Pump failure detect

[djpark]

I am required to develop a PIC based controller to monitor the status of submersible pump status and switch over to secondary unit in case of pump failure.

The pumps are typically 200W - 800W running at 240V and I am not allowed (actually not possible) to attach a sensor to the pump. So I am considering the current flow of the AC power source to the pump.

1) Is it actually workable to detect the pump (rotary motor I presume) failure by measuring the current flow? I have been thinking that if the motor gets jammed, the current would increase.

2) What is the simple (and cheap, of course safe as well) way to monitor the AC current flow?

Any advices would be appreciated.

-- dj

[ajt]

I did this on a recent project but with a 12V pump. I monitored pressure in the line as well as current the pump motor was drawing.

Monitoring the pressure is fairly easy as there are multiple pressure transducers available including surface mout varieties. Honeywell makes a good line of these and other sensors.
_________________
Al Testani

[libor]

I am developing exactly the same thing using a PIC, I am quite ready with the circuit, some programming still required (the main functions are working OK: failure if jammed, failure if running dry, failure if running to long, etc. I am working on some extra fancy functions, like power outage alarm, remote error reporting, journal logging, total wattage metering, optional LCD, etc.)
I used a Hall-effect current sensor from Allegro (ACS752SCA-050) to measure the momentary current the motor is drawing from the AC line without galvanic connection to it. This part is not very cheap (about 5-6 USD) but I found it to be the safest and reliable method to measure current in an isolated way.
I suppose (hope) we are not in the same country (i.e. not competing on the same market), I am feeling inclined to advise you further if you find it useful.

[dyeatman]

The CR Magnetics model 3110 is a split core type current sensor I have used quite a bit. It works very well for me in House power sensing applications.

http://www.crmagnetics.com/newprod/splitco.asp

[djpark]

Thanks to everyone for many good advices. I will be happy to hear more.

Though what we are trying to produce may do similar jobs, I believe that where it is applied to will also be very different. BTW, I am in Malaysia.

-- dj

[libor]

Regarding your first question of detecting a jam failure by measuring current:
yes, it is possible, though you should tune your firmware to a power limit adjusted to a particular type of motor (wattage), and filter out (neglect) the increased current when the motor starts. I made my PIC software to use field adjustable current limits, so the installer can adjust the pump monitor to the type of pump it is installed with.
I could even tune my software to detect the different current levels when the pump was pumping water or just running dry (rarely occurs that the floating switch stucks in a closed position, this is also a failure condition, lacking the cooling water would cause motor meltdown after a certain time)

[djpark]

Thanks for the reply.

Roughly how many percent of the current difference do I expect when the pump fails?

-- dj