# Using a capacitor

So when a cap says on the side : 10v 100uf does that mean only 10 v can be used on the cap? Or does it store 10v? I put a 10v cap on a servo getting 6v from ext power to remove jitter... It worked. I just want to understand how it worked.

A) what does the 10v actually mean?

B) how did the cap remove jitter?

Thank you.

Bstanko6: So when a cap says on the side : 10v 100uf does that mean only 10 v can be used on the cap? Or does it store 10v? I put a 10v cap on a servo getting 6v from ext power to remove jitter... It worked. I just want to understand how it worked.

A) what does the 10v actually mean?

It dies if you apply more than 10V.

Normally you use a factor of about two just to be safe - use a 10V cap on a 5V/6V circuit.

Bstanko6: B) how did the cap remove jitter?

By acting as a low-pass filter.

Why care about the amount of volts? Or amps?

Thanks fungus for replying, what I meant about farads, is what does the farad represent, and how does it affect whatever it is the farad represents. I know it's capacitance, but what dies a farad do?

Consider the capacitor as a water tank (plumbing used to be used as an analogy for electronics before the purists came along)

If the tank is sealed, the pressure rating (voltage) is the amount of pressure that can be asserted before the tank bursts. You can of course use it at lower pressure. (Voltage rating)

If you fit a drain tap (faucet) to the tank the water will come out at a steady rate (determined by the head of water in the tank) even if the filling water is fluctuating (Smoothing effect)

The larger the tank (volume) the more water it can store so the longer it can discharge after the inflow stops (Capacitance)

The "farad rating" is effectively the volume of the tank

Thank you, very good description.

10V on the cap means that any voltage larger than 10v will damage the cap. Caps act as short circuits or offers very little resistance to high frequency signals. Jitter is caused by noise which are high freq signals. Adding the cap bypasses the noise to ground.

A little bit off topic from the OP's question, but since people are trying to define what "rated voltage" means...

Rated voltage for a capacitor is more complicated than "more than rated and it dies." You need to also consider "when." The rated voltage gives some measure of how thick the dielectric layer is, which varies significantly by dielectric.

In all cases, the dielectric's thickness is such that the "rated" voltage allows the part to pass some qualification / life test. Typically it is 1000 hours with less than 0.1% of the parts failing. Sometimes this means a thicker dielectric, other times, it means reducing the requirements of the life test.

For example on X7R ceramics it is common to run the lifetest at 1000-hrs, 2X rated voltage, 125C. While X5R will be run at rated-voltage, 85C. This means the X5R can be thinner since it doesn't have to survive such harsh requirements. You could also imply the X7R dielectric layer more robust than the X5R.

[quote author=James C4S link=topic=204260.msg1504419#msg1504419 date=1386860055] A little bit off topic from the OP's question, but since people are trying to define what "rated voltage" means... [/quote]

I think this is a bit too specific. The consequences of going outside the rated voltage could be very wide-ranging:

from very minor -- like putting you outside the design criteria for some other part rating, like the temperature/dialectic constant coefficient or some other minor parameter that not many people rely on exactly,

to very major -- like catching fire and burning your house down, as some Tantalum caps might do.

To figure out where you are on that scale requires a lot more than you're going to find in a microcontroller forum. Fear "Very Major" and over-specify your parts and you'll be okay.

Farad is the amount of charge ( in coulombs ) which is stored in the capacitor, per volt of charge.

But it isn't really like a water tank.

If you charge the one Farad capacitor until it has a voltage of 1 volt, then it has one coulomb of charge in it. But if you charge it to 2 volts, then it has 2 coulombs. Each extra coulomb you can stuff in there, increases the terminal voltage by one volt. Until you reach the breakdown voltage of the device where it will fail.

The reason that is not really like a water tank, is that (liquid) water is basically incompressible, you cannot stuff much more water into a water tank by increasing the pressure. It is more like a compressed gas tank.

michinyon: The reason that is not really like a water tank, is that (liquid) water is basically incompressible, you cannot stuff much more water into a water tank by increasing the pressure.

Maybe more like an infinitely tall cylindrical tank. Or a cylindrical tank with base area proportional to capacitance and vertical height proportional to voltage rating.

That's what i meant by using the term "before the purists came along"

The analogy does not have to satisfy all conditions, it is a simplified simile to get a concept over and in that respect it has served its purpose.