Fish pond pump

[bills]

I have a 10 watt solar panel and it will run a small pump, how can I set it up to auto engage and shut off without damaging the pump? the solution must be able to be built with through hole on a perf board  or a pre fab board.
I would prefer not to use batteries,just have it pump when the panel has enough current to run the pump 10 watts@12 volts. any ideas?   

[SL4P]

Maybe best approach is to use a micro, and a battery management chip - with 'fuel gauge' capability.  Keep in mind the uP will also draw some (possibly tiny) current when the system is on the battery - whether charging or not.

Then with some experimentation, you can dynamically monitor the charge and discharge cycles - to identify a sweet spot for battery life and effective pump operation.

This method will also help you address variable sunlight exposure times and battery ageing.

[Seekonk]

What is commonly used for remote pumping applications with batteries is a linear current booster, look it up.  Solar panels have low current output at low light levels.  The LCB operates the panel at the panels power point voltage boosting current at a lower voltage, a MPPT controller basically.  I like using a small $2 micro as it is a ready made perf board But it can be done with any switch mode controller.  Small $2 2596 buck converters can also be used if a small motor.

[SL4P]

p.s. I should apologise that we've been responding to a 'battery' strategy, but unless you have extraordinarily reliable direct sunlight, your fish aren't going to be too happy with weak, short duration of water handling.

One issue is noting the 10w being adequate for the pump motor, but that 10w will realistically be 6-8w on a good day, and if the motor stalls, all that will be sunk in the wiring and controller. Buy a battery.

[Rerouter]

I'm at this level myself, look for a constant current driver, in stepdown configuration, have the micro measure the input voltage, turn on the constant current for a second, see how badly it drops, if it remains high enough, then leave it on as needed, if it drops, try again in a minute,

Or if you want to avoid jogging the pump, switch on some load resistance (hey solar is free?) and see if it drops, then decide to turn on your pump,

[Seekonk]

Are you doing a buck conversion at power point or just turning the pump on with enough sun?   Here is a link to the 2596 conversion, you can use two in parallel for more current.  The world has solved this problem and it is the linear current booster.  Batteries are a looooooooooooser.  You can turn the pump on and check the the PWM level every five minutes and wait if it has not reached a certain level. 10W is a little on the low side.

[Seekonk]

Have you given up?  I went searching around for linear current booster and there is not much that turns up.  I'm amazed that something so simple has a price tag of hundreds of dollars.  I have been doing basically the same thing heating water.  Just add a diode parallel to a DC motor and it becomes a LCB.  Shown is an experement I did which  monitors panel voltage with a UNO and drives a FET with PWM.   The booster module, 12V up to 36V, is needed for the heating element but not a motor.

[B.B.Bubby]

I have a 10 watt solar panel and it will run a small pump, how can I set it up to auto engage and shut off without damaging the pump? the solution must be able to be built with through hole on a perf board  or a pre fab board.
I would prefer not to use batteries,just have it pump when the panel has enough current to run the pump 10 watts@12 volts. any ideas?   

Did  you want to protect against dry running or over / under voltage or both?

I doubt a 10w panel could do any electrical damage to a small submerged pond pump, however undervoltage / low rotor speed could cause mechanical wear ( depending on design). A flow switch or sensor will protect against dry running and under speed.

or just stuff it asnd hook it up direct - keep it simple...  whats the worst that could happen?

[mikerj]

Or if you want to avoid jogging the pump, switch on some load resistance (hey solar is free?) and see if it drops, then decide to turn on your pump,

This is exactly what I did for a solar pump controller for a friends hot water system.  When the pump was off, the panel was loaded with an appropriate value resistor and a micro measured the voltage.  When the voltage was high enough for reliable pump operation, the resistor current path is switched off and the pump was switched on.  When the voltage fells too low (use plenty of hysteresis!) then switch back to the resistor. 

This could be done with an op-amp, a voltage reference and a couple of MOSFETs if just a basic switch action is needed.

[Seekonk]

A waste of a perfectly good micro when you could have used  linear current booster technology and given your friend a couple hours more of pumping.

[mikerj]

A waste of a perfectly good micro when you could have used  linear current booster technology and given your friend a couple hours more of pumping.

Jumping to assumptions inevitably makes you look silly...

This was a hot water system with a large solar collector on the roof, and a small electrical solar panel for the pump.  It was pointless switching the pump on unless there was enough sun for the solar collector to provide water that was hotter than the water in the tank, otherwise it would become a solar cooler.

There were a couple of DS18B20 one wire sensors installed, one on the hot water tank, the other on the solar collector output.  When the small panel voltage was high enough, the micro would enable the pump for a few seconds to move some the water in the solar collector to the outlet pipe, and would then measure the temperature.  If the outlet temperature was higher than the tank temperature, it would continue pumping, otherwise it would stop for an adjustable length of time and try again.  This meant that on very overcast days, the system would leave water to heat up in the collector until hotter than the tank before the pump was fully enabled.

The system worked extremely well after spending a few days getting the various parameters set, and the company that installed the panels were actually keen for me to manufacture them commercially as it worked so much better than the original system.

[Seekonk]

Sorry, it would still work better under a linear current booster control.  Deming says whatever you are doing now, no matter how successful, is wrong.  Try not to hold onto old ideas.

[bills]

Thanks for the help I am on vacation now and will be working on it when I get back.
 am thinking of using 2-10 watt panels as the pump I have now draws more then 10 watts.
The pump motor is a PM dc motor. I am concerned about having  current on a stalled dc motor when voltage drops.