Just occured to me WHY AC-coupling blocks DC, but 1 question

I just realized why a cap in series with a signal blocks DC, because of capacitive reactance! DC is 0 frequency so plugging in the numbers for Xc confirms... It was a random :idea: moment I had today as I was driving home. (I've been trying to get farther in AOE, with small baby steps)

One question.
If a polarized cap were used, would it then matter which direction the cap is inserted in series? I would think not, but thought I would ask.

Thanks!

the coupling cap generally blocks a DC bias on one end. Usually, that means you have a squiggly AC signal around some DC offset on one side, but not the other. For example, for a single-stage BJT amplifier, you'll bias the base to about halfway up the rail so it can conduct on both halves of the cycle. You want to be able to feed in a squiggly signal that's centered around 0VDC, so you use a blocking cap. In that case, if it's a polarized cap, you want the positive lead on the bias side, so the total DC voltage is positive.

Sometimes you'll have an interstage coupling cap, where you might have the collector of a BJT (which might sit at say 7VDC quiescent) coupled to the base of another BJT stage (that may be biased at say 5VDC). Obviously there you need to AC couple, and you'll want to put the positive lead of a polarized cap on the higher DC point, which would be the 7VDC collector.

Before anyone lays into me about DC coupling between stages like that, bear in mind I was just trying to make a crappy example

oh, I forgot the main point.

You want to keep positive voltage on polarized caps to keep them formed, no matter where in a circuit they are.

Thanks tmbg

> If a polarized cap were used, would it then matter which direction the cap is

An ideal cap: does not matter.

Many types of caps are near-ideal, in the sense that they don't have a DC polarity. Film, paper, glass, mica, plastic, wax-paper: all symmetric. Insulator does not care which side is positive.

The big exception is the Electrolytic.

Have you ever seen electroplating? Copper, gold, chrome? You mix a liquid with a metal-salt, stick in two electrodes, and connect a battery. One of them gets plated with metal. The other does not. Which one is determined by the battery polarity.

To make an electrolytic capacitor: fill an aluminum pot with alkali. Put in an aluminum electrode, preferably foil for maximum area and capacitance. Apply a battery between the pot and the foil. A large DC current flows. If you get the polarity right, a thin film of Aluminum Oxide forms on the foil. AlO is a good insulator, so the current slows to a stop. Now you have a very thin layer of insulating AlO, with a conductor on each side (the foil and the alkali liquid). A Capacitor! And because of the large area and super-thin insulator, a pretty large cap.

This works fine as long as there is DC voltage of the original polarity. Without any DC, the AlO is eventually re-dissolved, though with modern caps that can take many years. With reverse polarity DC, that AlO film is eaten-off very quickly. Now you have no insulator, just a can of conductive liquid. It acts as a resistor, with unpredictable but usually very-low resistance. In power applications, it sucks all available current and bursts. In audio coupling use, a reversed electrolytic will leak enough to throw your amplifier's carefully designed DC bias scheme out the window, and the amp will stop working.

So: short-term, the electro looks like a capacitor even if voltage goes a little in the wrong direction. But if voltage stays wrong, it might as well be a can of onion soup. So you only use elecros in places that have a for-sure DC voltage, and you make sure to put the cap in the right way.

A pleasure to read.

Analag

Yup - there's some pretty good essays from that hand..

..Thanks PRR..!

Beautiful, Paul. You've got quite the way with words and getting ideas across. :thumb:

Nice!