Simple Water Level Controller

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
Hi Guys,

I wanted some of your expert opinions. I hope you can help me.
I wish to make a very simple water level controller to detect whether a tank has either water or is empty. The tank itself will be self filling with the water but I do not want the tank to empty.
At the bottom of the tank will be a submersible water pump sucking out the water. If (for some reason) the self filling mechanism is blocked or fails to operate, obviously the water level will eventually decrease and the tank will empty - I do not want this to happen!
However, if this were to happen, I want the pump to switch off as soon as the water level drops to a predetermined height (nearly empty). The water level is detected by the (perhaps copper) probes which will be positioned just above the bottom of the tank. When the probes detect water, the pump operates and sucks out the water. When the probes detects no water (water level has dropped) the pump turns off.
The 2 circuits below ( I hope they are readable) operate from a 12vDC power supply. Circuit 1 is constructed with a relay and Circuit 2 is with a MOSFET. Which in your opinion is the better design....or neither? :)
If neither, perhaps you can suggest a better design or suggest any change in the components?

Many thanks.
 

Attachments

Last edited:

wayneh

Joined Sep 9, 2010
17,496
The standard approach – for "simple" – is a float switch. They're used in toilets, sump pump wells, and many other applications. Relying on conduction through the water is fraught with problems. Can it work? Sure.

I prefer Circuit 2 as long as the pump current draw is well within the specs of the MOSFET. I would replace the op-amp with a comparator (such as LM339, a quad package), raise the value of R9 to 10-100K, eliminate R11, and add a 3.3k pull-up resistor from the gate to +12V. (The comparator output can only go low or "open". You need the pull-up to turn the MOSFET on when the comparator goes high.)
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
PS. I have just had a look at the pump and it's rated as 12vDC 36W 3A.
It maybe tricky to find an inexpensive float switch with these ratings.
 

Alec_t

Joined Sep 17, 2013
14,280
Either switch in the post #3 links should be fine. You could use its contacts to control the on/off state of a MOSFET to switch the pump.
 

HW-nut

Joined May 12, 2016
97
You may want to use a capacitance sensing approach. This would not require electrical contact with the water, does not corrode and can be very precise.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
You can get simple float switches which detect the upper and lower level in one.
Or very simple to make, a small dia tube with sealed end has two proximity switches such as Honeywell ss400 series inserted into the tube one at the bottom level, one at the top.
A circular float with embedded magnet goes around the tube and triggers each prox at the set level.
Max.
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
Either switch in the post #3 links should be fine. You could use its contacts to control the on/off state of a MOSFET to switch the pump.
Yes, good idea. I will think about this, it would probably be a better idea than using probes. Thanks

You may want to use a capacitance sensing approach. This would not require electrical contact with the water, does not corrode and can be very precise.
What is capacitance sensing? I've not heard of this. Can you post a link as to what this is?

who to make FM Transmetter at home
What!?
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
Or very simple to make, a small dia tube with sealed end has two proximity switches such as Honeywell ss400 series inserted into the tube one at the bottom level, one at the top.
A circular float with embedded magnet goes around the tube and triggers each prox at the set level.

Max.
Thanks for the suggestion but I think a float switch operating a MOSFET maybe more suitable.
 

Dodgydave

Joined Jun 22, 2012
11,285
You're going to need to detect two levels, one empty and one just above empty, otherwise the pump will be starting and stopping rapidly, it would be better to use a flipflop using a 555 timer or other chip, to switch the relay on/off.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
You could still do it using reed switches, that is what the purchased ones use anyway.
The reed relay can switch a small relay direct.
max.
 

HW-nut

Joined May 12, 2016
97
Yes, good idea. I will think about this, it would probably be a better idea than using probes. Thanks



What is capacitance sensing? I've not heard of this. Can you post a link as to what this is?



What!?
Same technology used for touchscreens and keypads. Google "Capacitive-Based Liquid Level Sensing Sensor Reference Design" for application note.
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
Thanks everyone for your suggestions.
I think I have decided that the circuit below would be the most cost effective and simplest to build with a cheap float switch, (something like this: http://www.ebay.co.uk/itm/121969125861 ).
The capacitor in the circuit should delay/slow down the starting and stopping of the pump. The gate voltage to the MOSFET should be about 5v and I have included a simple LED to show when the circuit is operating - unless of course someone else has a better suggestion. :)

X-float swtich.JPG
 
Last edited:

Alec_t

Joined Sep 17, 2013
14,280
See DD's suggestion in post #10.
You don't need the potential divider on the gate of the IRF540. It will be happier with Vgs =12V.
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
You're going to need to detect two levels, one empty and one just above empty, otherwise the pump will be starting and stopping rapidly, it would be better to use a flipflop using a 555 timer or other chip, to switch the relay on/off.
See DD's suggestion in post #10.
You don't need the potential divider on the gate of the IRF540. It will be happier with Vgs =12V
.
Many thanks for your suggestions but, I do not think a timer circuit is entirely necessary in my plan. But you may think otherwise.
Let me put you entirely in the picture. See the picture below.
I am currently building a swimming pool and I want a stream, small pool and waterfall to flow into the swimming pool. The tank/reservoir which holds the water is to feed the stream, pool and waterfall from the pump. This tank should/will be constantly supplied with water from the swimming pool. If for any reason the tube which supplies the water to the tank is blocked (it hopefully shouldn't get blocked), then I want the pump to turn off using the float switch. The float switch itself can (in theory) be placed at any depth below the surface of the water.

Your thoughts gratefully received.....
Float Switch1.jpg
 

Alec_t

Joined Sep 17, 2013
14,280
I do not think a timer circuit is entirely necessary in my plan.
That's not what DD is saying. You can use a 555 for various things. It contains two threshold comparators which set and reset a latch, so you could make use of the latch for sensing two water levels if you needed to do that; but having explained your pool plan a single-level sensor looks to be enough.
 

Thread Starter

Ford Prefect

Joined Jun 14, 2010
245
That's not what DD is saying. You can use a 555 for various things. It contains two threshold comparators which set and reset a latch, so you could make use of the latch for sensing two water levels if you needed to do that; but having explained your pool plan a single-level sensor looks to be enough.
Yes thanks, after thinking about it realised what DD was saying (I didn't think of it), but now I think just a basic float switch to sense if the water level drops too far will be ok here.
 
Top