Noob capacitor filtering question

Allow me to humble myself and ask another retarded question-

can anyone describe in real easy terms how capacitors work in power supply filtering applications? I know this is a broad question, but I think I need a broad answer, something a little simpler to digest than horowitz and hill...

If you had to explain this to a third grade class, what would you say?

thanks!

dave

They act as batterys. after AC is rectified to DC there is still a ripple, a little sway is still there. so a filter cap will charge up. when the ripple (say... sways minus) happens the cap unloads some of its electricity, and then charges back up when it sways positive. so the DC gets smoothed out.


I think.... :grin:

Think of it as a tank of water.... at first, it's empty... so you fill it up with water from the city mains water pipe. When your tank is full, all is fine. When the city cuts off it's water supply (i.e. some idiot with a bulldozer broke an underground pipe), then there's nothing to worry about... you still have water in your tank. You can go poop and still have water to flush. Now your tank is not full anymore.... keep on using the water, and pretty soon you're going to drain your tank. And your tank will stay empty until it gets filled up again with water from the city water pipe. Hopefully, by that time, the city has water flowing again.

......Now imagine water is flowing into the tank in pulses (60 times per second in the U.S. That makes your toilet hum badly) As the water enters the tank, the pulse is smoothed out. Put a few tanks in a row and the outgoing current of water is pretty darn smooth.

ok I think I got it. So in theory, the more caps you have, the smoother the supply is going to be? Obviously not that simple, but...

dave

The main problem with putting a HUGE filter cap in place of a smaller one in an existing circuit is that a really big cap pulls a lot of instantaneous power as it first charges up, and this can damage the transformer and rectifiers if they were spec'd for a smaller filter cap. It is tempting to put a really big cap in to give better smoothing, but that first charge when you turn it on can really suck some juice through the circuitry behind it.

[quote author="Viitalahde"]One thing I sometimes wonder is that can a cap AFTER regulator be too big?[/quote]
If you make it too big, the shortcircuit protection of the regulator might turn on... Also remember to install protection diodes.

Meaning that does a small cap smoothen the little things better?

Click to expand...

Small caps filter the high frequency "stuff" better. If you are still trying to get rith of 50/60Hz hum after the regulator, a large cap will be better than a small cap.

Best regards,

Mikkel C. Simonsen

If you are still trying to get rith of 50/60Hz hum after the regulator, you messed up rather badly already... adding more caps doesn't fix that. It will reduce ripple but never completely eliminates it... that's the job of the regulator(s), after the main reservoir.

the more capacitance you put after the regulator, the more the chance that on power-off, the regulator is going to see a sustained period of retained charge being present at it's output... not a great scenario. -Of course the primary reservoir shouldn't drain as long as the caps after the reservoir are charged... an exception is if the same rough, unregulated DC is feeding more than one regulator... then if one drags down the main reservoir caps, the other regulator(s) will have power forced back up where the sun doesn't shine if there's a big cap retaining charge... that's where the backwards diode helps.

Some regulators can 'pick a fight' with large capacitances on their outputs, but a small ceramic cap can help a great deal.

Other than that Dave, you're not really oversimplifying at all; it really is basically just that simple. Of course inrush surge is higher on power up as capacitance increases, but as long as everything is within margins, it's all good!

Keith