Hi, I'm developing wireless sensor application running on solar. I'm planning on using supercaps instead of batteries cos it makes so many things easier. Say I have two supercaps, and I need to give them priority, i.e. wireless sensors would have the highest priority and the other supercap would be for additional things. What i would like to achieve is to charge one supercap first and then the other. Does anybody have idea for sensing this? Bear in mind this is solar application and supercap could be charged at different voltages (so maybe sensing no current going into cap would be way to go). To give you better perspective there is no external voltage, so comparator doesn't come into consideration. Many thanks for any suggestions.
Better way: power everything from one (or both to get more capacity) caps and turn off unnecessary devices when caps have low charge left. Many devices have some form of deep sleep consuming only neligible current. Or you can cut their power via transistors.
Maybe I don't understand the question properly, but the energy (Joules (alternately Watt-hours with a conversion factor)) stored in a capacitor is 1/2 C * V^2 so simply knowing the value of the capacitor (C in Farads) and monitoring the voltage (V) across it gives the state of charge. Differencing measurements over some period of time would give the charge rate over that period.
There are some practical considerations of note. Super capacitors have pretty wide tolerances so C likely isn't exactly what it says on the package and there may be some temperature dependence. Typically all of the energy in a capacitor can't be recovered since the downstream circuitry will have some minimum working voltage, so the residual energy at the minimum working voltage would have to be subtracted from the current value to give a "usable energy" value. There are probably other factors, but setting a reasonable lower bound for energy available as a function of capacitor voltage should be pretty straightforward.
MrMark:
Maybe I don't understand the question properly...
...so simply knowing the value of the capacitor (C in Farads) and monitoring the voltage (V) across it gives the state of charge.
Yes, that is if you charge the capacitor with fixed voltage. To make things clear for you, what I'd like to achieve: the problem is solar panels dont give you steady voltage, so if it's cloudy say the cap will charge at 4.4V, when it's charged then the second supercap will start to charge, so both supercaps will be charged. But then it starts to be sunny and the voltage rises to 5.5V but only the first supercap will start charging, when it's charged second will start charging.
MrMark:
Typically all of the energy in a capacitor can't be recovered since the downstream circuitry will have some minimum working voltage.
Know about this problem, planning to use joule thief.
Many thanks for answers, it's kinda shaping up in my head.
, so if it's cloudy say the cap will charge at 4.4V, when it's charged then the second supercap will start to charge, so both supercaps will be charged. But then it starts to be sunny and the voltage rises to 5.5V but only the first supercap will start charging, when it's charged second will start charging.
OK - When current stops flowing (or when current reverses) the cap is fully-charged. (You will have to allow for some leakage current.)
luckyx182:
Yes, that is if you charge the capacitor with fixed voltage. To make things clear for you, what I'd like to achieve: the problem is solar panels dont give you steady voltage, so if it's cloudy say the cap will charge at 4.4V, when it's charged then the second supercap will start to charge, so both supercaps will be charged. But then it starts to be sunny and the voltage rises to 5.5V but only the first supercap will start charging, when it's charged second will start charging.
Unless you're using some sort of maximum power point tracking converter between the solar panel and the capacitor(s), the solar panel behaves more like a constant current source (at constant light intensity) than it does a constant voltage source. The voltage level across the capacitor is entirely a function of its state of charge and it will pull the solar panel down to that voltage (ignoring voltage drops across whatever switch network connects the two). That is, if one simply puts a discharged capacitor across a solar cell under constant light, the voltage will start at zero and rise approximately linearly with time until it approaches the open circuit voltage of the solar panel, or more accurately, the current knee in the diagram below.
I'm not clear on the size and complexity of this proposed installation. For a relatively large scale installation optimized for efficiency then perhaps it's more complicated than I've suggested. In that case a converter that tries to keep the voltage across the solar panel at the peak of the power curve while charging the storage device might make sense, but this isn't trivial to implement.
Yes, that is if you charge the capacitor with fixed voltage.
You seem to be confused.
The voltage across the capacitor at all times directly reflects the total energy stored. The relationship is E (in Joules) = (1/2)CV2. That has nothing to do with how the capacitor is charged.
If you want to power your something from supercap you need very low standby current. I doubt you will get joule thief with suitably low quiescent current. Anyway if you get low voltage parts the energy left in the capacitor is neligible and not worth it.
I still think your idea is wrong from the beginning. It is much better to use one power reservoir and cut power to unnecessary devices when to low power is left.
MrMark:
I'm not clear on the size and complexity of this proposed installation.
It's very small, area of solar panel will be 330 cm2, something similar to jremington's suggested tutorial.
jremington:
You seem to be confused.
The voltage across the capacitor at all times directly reflects the total energy stored. The relationship is E (in Joules) = (1/2)CV2. That has nothing to do with how the capacitor is charged.
Yes, my fault, what I meant by charged capacitor is "charged" at certain voltage. Thanks for the tutorial, helped quite a lot.
Smajdalf:
I still think your idea is wrong from the beginning. It is much better to use one power reservoir and cut power to unnecessary devices when to low power is left.
I'm starting to think this is much simpler than my proposed idea and I could programatically set the threshold voltage when to power things.
I think I've got the idea how to make it, thanks everyone
