Waterpump cutoff (for broken pipe)

[rustycarr]

Thanks for the ideas that were put forward.
A sump float switch in the water tank is good and will turn off the pump before it runs out of water.
Just outlining my problem. Rainwater tanks and 240vac Pressure pumps. The pump turns on and off as you flush a toilet or turn on a tap. It senses the pressure drop and starts the pump. When the tap is turned off the pump keeps going until its up to pressure, normally a few seconds. When a pipe breaks or and end comes off or some jerk (me) cuts it with a lawn mower without noticing it the pump will then run continously until all the water in the tank is empty and after that the pump seizes up.
So what I need is something to sense that the pump is running abnormally long, say half hour continuously, and then shut it of.
So to wait until the tank is empty to shut it off is no good , because here in Australia water is precious.
Also ease of connection. I was thinking of something plug into the power lead. Sensing the voltage drop when the pump kicks in so start a timing sequence of say a 555 timer and relay. Or an audio amplifier trigger that is "listening" and a timer circuit.

I have talked to a pump irrigation expert and he hasnt heard of any system but he said it happens all the time..
Anyway thanks for the ideas, keep them coming and a solution might be found together.

 

[KeepItSimpleStupid]

You can sense current as well Also see SSAC. Just makes it more expensive. You can do it cheaper if you were building 1000. There is development costs in a microcontroller solution usinng EEprom, F
lash or battery backed up RAM. I/O modules or brew your own is also possible. Could also do it with a Smart Relay (A development system is required).

You can develop and test or wire 4 or 5 commercial parts together. PICs or a CPUStick based system is also possible.

 

[ronsimpson]

I now see the picture. I also live with too little water. The float saves my pump every day.

The first time the pump turns on it may take 5 minutes to fill the pressure tank and thus the pressure switch turns off the pump. There after it takes only 3 minutes of pumping to refill the tank. This happens every 7 minutes.

When the pipe breaks the pump runs forever, or maybe 10 minutes on 1 minute off.

I got from the used electronics store some time delay relays. (delay to off, and delay to on) Mine have a knob in top to set the timer. The relay/timer is powered from the pump electricity. Any time the pump has power the timer has power. Because the pump never runs more than 5 minutes and typically 3 minutes the timer gets reset by lack of power before the 7 minutes is up. It the pressure tank does not get filled in 7 minutes the timer opens/closes contacts and kills the power until a human resets it.

I did not draw the circuit because I don’t know if you can get a delay to on or delay to off timer and that inverts the logic.
You can get 110 or 220 volt relays that are easy to interface to the pump. I have some 24 volt timers that need a transformer to drop the pump voltage.

Garage door openers have a crude timer in them. Mine is fixed at 5 minutes. Might work.

 

[Reloadron]

Actually in thinking along the lines of KISS there may be a somewhat easy (well sort of easy) way to go about this.

OK, since the trigger for this will be along the pump run time and not using a float switch we can look at it that way. This could be done using a PIC I think with minimal programming skills or discreet components in a more complicated fashion.

Starting with the pump motor maybe detect the pump on by either current sense or use the signal that turns the pump on. Either way get a signal and condition it. That signal enables a count up in the PIC as long as it is present. Logic high on some input pin? If that logic level goes away (pump off) before the PIC times out all fine and well, reset the PIC count. However, if the on time exceeds a pre determined count and the PIC is allowed to time out then an output on the PIC goes high (or low) and stops the pump and gives an alarm condition. All this is doing is looking at pump run-time as a defined period. Exceed the period and stop the pump, remain under the period and reset the counter as long as the pump stops.

Not to say that a fail safe couldn't be tossed in with a float switch as a final off for the system.

Now if the problem is common and people burn up pumps running them dry you may have a nitch if you can build a cheap system to market.

Ron

 

[4pyros]

Here is an idea. Use a smart power strip to detect the pump turn on with out moding the pump at all. Use a power pack to power a 1/2 hour timer that will open a relay wired before the smart power strip and pump and it will stay off untill reset. what is the currant of the 220 Volt motor? Andy

 

[KeepItSimpleStupid]

Exactly. One of the gotcha's is if there is a power failure and the pump has overrun. My component solution handles this.

Aside:

I once had an interesting problem. Leave the valve in the same position it was before power went out. The turbomolecular pump and valve was generator backed up, so the 1 minute off before the generator would start would drop out the air operated valve. Closing briefly was fine.

There were a few other systems which had generator backed up cryopumps and if the power was off for 5 minutes or more, you should leave the pump off. If less than 5 minutes, it would be beneficial to leave it on. I never worked on that problem. If power was out for more than 5 min, it would take about 8 hours to defrost and restart the pump.

In yet other systems which had near liquid nitrogen refrigeration systems, the control system would have been a lot tougher with heaters and valves to control. The, not so good system, just dropped out the diffusion pump system and closed a few valves. Idealy, the amount of time power was out or temperature should have determined what happened during recovery. Heaters, valves and refrigeration system would need to be controlled. We did not have the generator capacity to backup the three 5 HP refrigeration systems.

Useful stuff, but management made the ultimate decision.