Capacitor ESR visual demo using the oscilloscope 

indeed

Usually you can do this for power supply filter capacitors BUT the large capacitance is going to have a larger "inrush" current when you first hit it with 12 volts but in this case it seems harmless.

Yes (ideal).

Electrolytics are used in audio power supplies a lot because of the much higher capacitance. They are trying to minimize ripple as much as possible. Ceramics are also avoided to avoid microphonics, but they are sometimes used right next to the op amp if a switching power supply is used. In audio amps with linear power supplies, you won't see ceramic capacitors because they aren't necessary, and it's better to not introduce the potential for microphonics.

It would be relative to a cap of the same type. For example, that 330uf Xicon cap was not bad, it was just not a low ESR type. It measured about 17mv across its ESR in my jig, however, if it measured 50mv, I'd consider it likely to be failing.

JohnAudioTech I guess the datasheet needs to be referenced to determine the design tolerance then. Thx!

Polarized electrolytic capacitors require DC voltage and in the application you describe its positive side, for instance, would idle at 1/2 the power supply voltage and as the signal came through it would vary down to nearly zero and up to nearly the supply voltage. On the other side it becomes AC (alternating current) but that's okay because the capacitor *itself* never becomes reversed polarized. Of course, non-polarized electrolytic capacitors exist for situations where a DC bias is not present.

Fried Mule I think John demonstrated that this is not true. The same value of capacitance in 3 different types of capacitors provided much different response. If the point is to keep the signal clean, it would appear that the ceramic capacitor was even better than the poly and both were far better than the aluminum electrolytic. But if you need bulk capacitance, it is hard to find huge capacitance in ceramic, so you can parallel an aluminum with a ceramic to get the best of both worlds. Make sense?

"Is that right?" That is correct. It is common to parallel a small capacitor across a big electrolytic capacitor. The "getting ready" you speak of is actually inductance in a rolled-up electrolytic capacitor. You can see some odd behavior in the oscilloscope that appears not to be explained. That's most likely the parasitic inductance of the capacitor. Abrupt changes in current will produce a voltage spike on an inductor. So, there's inductance present, not just series resistance.

@@thomasmaughan4798 Thank you very much for your great reply! But would that not mean it is better with lets say 10 small caps than 1 large and 1 small cap? Assuming both solutions have the same total value.

@@friedmule5403 "But would that not mean it is better with lets say 10 small caps than 1 large and 1 small cap?" It *might* depending on various circumstances. Large capacitors tend to have significant inductance that can be a problem at higher frequencies but is not a problem at low frequencies more commonly encountered with large capacitors used as power supply filter caps. The small mylar or cereamic capacitor is a high frequency bypass to keep stray radio signals out of the voltage regulator since that can produce oscillation that destroys the regulator and whatever is hooked up to that power supply. But there's a gotcha: the inductance of the large capacitor will form a "tank circuit" and resonate with the small capacitor at a specific frequency and at that exact frequency it will not only not bypass, it will enhance that signal. So the "Q" is reduced by a series resistor, a low value so we don't lose bypass effectiveness. A VNA (vector network analyzer) or BODE plot can determine the frequency at which this takes place; not that its a big problem but if you have CB or ham radio (for instance) then it could become important.

@@thomasmaughan4798 Thank you so very much, this makes so much sense! Are you a teacher or something? :-)

Persistent overheating of a power supply regulator suggests excessive voltage in the transformer and the transistors and resistors must regulate this down to whatever is the specified voltage. Spikes suggest an inductive load somewhere in the mix, particularly with switching power supplies. A small capacitor in the right place (more than one probably) and probably with a modest resistor in series becomes a "snubber" to control these spikes.

Not very easy. Speakers suffer power compression when the voice coil starts to heat up. It might be possible to measure that with some risk of damage. If the driver has a part number, you can search online.

So i guess the power of a speaker depends on the frequence. first find the ideal frequence of the speaker like you explaned in one of youre video's. And then use that frequence to test the maximum power. (with ear caps :-) )

JohnAudioTech I wonder if you could by monitoring the temperature of the magnet and increase the power slowly? Or maybe use a mic with real time analyzer app to look for change in noise content?

I also wonder about the 22 Ohm value, how you come up with that and if it's value should be changed for testing electrolytic VS small caps?

I can solder the circuit on proto board with better components for improved performance but this is thrown together for the video and to test the caps that came with the LM1875 kit. You could (should) use higher current drive for the low ESR caps to see the differences, but the idea here to keep it all the same to show the relative performance differences.

JohnAudioTech well done!

It could work if the rising and falling edges are fast enough.

Yes, I am doing this in my audio amp project.