to push it's wau up the tube increasing the pressure. thus opperating the micro switch. (There are also float type level switches on ebay that may do the job.

Les.
 

Alec_t

Joined Sep 17, 2013
12,253
Getting to the root of the problem, you could thread lengths of drinking straw or other plastic tube over the chain to reduce the chance of it tangling. Worked for my flapper valves.
 

AlbertHall

Joined Jun 4, 2014
11,530
If the cistern has the usual float which shuts off the water when full then you can glue a magnet to the float and use a reed switch to detect when the the float is in the 'up' position. If you use a changeover reed and use the normally closed contact, when all is normal and the cistern is full that switch will be open and will close when the water level is low. This switch supplies power to a timer ('555?) which will operate a buzzer after a time delay. This arrangement would use no power at all while the cistern is full.

I have built a system which refills the cistern from a rainwater butt and I have used a magnet and reed switch to know when to refill the cistern and it is easy to arrange and works reliably.
 

djsfantasi

Joined Apr 11, 2010
7,867
Or a simple float connected to a microswitch. Use the kind with a long lever, drill a hole in the end and attach a length of brass rod to a small rubber ball.

Wire to a small microprocessor and code a small routine to time when the water level is low.
 

Thread Starter

AnalogKid

Joined Aug 1, 2013
9,395
Les - I thought about pulsing the sensor (post #4), but it costs more power to run the oscillator continuously that it saves. And, the washing machine system is great if you have 110 Vac available.

Albert - duh. Missed that one.

#12 - almost exactly what my sch looks like so far, but no P-fets allowed.

Thanks all.

ak
 

shortbus

Joined Sep 30, 2009
8,879
Electrical circuit to solve a common mechanical problem:) Many people just don't get the chain adjustment correct, that is what causes this problem.
 

Thread Starter

AnalogKid

Joined Aug 1, 2013
9,395
How does that apply here?
He's talking about the chain between the flush handle and the flapper. If it is too long it can get tangled when the flapper is up during flushing, then hold the flapper open when the water has run out, preventing the tank from filling. His point is that if you prevent the problem at the source (prevent tangling) you don't need an alarm.

However, I have more electronic parts than plumbing parts.

ak
 

tcmtech

Joined Nov 4, 2013
2,867
He's talking about the chain between the flush handle and the flapper. If it is too long it can get tangled when the flapper is up during flushing, then hold the flapper open when the water has run out, preventing the tank from filling. His point is that if you prevent the problem at the source (prevent tangling) you don't need an alarm.

However, I have more electronic parts than plumbing parts.

ak
A trip to the hardware store and a $7 toilet flush flapper rebuild kit followed by about 5 - 10 minutes of basic plumbing work would solve the problem too.

9 volt battery plus 555 timer plus switch plus misc other parts along with assembly and tweaking them will eat up way more time and money than the rebuild kit will.
 

crutschow

Joined Mar 14, 2008
27,940
Okay, I wasn't thinking about the chain between the handle and the flapper obviously. :oops:
I've also had that problem with chains so now I use flappers that have a plastic strip instead of a chain (example below).
Never had one of those tangle.
upload_2017-5-1_10-37-18.png
 

AlbertHall

Joined Jun 4, 2014
11,530
It seems US toilets are rather different to UK ones. Mine has two buttons on the top - one gives a full flush and one gives a reduced flush. The buttons press down directly on the water release mechanism with no 'string' needed.
 
Top