# Are supercapacitors so super?!

A capacitor (or cap) is an electronic element that stores the energy. Pretty much like a battery, can be used to power things. Well, there are actually substantial differences in the way caps charge or discharge. But let’s focus on the amount of energy a capacitor can hold.

Did I say “power the things”? Well, this is actually how they are commonly used. They are first charged by the supply voltage and discharge soon after to stabilize (or update, smooth) the voltage whenever it drops or fluctuates. Millions and millions of cycles during their lifetime. Far more any battery can withstand. You can find dozens of such circuits in any electronic device. And similar ones – but the idea is always the same: store some energy and give it back.

The catch is that a “normal” cap can only store a tiny amount of charge. Seriously, imagine a red LED, just like one that indicates your tv is in standby. Say a cap is charged to 5v. One very commonly used in electronic circuits has a capacity of 100nF.

^

Whatever nano-Farads (nF) are, 100 nF can lit this LED for… well like 0.000016s. Which only Flash Gordon could enjoy, I believe:) “Standard” caps can be bigger, like up to 1mF (electrolytic), like 10000 times more charge that 100nF – still that does not change too much.

Now: let’s consider a SUPERCAP.

Yes,true, comparing the size – is usually much bigger than 100nF ceramic capacitor. Also, cannot be treated with high voltages (usually up to 5.5v, can be combined). But a 1F (again, do not worry about Farads for the moment) could lit the same LED for like… 160 seconds. What a difference! From a fraction of a second up to almost 3 minutes! Isn’t that SUPER?! Of course you can buy 100F and 400F capacitors as well. For the moment they are more expensive (like \$20 for a 400F), but prices go down at constant pace.

Ok, but how do supercaps compare to a battery, say a similar size coin CR3032? Well, such a battery can power a red LED for like 12 hours (or even days – if only properly treated with additional resistor limiting the current). Comparing to 3 minutes of a 1F cap – are supercaps really so super?!

Yes, they are, just serve a different purpose. Instead of “normal” batteries (which when discharged cannot be brought back to life), try to compare them to an accu like rechargeable batteries (like NiMH), or lithium-polymer cells (LiPo’s). Still – the energy density of a supercap is far smaller (like 5% of LiPo’s). But they can provide quite a current, much more than any alkaline battery or NiMH accu (but comparable to LiPo). On the other hand, it is quite complex to charge a NiMH cell, not mention LiPo or LiIon (2-stage charging required). Caps are soooo easy to charge – just put some voltage across terminals. They are quite predictable, will not blow into your face if you puncture them or die when you discharge them completely (like LiPos). Unless not overheated – can last even dozens of years. Caution is needed to prevent connecting them in a reverse direction or treating with overvoltage.

So when to use a supercap? Industry is hoping that at some stage supercaps will replace batteries in hybrid/electric cars. First prototypes are under tests – called capa-vehicles. Still new materials are needed to overcome limitations of capacity.

When it comes to your DYI projects, use them whenever you have a chance to charge and the discharge current is reasonable. For example, I use them to keep power of a real time clock for my Arduinos. Or for a GPS shield to keep the chip in standby. Or as a backup power to let Raspberry Pi close gracefully. I saw designs where supercaps work great with solar cells. I think there can be dozens of uses. So next time you are to use a battery – think of a supercap!

## More explanations?

LED have the “forward threshold” voltage. It is a minimal voltage that “triggers” a light diode to lit. For a red diode, the forward voltage is around 1.8v (depends on the current but let’s take some common value). So a 5v charged cap will make it lit as long as it provides voltage higher than 1.8v. 5v less 1.8v means a drop of 3.2v.

In terms of charge (say Q,measured in coulombs), which is a result of capacity C (e.g. 1F) and voltage V (between capacitors terminals), we start at 5v:
$Qs=C\cdot V = 1F \cdot 5V = 5C (coulombs)$

End end-up with 1.8v:
$Qe=C\cdot V = 1F \cdot 1.8V = 1.8C (coulombs)$

We have just discharged 3.2 coulombs. Here is the trick: charge per unit Q of time t is… current I:
$I(t) = {{Q(t)}\over{ t}}$

1 amp current is when a charge of 1C flows during 1 second. Let’s change the equation:
$t ={{Q(t)}\over{ I(t)}}$

Say we have 3.2 coulombs discharged at 20mA, which is 0.02A (well, units must match). This gives (at time t):
$t= {{3.2C}\over{ 0.02A} }= 160[s]$

That’s 160 seconds.

Let’s try again for a 100 nano Farad cap.

1pF = enough of these zeros:)

Initial charge is:
$Qs = (100 \cdot 10^{-9})\cdot 5 = (10^{-7} * 5) [C]$

and:
$Qe = (100 \cdot 10^{-9})\cdot 1.8 = (10^{-7} * 1.8) [C]$

where time t:

$t = (10^{-7} * 3.2) / 0.02 = (10^{-7} * 3.2 ) * 0.5 * 10^{2})= 1.6 * 10^{-5}=0.000016[s]$

In fact caps discharge linearly only at part of their characteristics. There are other ways to calculate discharge time that consider non-linearities – but this is just to show you the principle.

##### CR3032

A typical coin battery of such a kind drops down from 3v to 2v giving 250mAh. So if you source 20mA, it can theoretically last for like 12 hours. Possibly less, as 20mA will stress such a battery a lot – because normally they are to operate well under 1 mA. Or more – as such batteries usually have quite a internal resistance limiting the output current. This is why in many toys, LEDs are connected directly to a batter (which normally would require a resistor). So instead of 20mA – you will get only a few mA. A lot of depends on a particular manufacturer and quality of the cell.