For a simple circuit to disconnect the load whenever the battery is charging, just use a normally closed reed switch with a current coil around it to disconnect the load whenever the wind turbine is delivering enough voltage to force a charging current into the battery, and have a diode in series with the generator to prevent the generator from drawing down the battery. That meets all of the requirements except waiting until the battery is fully discharged. And it is simple, having only three electrical parts.

That does sound fairly interesting. Is there a specific diode that you'd recommend?

 

Thank you Wbahn! That's the second important thing I have learned from you.

 

The powers can be MUCH different and, in the case you are describing here, the charging power is MUCH higher than the discharging power. But remember that there is no such thing as is conservation of power (between charge and discharge) -- but there IS such a thing as conservation of energy.

If you use a fixed voltage source to charge a capacitor from fully discharged to the source's terminal voltage, the total energy delivered by the source will be twice the energy stored on the capacitor -- the remainder will be lost in the internal resistance of the source, capacitor, and interconnects, no matter how small they may be. If the charging time is very small, the power will be quite high since the average power is the energy transferred divided by the time taken to do so. The average power delivered by the source will be twice that absorbed by the capacitor. Similar -- but more complicated -- issues apply if the capacitor is replaced by a battery. If the terminal voltages of the two are not equal, then the current flowing between the two will dissipate energy as heat as it crosses the differential.

As the capacitor is later discharged to provide energy to some load (amp, LED, whatever) over a longer period of time, the total energy provided cannot exceed the energy that was stored on the capacitor (which is only half the energy provided by the original source) but, because it is being transferred over a longer period of time the average power is considerably less and the ratio of the two can easily be many orders of magnitude.

Can I assume that having a more efficient charging source or a more efficient delivery to the load or even a better load would correct or at least help with the parasitic resistances? Though I do have ideas to improve on each I think 2 of 3 would suffice.

 

That does sound fairly interesting. Is there a specific diode that you'd recommend?

I have not searched through diode specifications recently, but for a regular diode it must be a type with a very low forward voltage drop at the current available from the generator, and the reverse voltage rating must be at least more than the generator output voltage. OR, you can use an FET transistor wired as a diode. That should have a much lower forward voltage drop.

 

I have not searched through diode specifications recently, but for a regular diode it must be a type with a very low forward voltage drop at the current available from the generator, and the reverse voltage rating must be at least more than the generator output voltage. OR, you can use an FET transistor wired as a diode. That should have a much lower forward voltage drop.

Would it be possible if say the generator slowed or stopped producing less current than the battery charged has stored up work that way?

 

Would it be possible if say the generator slowed or stopped producing less current than the battery charged has stored up work that way?

For the connection scheme that I described, as soon as the current from the generator dropped below the current to keep the reed relay contacts closed, the system would revert to a condition with the battery powering the LED and the generator disconnected.

 

For the connection scheme that I described, as soon as the current from the generator dropped below the current to keep the reed relay contacts closed, the system would revert to a condition with the battery powering the LED and the generator disconnected.

That sounds pretty much what I need.

 

That sounds pretty much what I need.

And as the battery voltage would drop, it would take less generator output to start charging the battery. So it meets that specification as well.

 

And as the battery voltage would drop, it would take less generator output to start charging the battery. So it meets that specification as well.

I'm just working on a possible impossibility is all. Can't go into details because my mind hasnt yet given me all the details either hahaha. Just a hunch. It does work with a manual mechanical switch but that defeats the purpose. Talk about a conundrum in a candy coated nutshell.

 

It does work with a manual mechanical switch...

Anything that "works" with a switch can be done electronically instead of manually as long as you can specify the logic that the human has to go through to decide the switch position.

 

Unfortunately there's the issue. A purely mechanical switch isn't controlled by an electrical source thus doesn't add any more impedance or drain to the circuit than the connection it makes. The issue is an electrical switch that uses low power to perform the function with the limited power available. I wondered if a 555timer would do this or if solely using transistor's could perform the same function. (All theoretical ideas here)

 

Unfortunately there's the issue. A purely mechanical switch isn't controlled by an electrical source thus doesn't add any more impedance or drain to the circuit than the connection it makes. The issue is an electrical switch that uses low power to perform the function with the limited power available. I wondered if a 555timer would do this or if solely using transistor's could perform the same function. (All theoretical ideas here)

The ICs you likely need use very little power on their own. It’s hard to speculate without a schematic but I’d guess you can get it done using less than 10mA. Maybe a lot less once the folks here give their inputs.

 

The ICs you likely need use very little power on their own. It’s hard to speculate without a schematic but I’d guess you can get it done using less than 10mA. Maybe a lot less once the folks here give their inputs.

An MCU belonging to the PIC10 family uses an extremely low amount of current (about 75µA tops) and can easily be programmed as a timer

 

The ICs you likely need use very little power on their own. It’s hard to speculate without a schematic but I’d guess you can get it done using less than 10mA. Maybe a lot less once the folks here give their inputs.

It would only need to 1 the charging side and 0 the load the 0 the charging and 1 the load. Just a back and forth switch.

 

An MCU belonging to the PIC10 family uses an extremely low amount of current (about 75µA tops) and can easily be programmed as a timer

Does it also work as a relay?

 

Does it also work as a relay?

A relay is a separate circuit/component

 

It would only need to 1 the charging side and 0 the load the 0 the charging and 1 the load. Just a back and forth switch.

Yes, you can accomplish that switching with MOSFETs or relays. Using relays would require more power to activate the relay but there is such a thing as a latching relay that needs only a pulse of power to switch state, and then it draws no power. MOSFET switching would require essentially no power at all.

I mentioned ICs because you need something to manage the logic and take the place of the human. These would drive the MOSFET or the relay. I have no experience with microprocessors and would approach this project with comparators and maybe a 555 or such. Hard to tell until the logic is defined. As @cmartinez notes, a more elegant way to run the logic would be a low power microprocessor. Using one requires moving up the learning curve, though, and that's just not something I've done. If this is the only project you ever want to do in electronics, I don't think it would be worth it. But if you've got other ideas in your head, you'll be ahead in the long run by diving in and getting comfortable with a micro from the start. If that's the case, you might look at the Arduino. There are tons of online resources relating to it and you can get a lot of help.

 

A relay is a separate circuit/component

My apologies I meant on when needed and off when it isnt.

 

Yes, you can accomplish that switching with MOSFETs or relays. Using relays would require more power to activate the relay but there is such a thing as a latching relay that needs only a pulse of power to switch state, and then it draws no power. MOSFET switching would require essentially no power at all.

I mentioned ICs because you need something to manage the logic and take the place of the human. These would drive the MOSFET or the relay. I have no experience with microprocessors and would approach this project with comparators and maybe a 555 or such. Hard to tell until the logic is defined. As @cmartinez notes, a more elegant way to run the logic would be a low power microprocessor. Using one requires moving up the learning curve, though, and that's just not something I've done. If this is the only project you ever want to do in electronics, I don't think it would be worth it. But if you've got other ideas in your head, you'll be ahead in the long run by diving in and getting comfortable with a micro from the start. If that's the case, you might look at the Arduino. There are tons of online resources relating to it and you can get a lot of help.

Certainly not my only project. I have many that I plan on doing. I probably have components necessary to do what I want but listing them all to know for sure would burden us all. I'm new to this but certainly passionate about it.

 

I think in that case you should start playing with the Arduino. Make some LEDs flash and start learning the basics. As you add to your bag of tricks, it'll be "easy" to knock off the type of project described in this thread.