Break-in. Best Practice or BS?

[moldyoldy]

I'm interested in some intelligent discussion regarding cone driver "break-in". My current reasoning tells me that a cone driver's suspension has the proper (intended) compliance upon assembly, and that this same compliance should continue for the life of the driver. However, many claim that cone drivers must be "broken in" by continuous long excursions before they will perform as designed. I contend that if a driver's compliance can be altered substantially in a few hours use, that such change would continue with further use, and soon render the driver useless. Do any manufacturers recommend break-in?

[Earl K]

Here,

- See the text that's circled and underlined .

- Obvously ( from the gif. ) extreme mechanical movement does alter the compliance . I doubt that every woofer shipped by JBL goes through such a maximum power test .
- I've broken in many foamed compliances. Each woofer type and compliance type will have a uniquely identifiable shift ( within a limited range ) of its Fs ( effected by compliance ) . This will shift the ts parameters around .
- I guess if a person isn't measuring Fs, then they'll be unaware of this shifting.
- The compliance ( therefore the Fs ) does stabilize at a set figure. But eventually, with continous maximum input wattage ( constant hard usage ), these figures will eventually go down as the hinges ( the spider & the compliance ) wear out .

[Don Mascali]

It is customary to "Break in" a new VLF or LF driver before testing T/S parameters. They change significantly in the first couple of hours of operation.

On the other hand, I always considered the suggestion to listen to new systems for a while before making a judgement as allowing time for your ears to become atuned to the new sound and forget the last system. Ears seem to have a very short memory span.

[Zilch]

There's a measurable, indeed, significant, shift in Fs (and other parameters) with break-in.

JBL specs incude it as part of the test procedure; two hours, as I recall. I just run new drivers overnight at more modest levels than JBL's power test.

My understanding is that it is required for the new materials to establish their flexural "seat," and while the process does continue, it's asymptotic.

Presumably, both spiders and surrounds are involved at low frequencies, but it's easily envisaged that preferred flexural patterns would also develop within the cone and dust dome at higher frequencies.

Interestingly, I did not detect a significant change in Fs when breaking in butyl-surround LE14H-3 drivers; it may be they were run-in at the factory, tho....

[Tom Brennan]

"After several hundered hours use the micro-dynamics became more coherant and the soundstage both broadened and deepened"

And all that bullshit.

[Don McRitchie]

There's a measurable, indeed, significant, shift in Fs with break-in.

JBL specs include it as part of the test procedure; two hours, as I recall.

I believe that this is just done to simulate operating conditions. Documentation that I have encountered on this subject indicated that if you then allow the driver to rest for some period of time and test it again with no power conditioning, it will usually measure back at its original pre break-in numbers.

I remember John Dunlavy posting on this subject on Usenet a few years ago. Back at the time that he was running Dunlavy Audio Labs, he was reliant on third party drivers for his systems. As a result, he did extensive testing on break-in and other factors in selecting his drivers. In most cases, while there were measurable changes after power testing, they usually returned to their pre power test state after a period of rest. In other words, most "break-in" phenomenon he discovered were just temporary dynamic changes and not a permanent physical change.

[Zilch]

Yet in the example Earl cited above (2242H), JBL publishes the post-power-test value (35 Hz) as the T/S Fs.

:dont-know

If the change is not permanent, then woofer testing for box design is a wank? I'm always looking for measured Fs to closely match published T/S....

[Don Mascali]

The average DIY hobbiest probably uses the published T/S data and can hit the mark on tuning. If you were to use the initial data you could miss it by quite a bit.

The speakers would only sound good when you came back from vacation and would creep away from optimum tuning after a little regular listening.

[Zilch]

So, my box tuning varies significantly depending upon how long it's been idle?

This would be a heretofore "undocumented" feature.

[I'd better go turn 'em all on RIGHT now!!... ]

[Don McRitchie]

Yet in the example Earl cited above (2242H), JBL publishes the post-power-test value (35 Hz) as the T/S Fs.

This is not in conflict with what I wrote. Any design should be based on parameters that are consistent with what is expected when the driver is in operation.

There has been much written on this topic on the net if you care to do a Google search. As with any phenomenon, there are exceptions. However, the consensus amongst professionals and experts seems to be that most drivers require little to no break-in. My favorite quote on this subject is by noted speaker designer Ken Kantor: "The thing is, it's a self-solving problem. By the time you come to think you might need to break in your speakers, you already have."

The exceptions generally have to do with drivers designed with inordinately stiff suspensions and specialty designs like electrostats.

[moldyoldy]

Wow! I went to eat after starting this thread, and found this great discussion upon my return!

Actually, I had already researched the topic a fair amount before posting, but I tend to take most of what I read with a grain of salt, so I wanted to hear the opinions of those I trust (you guys). Most info I found came from two camps, (a.) those that wanted to sell me their product, or (b.) those that didn't have a clue and were merely repeating some drivel they picked up on some (other) forum. The occasional recomendations to break-in from mfrs I found offered little or no technical reasoning or test data, so I concluded their real motive was more to break-in the user than their drivers; by recommending extended break-in periods, the user would be more likely to grow accustomed to the sound (and less likely to return the product immediately).

I did find one reference that offered signifigant test data, as well as AES and other credible references;

http://www.audioholics.com/techtips/...kerBreakIn.php

The author's findings were largely in agreement with Don's (thanks for your input!) statements, in that extended long excursions did (slightly) 'loosen' mechanical compliance, but the effects were largely very short-term, and the original mechanical compliance returned after a short period at rest.

So, at this point, (though I'll keep an open mind to be convinced otherwise), I'll conclude;

BREAK-IN IS BS. (Edit; verdict in temporary suspension)

(RE-EDIT); BREAK-IN OF SPEAKERS BY THE END-USER IS NOT REQUIRED, WHICH IS ALL I MEANT TO PROVE IN THE FIRST PLACE.

[Ducatista47]

Audioholics has quite a library of debunking articles and studies. When I hear some audiophiles piffling on or get too full of myself, I give their site a look and come back down to earth.

Clark in Peoria

[moldyoldy]

For those that would like to do some testing to confirm or debunk the break-in issue, I'd like to suggest that in addition to any "active" testing you perform, you also try some "passive" tests. Mechanical compliance can be measured by a force/deflection-distance test, and individual suspension components (spiders and surrounds) are rated this way, typically in mm per gram. Any savvy experimenter can devise a way to do this without accellerometers or other fancy gear. Next time you have a new woofer on the bench (remember the issue is breakin of new drivers), try this;

1.With the driver face-up on the bench, lay a straightedge across the frame (not the gasket, which could be later compressed and compromise the test), and accurately measure the distance to the apex of the dust cap, or another repeatable reference point.

2.Place sufficient weight at the cone/cap junction to depress the cone signifigantly, preferrably enough to simulate the inner Xmax limit. (BBs work well, as they evenly distribute the weight, and can be counted if you have no gram scale).

3. Measure and record the distance of deflection, as well as the weight it took to get there.

4. "Break-in" the driver any way you care to, as long as you don't exceed Xmax doing so. On completion, let the driver sit overnight.

5. Remeasure the cone deflection using the same weight and measurement points as before. Compare the results.

[Zilch]

The author's findings were largely in agreement with Don's (thanks for your input!) statements, in that extended long excursions did (slightly) 'loosen' mechanical compliance, but the effects were largely very short-term, and the original mechanical compliance returned after a short period at rest.

So, at this point, (though I'll keep an open mind to be convinced otherwise), I'll conclude;

BREAK-IN IS BS.

It's unclear that we just read the same article, actually.

While it it reassuring to learn from Sanfilipo's simulation analysis that the changes documented to actually occur have little effect on performance in the completed system, he clearly describes two classes of break-in parameter shifts.

The first is a permanent shift associated with the mechanical factors I described above, clearly of significance to anyone designing a system around measured parameters:

"When a newly-minted driver rolls off the assembly line it will typically not measure or otherwise perform as one of its well-broken in siblings does. In large part these differences arise owing to the fact that said driver’s spider (often crafted of a varnish-impregnated linen) is not as compliant as it will become once it’s broken in properly. Depending on number of mechanical factors relating to its design, construction and materials composition, a net post-burn in compliance increase on the order of 5% - 10% would not be unusual."

Sanfilipo assumes this initial break-in occurs during the process of manufacture and testing of completed systems, which he admits may not be the case, but, according to his analysis, this is of no consequence; the influence upon performance of the in-cabinet finished product is minimal.

However, to a system designer relying on actual measured driver parameters, that difference is quite significant indeed, and the initial break-in must be accomplished for reliable results. A new driver received in a box from the factory will not necessarily have been subjected to the requisite process, and a newly reconed driver almost certainly will not:

"Required break in time for the common spider-diaphragm-surround is typically on the order of 10s of seconds and is a one-off proposition, not requiring repetition. Once broken in, the driver should measure/perform as do its siblings, within usual unit-to-unit parameter tolerances.

Probably the most common approach used by manufacturers who purposely take the time to break in raw drivers is to apply a sine wave signal, at a frequency equivalent to the unit’s free air resonance, delivered at amplitude sufficient to thoroughly stretch out the spider, without damaging the unit, of course

An alternate approach referred to in the literature is the use of broad band noise. However, this approach is inefficient when compared to the sine- wave-at–free-air-resonance approach."

I use the inefficient broad band noise approach for some assurance that whatever changes may occur happen across the entire frequency spectrum.


Sanfilipo further describes a second, dynamically reversible parameter shift, of essentially the same order of magnitude, that occurs in actual operation, which he attributes to "temporary elastic deformation of the surround," from which the driver spontaneously recovers:

"Cease stimulus and the driver’s compliance will return – in most cases within seconds or minutes depending upon surround design, material composition, ambient temperature and so on - to its pre-stimulus value; the compliance changes are temporary."

While the changes of this class may be of interest to the designer, and he may indeed consider and compensate for them, they are apparently not altered by the break-in process.

Thus, whether break-in is BS or real and significant varies depending upon the purpose and perspective. It may matter not a whit to the end user, but it is extremely important to the system designer....

[moldyoldy]

Can't pull any tricks on the Zilchster.....

I knew if anyone would thouroughly read (and grok) that article it would be you!

OK, I said I'd keep an open mind, so I'll give you the "10s of seconds". Heck, let's be generous and call it a full minute. Still, I'd hope that any reputable mfr would have surpassed that in even the most cursory of QC testing, long before the consumer opens the box. (I know, ass-u-me).

I also fully agree with your designer's standpoint issues, but most of us aren't on that end of the food chain. Remember, this is in General Audio, not DIY. I'm seeking the best answer for the masses. I suspect that drivers are being damaged by inappropriate methods used by the uninformed.

I have done the force/deflection test on two of my own recones, a 411 with foam, and a 421 with a doped accordian. Free-air, Xmax excursion, sine wave at Fs for 30 minutes each (had to leave the room, as I couldn't bear to watch). No measureable difference in the mechanical compliance of either.

To be fair, I'll suspend the (premature) verdict while the jury deliberates.....