Bad capacitors

Sounds like normal cap leakage, as measured by a high-impedance scope input. If so, it doesn't need to be fixed.

Measure the voltage across a 1K load to ground.

John

Sounds like normal cap leakage, as measured by a high-impedance scope input. If so, it doesn't need to be fixed.

Measure the voltage across a 1K load to ground.

--------------

Lol. That may make you think it's not there but it is and has nothing to do with the scope.

The 2 issues are:

1. High series resistance. This in series with the leakage resistance forms a voltage divider. This is why your solution doesn't work as you just reduce the DC voltage along with reduce the time constant which goes against maximizing the bandwidth.

Electrolytic caps can have as low as 1M leakage resistance and the common 1M resistance in series for dc blocking results in 50% of the DC voltage across the "load". Teflon caps can have high as 10^14 Ohms which results in almost no dc getting through(effectively what your suggesting as the ratio is about the same).

1. The leakage resistance depends heavily on time because of the self-healing properties of electrolytic and if the time based characteristics you are describing.

One way to measure the leakage resistance is to put a know fixed voltage across a cap in series with a resistance. The leakage resistance forms a simple voltage divider. Measure the junction voltage and then you can compute the leakage voltage. You can use a Wheatstone bridge if you want.

There is not much one can do about the leakage resistance as it is part of the design of the cap. Get better caps. Putting caps in series only makes things worse unless you put a low leakage cap in series with the high one, which generally defeats the purpose unless you want to save money.

An expensive low leakage cap and an inexpensive high leakage cap of the same capacitance will have a much lower leakage than two inexpensive caps in series. This doesn't work for paralleling. Two expensive caps in series won't necessarily be much better but may have other benefits.

You can get fancy and create some active caps if it's absolutely necessary but generally much easier and cheaper just to get good caps in the first place. Although active caps can turn a very high leakage electrolytic of several microfarads into an almost 0 leakage cap and can even help correct some of it's voltage dependent capacitance.

Of course it does. The cap leakage is some number of microamps. If dumped into a 1M scope, or into a 10M probe, the scope will get a lot of voltage across it. A low impedance load, like a 600 ohm or whatever audio device, will have a far smaller voltage drop.

If he measured it with an "infinite impedance" bench voltmeter, he'd probably see volts of DC... essentially all of whatever's on the high side of the cap.

He has many microfarads of coupling cap. It's an audio system. It's

What "1M resistance in series" are you referring to?

John

Late at night, by candle light, John Larkin penned this immortal opus:

ITYHBT

- YD.

Bantam loon ?:-) ...Jim Thompson

Is the audio amp not working? If not, then close it up and leave it.

Why do people still don't know about electrolytics for AC purposes?

Is not. He is seeing DC and thinks the caps are bad. All he's really seeing is the voltage divider formed by the caps and the hi-Z scope.

Electrolytics can also generate a little voltage of their own, chemically. But that doesn't mean they are bad. A 1M or 10M scope will show such voltages, but a 600 ohm load would dischange them pronto.

The point is that a small voltage seen by a hi-Z scope does not mean the caps are bad.

Audio boxes are designed to drive loads. If the load were, say, a speaker, the DC offset he's see would be even smaller.

Why don't you ever address the subject, actually say something on-topic? All you know how to do is lamaly ridicule other people's posts, without saying anything about the actual issue.

I talk electronics, you whine and cluck.

John

No, I think he just doesn't understand much about electricity.

John