[jneutron]

With superconductors you have effectively zero resistance so signals are free to travel in any direction they please. Superconductive materials create a path of least resistance, which is the concept of how the bybee filters work. They are an inside out ferrite core where they attempt to create a super conductor that pulls the signal through a ferrite-like outter shell that converts that noise to heat. They don't quite work like that, but that's the idea.

Superconductors actually work in a somewhat different fashion.  And, there is no superconductivity at room temperature, nor is there any "near" superconductivity at room temperature.


Aluminum electrolytics are made with two foils seperated by a paste.  When the cap is rolled up jellyroll style, the capacitor is rated at approximately zero volts.  Approximately simply because there is a little bit of aluminum oxide at the start.  The capacitor is then "formed by the application of a dc voltage.  This causes an oxide to form and grow on the aluminum surfaces.  As this oxide gets thicker, the rated voltage goes up and the capacitance value goes down.  When the leakage at the forming voltage has reduced to a part specific value, the capacitor is considered "finished", and it is sold.

The forming process cares not where the leakage within the structure comes from.  There is a possibility that portions of the paste are thicker than others, there is a possibility that the Al2O3 is of varying thickness along the full area of the cap.
In all cases, when a capacitor is energized by the application of voltage, there will be a leakage current which can drift with time, and there can be additional forming in the structure where the thickness of the oxide is reduced.
Aluminums are a dynamic element.  They react to the impressed voltage.  Tantalums I cannot say, I have never had to re-form tantalums.  In general, they are easier to blow up, leaving nothing left to reform.

ps..Rod is incorrect with respect to capacitors having no inductance.  A jellyroll structure which has the leads connected at opposite ends will have a solenoidal magnetic field.  I am surprised to see that error.

Cheers, jn

[Rclark]

Oh boy, the heavyweights weigh in, looking forward to reading this later.

[*Scotty*]

jn, if I am not mistaken axial lead wrap and fill polymer film caps also have a solenoidal magnetic field.
Scotty

[jneutron]

jn, if I am not mistaken axial lead wrap and fill polymer film caps also have a solenoidal magnetic field.
Scotty

They can indeed.

The biggest contributor is where the leads are connected.  If both leads are connected at the same location, then all the current will neutralize.  If the leads are at opposite ends of the foils, then the currents will not neutralize when the cap is rolled.
Cheers, jn

[*Scotty*]

Kind of makes you wonder if the can cap physically vibrate or change its physical dimensions when an AC signal is applied due to magnetic field effects. Lorentz Force maybe. Thoughts of added IM distortion run through my mind.
Scotty

[Ethan Winer]

the information you presented does not disprove that capacitor breakin occurs.

I didn't say capacitors never change value or other properties with applied voltage over time. What I dispute is that appropriately selected caps change audibly as is often claimed. It's clear to me that what changes the most is people's perception.

Do you have any links from other sources to substantiate your opinion of .1db?

The burden of proof is on those who make claims that defy years of acoustics research. Further, I doubt you'll accept anything I post, no matter the source. But here you go anyway:

http://hyperphysics.phy-astr.gsu.edu/hbase/sound/db.html#c4
http://www.indiana.edu/~emusic/etext/acoustics/chapter1_loudness.shtml
http://www.deltamedia.com/resource/db_or_not_db.html
http://www.scribd.com/doc/50529329/10/Just-noticeable-di%EF%AC%80erence-JND
http://www.astralsound.com/the_decibel.htm

--Ethan

[BPoletti]

Link please.

http://www.herronaudio.com/images/Measurements.pdf

Page 3.

Maybe we can get some of the old Stereophile gang to chime in.  I heard that you're really close friends with all of them.

 

[jneutron]

Kind of makes you wonder if the can cap physically vibrate or change its physical dimensions when an AC signal is applied due to magnetic field effects. Lorentz Force maybe. Thoughts of added IM distortion run through my mind.
Scotty

Good thoughts are indeed running through your mind.

I'd expect IM to be a measurable entity of course.
Given the current levels, I'd be surprised if that Lorentz guy reared his ugly head in regular caps though.  It's not like we're running a kiloamp per square mm where the forces can add up.

When a cap is made, it is assumed that the capacitance is uniformly distributed through the area of the electrodes, and that the current flowing through the cap also distributes uniformly.  The first is a function of build consistency, the second can be altered by the frequency being pushed and the topology of the connections.  If the connections are such that the current path is a net solenoid, then you are using a parasitic foil inductor within the cap.  A foil inductor does not act nicely at frequency because the current will start to redistribute as a result of proximity effects (aka skinning/Lenz effect).  When this occurs, then all the dielectric path will no longer be available for the current, and the capacitor's effective value will begin to drop.  Note that this is not inductive reactance per se, but a consequence of the inductance causing current redistribution.

I have absolutely no idea if these effects can be audible before, during, or after any capacitor burn in, but merely note the physical effects.

Cheers, jn

[Big Red Machine]

To give a pure mechanical perspective on build consistency:  I started my career working for a telephone handset company and we manufactured our own aluminum/magnet speaker cones.  Given a metal stamping process, and tools do wear and material does vary, even within the same roll, the frequency response of the devices would vary.  Could it be heard?  Sometimes you could outright hear a "bad" one but every unit was tested with a frequency response curve taped to the display that it had to fall within.  We could shim the cone in or out if needed but that was not desirable for high volume manufacture.  So I think that in the manufacture of ANY device, the variation is impossible to avoid.  Will some of these devices change over time?  I am sure they do.  Can I hear it?  Maybe.

As an FYI, companies like Audience, who make capacitors, take the best of the best off their cap line and put them in their Power Conditioners.  Because they vary and they know the parameters they want in their own products they skim the "best" for themselves.  I wonder what they know about change over time, if any?

[Steve]

The burden of proof is on those who make claims that defy years of acoustics research. Further, I doubt you'll accept anything I post, no matter the source.

I must respectfully disagree as you are claiming .1db as fact, not opinion. Therefore, the burden of proof is on you to prove you are correct.

In fact, each link does not address the subject matter, which is individually addressed below.

http://hyperphysics.phy-astr.gsu.edu/hbase/sound/db.html#c4

A useful general reference is that the just noticeable difference in sound intensity for the human ear is about 1 decibel.

The problem is that we are dealing with frequency response changes, not spl intensity levels across the board. So not applicable.

http://www.indiana.edu/~emusic/etext/acoustics/chapter1_loudness.shtml

that is, the minimal perceptible change in amplitude — varies by the starting amplitude and frequency, but in general it ranges between 0.2 and 0.4 dB.

Again, not applicable as we area dealing with frequency response changes, not spl level changes across the board.

http://www.deltamedia.com/resource/db_or_not_db.html

But 1 dB is generally accepted asthe smallest change in level that most people can detect.

Again we are dealing with frequency response changes, not spl levels across the board.

http://www.scribd.com/doc/50529329/10/Just-noticeable-di%EF%AC%80erence-JND

The JND is usually measured using pure sine tones and actuallychanges depending on what frequency is presented to the subject and atwhatdB level the tone is presented. The general rule of thumb, however,is that the JND is about 1dB

We are not dealing with spl levels across the board, but again frequency response changes. Is there another section of the book that does address the issues?

http://www.astralsound.com/the_decibel.htm

our hearing doesn't easily distinguish between differences in level of less than about 3 decibels, but with practice differences of 1 decibel (about 9/8 difference in sound pressure) are perceptible. To most listeners a difference of 1 decibel is "just noticeable", 3 decibels is "clearly noticeable

Again we are not dealing with spl levels across the board, but frequency response deviations.

When one is changing the input Z of the amplifier from 50k ohms to 20k ohms, one is changing the frequencuy response.
As such your links do not address the subject at hand.

As such I have presented my findings, so bowing out.

Cheers.

[Ethan Winer]

Wow, nice to see this thread resurrected. 

http://www.herronaudio.com/images/Measurements.pdf
Page 3.

I see nothing about test methods, frequencies used for the source, number of people tested, how many times, blind or not blind, and so forth. Just one guy making a statement in an audiophile magazine. Versus the links I posted, two of which are on university web sites.

The threshold of "just noticeable difference" for audio volume levels has been well known and universally accepted for many years. My own extensive experience listening in professional monitoring environments backs up the conventional wisdom. The idea that people can reliably hear a 0.03 dB change in volume is like saying you can see the addition of one candle when looking straight into a 2 KW halogen spotlight.

However, I'll be glad to change my opinion if shown credible proof.

--Ethan

[sts9fan]

Great to have this back. Let's stay on the topic of caps. 

[brj]

I'll echo the kudo's for resurrecting this thread.  There are still a few non-productive posts, but there is a lot of excellent content as well.

To JohnR's point - does capacitor burn-in matter enough for people to wax poetic?  I won't claim to have enough experience to say either way, but even if I firmly believed that it didn't matter, I still like to know how things work!

Given that...

When the leakage at the forming voltage has reduced to a part specific value, the capacitor is considered "finished", and it is sold.

Jn, can you clarify exactly where the current is leaking "from" and "to"?

ps..Rod is incorrect with respect to capacitors having no inductance.  A jellyroll structure which has the leads connected at opposite ends will have a solenoidal magnetic field.  I am surprised to see that error.

Just to be clear, is there another fashion in which one could connect the leads in a capacitor, besides one lead connected to each end of the rolled conductor?  (I don't believe you were implying that, and I can't think of how you'd do this and still preserve the behavior of the cap, but I thought I'd check.)

Thanks!

[jneutron]

Jn, can you clarify exactly where the current is leaking "from" and "to"?

For an aluminum lytic, they make it using foil and paste.  After all the manu steps, it does not have a good voltage rating.  The forming process actually forms the oxide that the capacitor relies on for it's capacitance.  During that forming process, it is leaking current from one terminal to the other. 

Just to be clear, is there another fashion in which one could connect the leads in a capacitor, besides one lead connected to each end of the rolled conductor?  (I don't believe you were implying that, and I can't think of how you'd do this and still preserve the behavior of the cap, but I thought I'd check.)
Thanks!

If you connect as per rod's drawing, one lead at each end, then you maximize the inductance.  This is because the current only travels in one physical direction.  Any current that has flowed through the dielectric to the other foil, still travels in the same direction.

If you connect both terminals at one end of the foil, the current will travel down one foil, cross the dielectric, then travel back along the other foil.  Because they are so close together but opposing current directions, the magnetic fields will cancel out for the most part.  If both connections are at one end, then the foil is rolled up around the center, all current will spiral out from one terminal, and spiral back in to the other, so it is the equivalent of two counter-rotating coils.  The theory is exactly the same as a bifilar wirewound power resistor.

ps..here's a link to a reasonable chemistry explanation:

http://electrochem.cwru.edu/encycl/art-c04-electr-cap.htm

Cheers, jn

[*Scotty*]

This sounds like you are describing the Jensen 4 pole electrolytic cap based on the Sprague capacitor developed long ago for use in switching power supplies. Speaking for myself I have certainly heard electrolytic caps before and after they have formed up. In some cases the manufacturer states that their caps will not meet spec until they have had voltage applied to them for some period of time.
Scotty

[Rclark]

Yeah I was gonna ask about that. Designing caps must be difficult in that you have to add burn in time to the design process.

[*Scotty*]

Rclark, if you read the article at the link that jneutron posted, the production process used to make an electrolytic cap is pretty throughly explained. I don't think the forming process gives them much a problem design wise, they just go ahead and allow for it in the specs they quote.
 I think the challenge to the industry is how to design a very low ESR capacitor for use in computers and low voltage switching power supplies that is economical to manufacture. I believe the latest break-through on this front is the solid polymer aluminum electrolytic capacitor. I think the first one was made by Sanyo and marketed as the Os-Con.
Scotty

[Rclark]

 Has anyone had experience with a fresh crossover versus broken in? And by that I mean with drivers that are also broken in as well?

[*Scotty*]

Has anyone had experience with a fresh crossover versus broken in? And  by that I mean with drivers that are also broken in as well?

Danny Richie has extensive experience with both crossover part break-in and driver break-in. PM him with your questions? I am sure he would be happy to answer your questions based on my interactions with him. See link http://www.audiocircle.com/index.php?action=profile;u=87
My opinions on this subject are based on personal experience and I am not a professional speaker designer as Danny is.

[Rclark]

no point in PM'ing, might as well have everyone let it out, heck the way this thread is going, it might as well be a sticky on the issue.

I want to thank everyone who has answered, I don't see getting such responses on any other forum.

I wonder now how cap designers choose the materials they do. What made someone choose beeswax? platinum? What is the design process for an audio capacitor these days? How is burn in factored? It must take ages to select a cap for production...