How to Work With and Store Very Low Voltage?

[tommygdawg]

Hello all! First time here, but have been following Dave's videos for quite a while, figure this was the place to come with my burning questions!

I'm trying to make a very basic miniature wind turbine to charge a battery. I'm using a small DC motor attached to the turbine as my generator. Just giving the turbine a spin with my hands I can get anywhere between 100-250mV. Just for fun, I'm also playing with those little TEC units which also give quite small voltages, usually around 100mV or lower.

My question is, how do I make this usable? I've done a fair bit of Googling and playing with my multimeter and I've come up with some ideas, maybe you all can tell me if I'm on the right track.

What seems to be the best idea for storing electricity with very low voltage is to have a bank of capacitors wired in parallel. For simplicity sake, let's say I have 10 caps each getting 100mV, which would give me a total 1V (I'm aware my total capacitance will go down in this configuration). Lets say once each cap reaches 100mV (or perhaps at a set time interval, regardless of cap charge), they switch to a series wiring and thus all dump their charge in one go, giving me 1V to my destination. My question here is, how on earth do I wire that? It's mind boggling!

Secondly, it took me ages to figure out that my problem even keeping a cap charged was the cap draining back to the source. For instance, I would attach a capacitor to a small solar panel, and if I exposed the panel to light, the cap would charge. If I covered the panel, the cap would consistently drain quite quickly, if I unplugged the panel, the cap stayed fairly consistent in charge. My conclusion was that when the panel is attached but not receiving enough light to generate hardly any voltage, the cap starts draining back into the panel, until the very small voltage output of the panel more-or-less matched the charge of the cap. Is this correct?

Now that I think I figured that out, let's simplify everything: 1 cap and 1 power source (turbine, solar, whatever). I'm only getting around 100mV. How do I keep the cap from draining back into my turbine or solar panel once the source stops generating? At voltages this small, even a Schottky diode would be useless. If I'm understanding things correctly, once I figure *this* part out, then I can move on to figuring out how to wire them up with a switch or timer that will flip the configuration from parallel to series every so often.

I know that's a long post, but I tried to be as detailed as possible! Any help would be greatly appreciated.

Thanks!
 - Tommy

[mariush]

Look into energy harvesting ICs or boost regulators that start at very low voltages.

See for example : http://www.linear.com/parametric/energy_harvesting 

Other companies have similar selections of chips.

[macboy]

You might want to look into "Joule Thief" circuits.

[max_torque]

The issue you are finding is that at very low voltages, the power available to actually do anything is tiny.  Whilst modern microprocessors can flip bits at say 1.8V, in the real world, anything that takes any power will need more voltage (and more power).  You say you want to charge a battery, well, with say 250mV, even if you could source 1A (which i bet you can't) that's just 1/4 of a Watt, which for any battery bigger than a coin cell, isn't really going to do a lot.

If you just want to light an LED to show something, then rewind (or replace) you generator to produce a higher voltage at the same speed, so you can now use a normal diode as the anti-reversal method  (or attempt to gear the generator higher, so it spins faster)

[katzohki]

Yeah, I really don't think there's much you can do with such low power levels, but in general I think your idea of storing in parallel and sourcing in series is an interesting idea to experiment with. You may not create the next big thing, but it will surely be a valuable learning exercise.

I bet you could switch the arrangement from series to parallel and back using some simple transistors, but with such low voltages you might run into the diode drop in them to be a problem. I guess relays would work, but switching them will probably use all of your stored energy.

See if you can get some supercaps (like 1 Farad or something) or some LiPo batteries to experiment with.

[tommygdawg]

Thanks for the thoughts, all. It should be noted that I'm not at all trying to create the next big thing. I'm just experimenting and having fun and trying to figure out how all this stuff works. I'll look into picking up a couple chips from Linear Technology and experimenting with those.

[dannyf]

I would use capacitor + relays (solid state or otherwise). The relays would configure the capacitors to be charged off by your source and then in serial to power an external load.

[vitUP]

What seems to be the best idea for storing electricity with very low voltage is to have a bank of capacitors wired in parallel. For simplicity sake, let's say I have 10 caps each getting 100mV, which would give me a total 1V (I'm aware my total capacitance will go down in this configuration). Lets say once each cap reaches 100mV (or perhaps at a set time interval, regardless of cap charge), they switch to a series wiring and thus all dump their charge in one go, giving me 1V to my destination. My question here is, how on earth do I wire that? It's mind boggling!

http://en.wikipedia.org/wiki/Charge_pump

[vargoal]

With respect to the capacitors draining through the solar panel I would try using a diode to prevent current from flowing away from the capacitor

[magetoo]

A diode will prevent current from flowing backwards, so that would work for preventing the capacitor from discharging.  But a diode is not a perfect one-way valve - you also need a small voltage across a diode to have any significant current flowing through it in the forward direction.

OP has very little voltage to work with, less than necessary to make a diode conduct.  That's what he means by "at voltages this small, even a Schottky diode would be useless".

(Sorry if I'm just restating the obvious here.)

[tommygdawg]

That is indeed my main problem. I've just learned about super diodes, or precision rectifiers, I'm still in the process of researching them. Seems like they would be perfect for my needs, but there's got to be a catch to using them, aye?

[edavid]

That is indeed my main problem. I've just learned about super diodes, or precision rectifiers, I'm still in the process of researching them. Seems like they would be perfect for my needs, but there's got to be a catch to using them, aye?

No, not relevant.

You could just add a gearbox to your motor, and make a stack of TECs.

[tommygdawg]

Why is it not relevant?

[electr_peter]

Super diode/precision rectifier uses additional circuitry and additional external power to implement "ideal diode" action with no voltage drop. It is a signal conditioning trick useful in some applications. It is not at all useful for your case.

[con-f-use]

[Rick Law]

...I'm using a small DC motor attached to the turbine as my generator. Just giving the turbine a spin with my hands I can get anywhere between 100-250mV...
...My question is, how do I make this usable?...What seems to be the best idea for storing electricity with very low voltage is to have a bank of capacitors wired in parallel.

For simplicity sake, let's say I have 10 caps each getting 100mV, which would give me a total 1V (I'm aware my total capacitance will go down in this configuration). Lets say once each cap reaches 100mV (or perhaps at a set time interval, regardless of cap charge), they switch to a series wiring and thus all dump their charge in one go, giving me 1V to my destination. My question here is, how on earth do I wire that? It's mind boggling!
...
 - Tommy

I discern that you may be mixing a couple of concepts together and that may be confusing you.  If you separate them, you may understand the problem better.

Voltage does not equal Energy.  Increasing the voltage in this context does not increase the energy.  You have one "turbine" (1 dc motor) and the amount of energy generated by the turbine doesn't increase by storing it different.  Increasing the voltage will increase efficiency and will allow you to cross certain threshold such as activating the diode avalanche.  But the total energy you stored is still just the energy of the lone turbine can create - stored in one or spread into 10 capacitors.

Think of it as butter on your slice of bread.  You have 1 lump of butter (1 turbine) and it is not enough for your 1 slice of bread (1 capacitor).  If you scrape off butter from other slices of bread and consolidate on your slice of bread, that would give you more - so far so good.  But if you start with that same 1 lump of butter, spread it on 10 slices of bread, then scrape them off to consolidate them onto 1 slice of bread, you still have just 1 lump total on your 1 slice of bread (assume perfect efficiency).

On the other hand, if you have 10 lumps of butter to begin with...

So, perhaps think about arrange the 10 turbine (DC motors) in serial each generating 100mV, now you have 1Volt and actually 10x the amount of energy.  It will actually have 1volt and 10X the energy.

Spreading it to 10 caps and re-consolidate will only work if the total amount of energy is sufficient and only problem is voltage being too low, which doesn't seem to be the case described.

[tommygdawg]

Thanks, Rick! Your analogy is very helpful! I have a question regarding it.

The idea of the 10 caps is that I can only fill each cap to a max of 100mV. So, if I have 10 caps at 100mV, would I then not effectively have stored 1V (assuming I could then put them together in parallel)?

I guess what I'd like to do is accumulate energy. Surely there must a way to do that? IE, instead of spinning 10 turbines for 10 minutes, I'd spin 1 turbine for 100 minutes. This was my thought with the capacitors, but perhaps I'm still confused.

[Rick Law]

Thanks, Rick! Your analogy is very helpful! I have a question regarding it.

The idea of the 10 caps is that I can only fill each cap to a max of 100mV. So, if I have 10 caps at 100mV, would I then not effectively have stored 1V (assuming I could then put them together in parallel)?

I guess what I'd like to do is accumulate energy. Surely there must a way to do that? IE, instead of spinning 10 turbines for 10 minutes, I'd spin 1 turbine for 100 minutes. This was my thought with the capacitors, but perhaps I'm still confused.

I don't know low energy well enough to answer your question, I suspect it is not something off-the-shelf component can handle. 
Perhaps someone more familiar with this subject matter will be able to render more help.

[IanB]

Thanks, Rick! Your analogy is very helpful! I have a question regarding it.

The idea of the 10 caps is that I can only fill each cap to a max of 100mV. So, if I have 10 caps at 100mV, would I then not effectively have stored 1V (assuming I could then put them together in parallel)?

I guess what I'd like to do is accumulate energy. Surely there must a way to do that? IE, instead of spinning 10 turbines for 10 minutes, I'd spin 1 turbine for 100 minutes. This was my thought with the capacitors, but perhaps I'm still confused.

You need to think about the problem in terms of power. When you spin the turbine blades you feed a certain amount of mechanical power into the turbine. The turbine is connected to a generator and the generator converts some of that mechanical power into electrical power. The ratio of electrical power out to mechanical power in is the efficiency of the system.

Now the electrical power coming out has what we might call a "quality". Low quality power is delivered at a low voltage, while higher quality power is delivered at a higher voltage. We use the word quality here because lower voltage power sources are in general less useful than higher voltage power sources.

So the question becomes, how to increase the quality of the generated power? One way, the simplest way, is to spin the turbine faster and harder. You won't get much out of it if you just spin it with your fingers, because your fingers don't have much power in them. Try spinning the turbine in a strong wind, with bigger blades, and maybe with a gearbox to speed up the generator. These all could help to get more power and higher quality power out of the system.

Another way, assuming the voltage is still too low, is to put a step-up voltage converter on the output of the generator. Maybe you can turn 0.5 V into 2.5 V using electronics, and then store the output in a battery or super capacitor. The step-up converter won't increase the power available, but it will make the available power more useful.

[DJ]

Think of capacitors as buckets. Some are leaky (self discharge rapidly) and some are not (air/teflon and other dielectrics in decreasing order). You want good buckets to store charge, and efficient (low-loss) valves to fill, empty and transfer the contents of those buckets.

Low potential (pressure/voltage) circuits are available (Linear Technology apps are a great start), but I would look first for an efficient generator that outputs a higher voltage (a few volts at least). This can make switching and other parasitics less of an impact.