Do Audio Speakers Break-in?

[YSC]

I wouldn't be making any generalizations like "all break-in is imaginary". Amir tested one speaker, and he's limited by the precision of the testing method (granted fairly fine resolution and to the limits of what the Klippel currently is used for). In my example it was an old driver that sat on the shelf for a long time. Who's to say what glue or materials are used in every speaker's spider and surround? Also, we don't know what, if any, run-in was done at the factory for the speaker Amir tested. I'm not saying I believe that most speakers have any break-in, only that it appears that it is unlikely that most speakers have break-in. We should all go back to our armchairs until more testing and data is done. I'm certainly not going to assume that break-in is necessary because I never would without evidence, but I'm also scientific enough to realize we do not (yet) have proof that break-in has no effect.

As a side note, I've over-amped a couple speakers over the years. Cheap speakers can show dramatic failures, but what happens to high quality drivers when they get over-driven? It could be elucidating on the sensitivity of the Klippel to try over-driving a couple drivers and see what happens first- does the speaker fail audibly or visibly (damage to back of voice coil for example) first or does the Klippel detect a difference in performance? What other kinds of defects can a Klippel detect? A tear in a cone? Damage a speaker on purpose and find the point where we can measure it...

I am pretty sure if any of those damages especially if coil is damaged it will show up as great FR deviation and distortions, as in the Behringer speaker he tested. Klippel is sensitive enough to detect the tiny speaker variations in the range our ears can't percieve, as in most measuring equipment nowadays, if the break in change is so minimal it can't be detected by a calibrated mic, it is practically non-exist as our ears won't remotely be able to detect that subtle change.

 

[Head_Unit]

Video showing measurements on the same driver new vs. used-->
Woofer Break in

This is evidence that there IS a break in difference with different TS value before/after. Watch the last video If you are in doubt.

Well NO, as someone commented. It may be the same model but it is two different units, therefore proving nothing. Measure a bunch of those new over time and I bet they vary just as much; certainly projects I worked on had a tolerance for Fs. What we need is more actual data, which I looked for once but found precious little of. Time to get a DATS or something...

Many manufacturers even say that speakers need a few hundred hours of breaking in.

You cannot believe that, since it serves to have people keep speakers long enough to get used to the new sound and not return them. Some manufacturers may believe there is break-in but having worked as a marketing whore let me tell you not everything spit out is gospel truth.

 

[Deleted member 48726]

Well NO, as someone commented. It may be the same model but it is two different units, therefore proving nothing. Measure a bunch of those new over time and I bet they vary just as much; certainly projects I worked on had a tolerance for Fs. What we need is more actual data, which I looked for once but found precious little of. Time to get a DATS or something...

I need to understand something about this argument.

If this is not valid data, because of a single unit tested, when does it become valid data? Upon the 2nd? 4th? 10th?
And must the units / speakers / amplifiers / DACs be from the same batch or from separate ones?

And in the line of this argument, can we then discard every single test made here on ASR and other places? Surely the argument must be valid the other way around or is it a "my way or the highway" situation..

A can of worms this is.

 

[DualTriode]

All break -in entirely imaginary.

All generalizations are false including this one.

 

[CtheArgie]

Can we PLEASE separate the concept of break-in, which relates to new speakers, from the effect of (long term) aging of the moving parts of speakers.

The second point in PS’s dribble relates to breaking in crossover components. Here, I think there’s more evidence it is mythology.

On the other hand, I have a dear friend from teenage years (she was so cute!) who is a convinced astrologist.

 

[thecheapseats]

Can we PLEASE separate the concept of break-in, which relates to new speakers, from the effect of (long term) aging of the moving parts of speakers.

in my experience, speakers are consumables like toner for laser printers or automobile tires... on my fourth pair of k+h (now neumann) in forty-plus years...

 

[Kal Rubinson]

Can we PLEASE separate the concept of break-in, which relates to new speakers, from the effect of (long term) aging of the moving parts of speakers.

How do you do this? Is there evidence of an intervening period of stable performance to separate them?

 

[mhardy6647]

My postulation is:

B = P / 100.
Where B is the break-in time in hours and P is the price per pair in €.

So -- I'm reflecting on your model.
It seems paradoxical that, given multiple drivers and crossover components (in a passive loudspeaker, of course), that the macroscopic break-in time of the whole system could be modeled so simply. Does this mean that there's a single component that is (as we'd say in my line of work) the rate-limiting step in the observable quantity, B... or does the model need to be extended to include more parameters (including some, perhaps, which are nonlinear with respect to time) for individual components -- and perhaps even the box itself (glue and other polymers outgas, you know)? Oh -- and I almost forgot the wire inside the speaker box. I am off my game today.

We'll come back to the impact of currency fluctuation on break-in time at a later date.

As humans we crave to simplify and make things we don't quite comprehend into a black and white turf war. This is no different than blind faith.

Wait... Blind faith would be objectively preferable to sighted faith, right? Double-blind faith -- even better!

All generalizations are false including this one.

As Epimenides the Cretin Cretan memorably noted.

 

[CtheArgie]

How do you do this? Is there evidence of an intervening period of stable performance to separate them?

This is exactly my issue/question. What are we (or Paul is) referring to? What happens in the first few weeks from purchase or what happens five to ten years down the road? I’m sure there are different (theoretical) mechanisms.

 

[Lsc]

I never understood why speaker break-in being real or a myth matters so much. In both cases the outcome is the same regardless of what you believe provided you keep the speakers for a while.

For the speaker break-in is a myth crowd, why not just let the folks who believe it enjoy their speaker break-in period. What difference does it make? And vice versa, same thing - enjoy the speakers right away.

 

[Keith_W]

I was curious about whether electronic components really have no break-in or not, so I did some googling and found an old thread on DIY Audio. The author of the thread measured a capacitor on his oscilloscope, let it burn in for a couple of days, and re-measured it on days 2 and 3. He found that a spike on the capacitor's output decreased in amplitude after a few days.

Having said that, 10 pages of googling found no other corroborating evidence apart from listening anecdotes in various subjective sites. I did another search looking for evidence in material science and engineering journals and found no evidence. So my conclusion is that there is no evidence in academic journals that properly ascertains the assertion of break-in of electronic components. However, there is no scientific evidence against it either. Unless there is evidence that electronic components do not change with use, then the only reasonable conclusion is that both sides (for and against) are unsupported by evidence.

 

[DonH56]

Electrolytic caps in particular do have a "break-in" period. Others, not so much.

IME/IMO - Don

 

[theREALdotnet]

Ask any high-end audio dealer, break-ins are a real concern for them.

 

[Deleted member 48726]

How do you do this? Is there evidence of an intervening period of stable performance to separate them?

Well. The easy thing is to follow the manufacturers procedure. After this time, the driver should adhere to spec.
Break in is not the same as wear. A stiff resin coated spider, fresh from the line, may not need much time to get mallable. When it has it may not change it's properties for decades.

 

[valve_integrator]

This is exactly my issue/question. What are we (or Paul is) referring to? What happens in the first few weeks from purchase or what happens five to ten years down the road? I’m sure there are different (theoretical) mechanisms.

I have some excellent wide range drivers from the late 60s no less. (dual cone CG type from RA date 1969).
When measured with proper studio professional condensor mics, they demonstrate very low distortion and excellent frequency response and typically in the 98-100db 1w/1m
Typical dual cone break up above 2khz can be controlled by rolling off well before + using DSP and I have read figures of THD 0.54% at 1KHz, 80db @ 1m

They have doped cambric surrounds which are actually known to improve with age.
Many consider them far superior to the CORAL FR types which are obsolete and hideously expensive.
(Look at the 2% on axis distortion at 100hz!)
I don't personally believe ferrite or alnico weaken with age.

RA did provide some stuff for the BBC LS3/5a, for which exist very numerous forums.
My obvious worry is not overloading mine because they do die (overheated VC -plenty cooked so far).

Now one has to say most of the stuff I read above must be utter b..llx.
Major organisations like the BBC had their own labs and designs.
You can't diss the Beeb, you won't go far trying it.

If there were changes in speaker drivers in such obviously measurable monitors as the LS range, we would have known about it by now.
The BBC were far more thorough in their QC 20-30yrs ago than Amir here, yet I don't hear a thing from them about driver nightmares or massive aging hassles....in fact most BBC LS3/5a seem to hold their values pretty damn well.

"The LS3/5A is probably the most tweaked loudspeaker in the world! The original BBC Research team at Kingswood Warren spent many weeks refining the design and making tiny adjustments. At today's prices the development cost in excess of one million UK Pounds"

"im Finnie also told me how some of the LS3/5A enthusiasts at BBC Kingswood Warren considered it worthwhile optimising the crossovers. This was done by using a high- accuracy measuring bridge to check the value of crossover components, especially the capacitors.
Closely matched components were selected and where required were finally brought up to correct value by adding small values of C in parallel.
The aim was to bring the crossover component values as close as possible to their theoretical values in order to ensure that two LS3/5As in a pair were very closely matched."

commercial success?

Currently the following companies are (still) manufacturing the LS3/5A under licence from the BBC:

Falcon Acoustics

Graham Audio

Stirling Broadcast

Rogers International Ltd

Btw I loved the cable cooker links.
Gems like this are hard to ignore.

"
It's easy to tell when a cable has been "overcooked" -- the sound becomes comparatively dull, "bleached," and the soundstage tends to shrink.
Fortunately, this condition typically passes after the cables are re-installed and played in the system for several hours.

Alan Kafton asserts that cable break-in is long-term but not permanent. He suggests that cabling (especially in the dielectric materials) benefits from a periodic "recharge" of 12 to 24 hours every few months."

'nuff said.

 

[antcollinet]

Unless there is evidence that electronic components do not change with use, then the only reasonable conclusion is that both sides (for and against) are unsupported by evidence.

So - back to Russell's teapot.

 

[antcollinet]

Electrolytic caps in particular do have a "break-in" period. Others, not so much.

IME/IMO - Don

Do you have any data for that? Being that they are not totally solid state, I could see how it were possible. They definately age, as the electrolyte dries out.

But for break in, you would have to show that there is a specific mechanism of change that is quite rapid in early life but then after a few hours of use stops changing, or reduces rate of change almost to zero.

Presumably manufacturers are able to measure that behaviour - so why don't they publish it, like they do all the other specifications (such as the aforementioned lifetime)? This question applies to all other electronic devices that people ascribe "break in" to.

Plus, in the audiophile definition, you'd have to show how those processes both change, and improve the sound, rather than make it worse - or more likely - do nothing.

 

[valve_integrator]

Do you have any data for that? Being that they are not totally solid state, I could see how it were possible. They definately age, as the electrolyte dries out.

electrolytic break in?
Grief, the only thing we have ever done over the years was "reforming" them.
There is good evidence reforming is required particularly for that high voltage stuff (old aged 600-1500V types).
With that 1meg ohm resistor you can actually see the voltages hovering around as it cleans up.

Today 'lytics have shrunk so much, we can only ascribe to vastly improved manufacturing techniques and much lower production spread as a result. Sadly anything over 600V has become unobtanium.

I do see so called "burn-in" on valves. (I have at least 250 of one mil spec type, which are supposed to be vastly better QC than ordinary consumer stuff).
When run up to full voltage from new, they usually run a good deal more current to begin with, then settle.
It makes selecting stuff out into identical pairs and quads, kinda hit 'n miss.

I don't believe any of that crap about "cryo treating" stuff though.
If it were so good, why don't SEAS and others drop their drivers in liquid nitrogen at the factory, like the loonies are doing with their pure silver speaker cables?

 

[Kal Rubinson]

Well. The easy thing is to follow the manufacturers procedure. After this time, the driver should adhere to spec.

That presumes that the manufacturer's recommendations are based on physics and engineering and not on marketing.

Break in is not the same as wear. A stiff resin coated spider, fresh from the line, may not need much time to get mallable. When it has it may not change it's properties for decades.

In principle, yes. However, the issue is not break-in vs. wear in a single element, e.g., the spider, but the physical break-in vs. neural adaptation vs. marketing for the consumer product.

 

[DonH56]

Do you have any data for that? Being that they are not totally solid state, I could see how it were possible. They definately age, as the electrolyte dries out.

But for break in, you would have to show that there is a specific mechanism of change that is quite rapid in early life but then after a few hours of use stops changing, or reduces rate of change almost to zero.

Presumably manufacturers are able to measure that behaviour - so why don't they publish it, like they do all the other specifications (such as the aforementioned lifetime)? This question applies to all other electronic devices that people ascribe "break in" to.

Plus, in the audiophile definition, you'd have to show how those processes both change, and improve the sound, rather than make it worse - or more likely - do nothing.

I'd have to dig, don't have it handy, but it came from the manufacturers' data sheets. That said it is not something I have looked at in years, not my day job. It's something I remember from Cornell-Dubilier, Sprague or similar when I was rebuilding my old tube preamp (600~800 V electrolytic caps). The procedure was the same as "reforming" older caps, basically charging slowly through a high-value resistor, then the value would settle in usually about 10~20% above the rated value (so at end of life they would be about 20% below nominal value, a typical sort of rating for electrolytic caps). It could well be I was looking at "forming" vs. "break-in" but the effect was the same to me at the time (20~30 years ago).

For low-voltage, high-value capacitors used on boards in my day job (note I am not the "capacitor guy") the Murata FAE I was working with on some filter designs said there was some "annealing" that takes place at initial turn-on so there is a brief low-capacitance spike that clears almost instantly. The low-C state was what I had seen and wondered if I had bad caps, but it was not repeatable and I only noticed it as I was looking at transient behavior of some new boards. I do not remember the entire discussion, again not my focus, but have a vague memory of pinholes clearing and dielectric reforming in the first few (milli-)seconds of initial voltage application.