Hey guys,

I'm new to the forum and hope some of you can aid me in this electrical dilemma Im having.

I'm currently working on a project where I have about 20 different cells generating some voltage. All the voltages are close to 1V, however all are slightly different (ranging from 0.8 - 1.1V). Individual cell current is an average between 100-150 mA. Ideally I would like to harvest all of the produced voltage and/or current. The problem im having though, seems to come from the difference in voltage outputs. The voltage difference forces the electricity to flow between the cells "balancing" each other out rather than to exit via the harvest wire (At the moment i have harvest wires from each cell connect into one main harvest wire but im realizing that I may need to change this).

My first idea was to use a diode, however the leakage plays a big role with my low V and mA.

Now im working on integrating a MOSFET into the system because the leakage there, from what I searched, is somewhere in the 0.1V range and I can sacrifice that. I plan to power the MOSFET bridge by using an arduino board.

So my questions are:

1. Would it be possible to harvest the produced energy via mosfet?

2. Does each individual cell need an individual mosfet?

3. If I wire the gate lead to arduino, source lead to the cell's harvest wire, and drain to the voltmeter, would that be correct?

4. Any recommendations on which MOSFET will work best? I searched microsemi.com and there are so many different kinds that I'm not sure which will work.

5. Are there any other methods that I should research to harvest the electricity?

Thanks guys, very much. And please excuse my ignorance if anything is out of place. I'm doing my best to research myself as well and not just looking for a free ride. Thanks again!

 

I think you mean the leakage in MOSFETs are in the 0.1 uA range (not volts -please correct if I don't undertand).

1. Would it be possible to harvest the produced energy via mosfet?

Yes, particularly if you operate them as boost converters. Run current through an inductor, then turn off the MOSFET and then let the resulting flyback pulse dump the current to the output, perhaps a capacitor that integrates the current from all sources) through a switch, either an active switch or a Schottky diode.

2. Does each individual cell need an individual mosfet?

Don't know, but a MOSFET per cell would apparently work.

3. If I wire the gate lead to arduino, source lead to the cell's harvest wire, and drain to the voltmeter, would that be correct?

Not really understanding where you would be going with this, if the voltage on the harvest wire is positive and the MOSFET is a P-channel MOSFET, then when the output of the Arduino goes netagive the current through the drain will increase. Note that the voltmeter might give a reading when the Arduino's output is positive because of the leakage current through the MOSFET (o.1 uA x 10 Meg = 1V, doubling every 10° C. You might want to use a load resistor for testing.


4. Any recommendations on which MOSFET will work best? I searched microsemi.com and there are so many different kinds that I'm not sure which will work.

Selection of the MOSFET is probably not critical to the circuit working. The International Rectifier IRFH4209DPbF might be a good choice, but wait until your circuit is designed and all of your voltages and currents are known so that you can select the MOSFET that will give optimum overall performance (particularly in terms of efficiency and cost).

Get one with a high enough voltage rating and the lowest possible on resistance balanced against gate drive losses.

The S.T. Micro STN3NF06 might be a good starting point for this voltage and power range. They are about 25 cents each from Electronic Source http://www.es.co.th/

STN3NF06 datasheet:
http://www.es.co.th/Schemetic/PDF/STN3NF06.PDF

5. Are there any other methods that I should research to harvest the electricity?

At first blush converting to a current from individual cells in an inductor then adding the currents from all of the inductors via switches such as Schottky diodes seems to be the most intuitive, but flyback supplies tend to be inefficient, which might be an important factor.

Other schemes such as using flying capacitors that alternate between being charged from individual cells then switched to be in series to get a higher voltage maybe be more efficient.

Flying capacitor example:
http://encyclopedia2.thefreedictionary.com/Flying capacitor

Incidentally, Texas Instruments has a line of energy harvesting products; if you haven't reviewed them yet, you might want to have a look. Could get some fresh ideas.
http://www.ti.com/lsds/ti/power-management/energy-harvesting-and-solar-charging-overview.page

One hopes that there will be other ideas coming forth from this forum.

 

If the cells are put in series, output current is that of the lowest cell current . Seperate cells by current & form strings of equal current & as close as possible, equal V. Dump all strings into a load resistor adjusted to drop open ckt. V to 80%. Now you have poor mans MPPT.

 

What are you going to do with all this, "free" energy that you sucked out of old batteries?