Speaker Break-in...Fact or fiction? (1 Viewer)

[Tom Brennan]

Regardless of the actual merits of break-in I don't think that listening to 2 sets of identical speakers, one "broken-in" and one not is very telling. You have the difference that would be caused by variations of the speakers within the maker's tolerances and also the 2 sets of speakers can't occuppy the same space at the same time. Speaker positioning causes a noticeable difference in sound.
www.chicagohornspeakerclub.org

 

[Chris Tsutsui]

So if speaker break-in is not true... then that means the rubber surrounds that are first molded are the same stiffness throughout the entire life of the speakers.

The laws of physics can easily describe how something like a speaker driver will change after undergoing flex, wear, and tear.

However, the argument at hand should probably be more along the lines of "Does my ear detect speaker break-in". Which we now enter the subjective realm of human hearing where all you will hear is opinions on your ability to hear.

So Frank, do you at least admit that a driver changes... even if it's inaudible to your ear?

 

[Michael R Price]

Has anyone noticed a tweeter break in? Or is this just having to do with woofer suspensions?

 

[Chu Gai]

I don't think its so much that break-in isn't true Chris, it seems from what I've read and understood is that the changes that occur in break-in seem to be rather offsetting and further that the variations are well within manufacturing variations for drivers. It's as if i were to shorten a wall on one side but extend it on the other. If that's the case then break-in is much ado about nothing insofar as the characteristics of the speaker changing over time. The studies that have been done suggest that the surrounds do stretch or become more compliant with play (that's in part to the heat generated) but that they return to normal. Kind of like if you were to measure a rubber band...give it a good stretch and remeasure. It's a little longer. Now give it a few hours and remeasure once more. Provided you haven't stretched it beyond its modulus, it'll come back to the same length. Like I've said previously, if nothing else, this information is good for the consumer and can prevent one from being snowed by the 'occasional' misinformed or desparate salesperson.

 

[Mal P]

Yes, absolutely. Driver parameters certainly do change from "out of the box" and therefore speaker parameters can change. Is this audible? It depends on the degree of change.
But I do have evidence to prove my claims, or at least, evidence provided by a popular speaker manufacturer:
Driver break-in test
I own the speaker that the driver mentioned above is used in, and I can say confidentally that the bass definition had audibly changed with the same piece of music at 0 hours break-in, as opposed to 20 hours break-in.
Here's a cut and paste:
Results of detailed running in.
Driver: Excel W21EX001
Manufacturer: SEAS Fabrikker AS Norway
Serial #: 2/00-50 and 2/00-48
Manufactured: February 2000
Overview.
Tests were conducted to qualify the existence of, and/or quantify the
magnitude of, changes in electroacoustic parameters of a loudspeaker driver
over time.
The aim of the test described herein was to measure the parameters of a new
driver, by impedance measurement and analysis, and compare them with the
same parameters measured after specified periods of 'running in' with a
suitable test signal. The results were then examined for evidence of time
related changes.
No attempt was made in this test to examine the parametric behaviour over
long time scales. This will be done in another series of tests to accurately
establish the existence or otherwise of such long term changes.
No attempt was made to examine the relative effects of different 'running
in' waveforms.
Test Conditions.
Tests were conducted in VAF Research anechoic chamber, solely for the
purpose of noise isolation. No acoustic measurements were taken. The Drivers
were suspended asymmetrically in free air at least 600mm from the nearest
(heavily absorptive) surface.
Temperature was controlled at 21degrees C +/- 2degrees as dictated by the
thermostat hysteresis of the airconditioning system.
Test signals.
The 'running in' signal was generated using proprietary VAF signal
generating software. The signal comprised a swept sine wave from 10Hz-100Hz
with a sweep time of 1 second. This was applied to the drivers at a level of
4 Volts p-p (2 Volts peak, 1.414 Volts RMS, approx 0.25 Watts). This level
was selected arbitrarily because it results in significant excursion of the
driver cone (approx +/- 2mm) at the lowest frequencies but does not approach
the maximum linear excursion of the driver (specified by SEAS as +/- 6mm).
The impedance measurement waveform was generated by a proprietary MLS
Measurement System and constituted a low frequency MLS signal sampled at
1.9kHz. This was applied at a level of 4 Volts p-p. Impedance measurement
bandwidth was 10Hz - 200Hz. No smoothing or windowing was applied.
The resonant frequency Fs was taken to be the frequency at which the
impedance phase angle passed through 0 degrees.
This method has been verified to our satisfaction in comparison with manual
swept sine wave tests and commercial MLS impedance measurements.
It has the advantage of exposing the driver to very low average power and
hence minimises thermal effects.
Method.
The impedance of the driver was tested by application and analysis of the
MLS impulse. The driver was then subjected to the running in signal for a
fixed period as shown in the results below. The parameters were again
measured immediately to allow for quantification of driver thermal effects.
The driver was then allowed to cool by natural convection for ten minutes.
The parameters were then measured again and the cycle repeated.
Additionally, a control driver was measured once at the commencement of the
test and again at its completion. No 'running in' was applied to this
control driver. results are included below for comparison.
Results.
Sample 2/00-50
Conditions / Total RunIn / Fs / Qts
---------------------------------------------
Initial Test / 0 min / 35.47 / 0.49
Run in for 1 minute
Cool down 10 minutes / 1 min / 34.7 / 0.42
Run in for 1 minute
Cool down 10 minutes / 2 min / 34.64 / 0.43
Run in for 1 minute
Cool down 10 minutes / 3 min / 34.63 / 0.44
Run in for 1 minute / 4 min / 34.34 / 0.43
Cool down 10 minutes / 4 min / 34.54 / 0.44
Run in for 1 minute / 5 min / 34.41 / 0.41
Cool down 10 minutes / 5 min / 34.65 / 0.43
Run in for 5 minute / 10 min / 34.33 / 0.43
Cool down 30 minutes / 10 min / 34.68 / 0.43
Run in for 10 minute / 20 min / 33.35 / 0.42
Cool down 30 minutes / 20 min / 33.65 / 0.43
Run in for 20 minute / 30 min / 32.14 / 0.42
Cool down 30 minutes / 30 min / 32.80 / 0.42
Run in for 30 minute / 60 min / 32.03 / 0.41
Cool down 60 minutes / 60 min / 32.06 / 0.42
Run in for 60 minute / 120 min / 31.23 / 0.41
Cool down 60 minutes / 120 min / 31.62 / 0.41
SEAS Specification Fs: 31 Qts: 0.41
Sample 2/00-50
Initial Fs: 35.38 Qts: 0.48
Final Fs: 31.62 Qts: 0.41
Conclusion.
It has been demonstrated that the Fs and Qts of a high quality bass driver
change from their initial values with use. The rate of change appears to
decrease as the final value (in this case very close to the values specified
by the manufacturer) is approached. This would support the theory that this
driver will settle down to parameters somewhat different from those tested
when the driver was new. The difference in Fs in particular is approximately
11% which, with the changes in Qts also noted, could significantly alter the
final alignment of a system using this driver. Note that changes in Vas were
not monitored and so an accurate estimation of the alignment error is not
possible.
Best regards
Philip Vafiadis
VAF Research Pty Ltd
www.vaf.com.au

 

[Frank Zimkas]

Chris,
The short answer is yes, I do believe that the characteristics (sp) of a speaker can change over a long period of time. Are those changes audible? I'll be darned if I know! Thats why I started this thread. There seems to be evidence on both sides of the arguement.

Hey I'm just trying to learn something here, thats why I asked for hard data and not opinions. We've all got opinions on this subject, how we formed them is a completely different subject. A lot of what we hold to be true about Ht or Hi-Fi had to come from somewhere. Was it a salesman that was blowing smoke in the general direction of our collective hind quarters or did it come from a knowledgeable/reliable source.

My personal opinion is that the speakers don't break in but that over time they can "Sound" different because we get used to them. I may be way off the mark here but thats my opinion, don't let it effect yours.

 

[Mark Austin]

Thank you Mal, for the very informative post. Maybe someone could forward this to Nousaine and get his thoughts.

 

[Chu Gai]

You'll note Mark that SEAS stated they didn't measure Fas which Nousaine and other's had. Perhaps their later work if it becomes available will illustrate the offsetting changes. The Qts change I believe had to do with the recentering of a fresh driver which Clark found and VAF also determined. As you saw, after the initial burst of signal, the value changed from it's initial value and then settled in. So what specifically is different Mark?

 

[Larry B]

Chu:

Larry: hardly a conspiracy theory as various measurements have been performed

By "conspiracy theory" I was referring to your proposal that speaker break-in is actually a marketing ploy designed to get consumer to keep their speakers past the point at which money-back returns are allowed.

Reagrding Nousaine's measurements, am I correct in assuming (oops...) that you believe that measurements reveal the entire audible properties of a speaker? If so, then it should be possible to determine from measurements alone whether one will like a particular speaker without even listening to it.

Larry

 

[Larry B]

Tom:

Regardless of the actual merits of break-in I don't think that listening to 2 sets of identical speakers, one "broken-in" and one not is very telling. You have the
difference that would be caused by variations of the speakers within the maker's tolerances and also the 2 sets of speakers can't occuppy the same space at the
same time. Speaker positioning causes a noticeable difference in sound.

Then it should not be possible to ever compare any speakers to one another.
Larry

 

[Mark Austin]

What is different?

SEAS Specification Fs: 31 Qts: 0.41


Sample 2/00-50

Initial Fs: 35.38 Qts: 0.48
Final Fs: 31.62 Qts: 0.41

 

[Chu Gai]

Larry:

I've been doing this speaker design thing as a professional
consultant for something like a quarter of a century. Over that
period, time and time again, I have been humbled by the
rediscovery of how UNreliable the human ear is as an analysis
tool. From the expectation bias that plagues us all to fatigue
and more.
Moreover, the ear might be able to tell us that something's
amiss, but it is LOUSY at telling us what and by how much. You
might spend days futziong with trying to fix "something" that
manifests itself audibly as masking the impact transients on a
triangle being continuously rung and get absolutely nowhere.
But, taking the same speaker and running a detailed spectral
decay plot told me in under 2 minutes EXACTLY what the problem
was and how to solve it.
Certainly, listening can help tune a unique design to an
individual's personal PREFERENCE, but that is NOT the same as
designing a system, testing and identifying its problems, and
then verifying that your solved those problems without creating
others. For that, the experience is that listening is
frustratingly misleading and unreliable in the extreme.
--
| Dick Pierce |
| Professional Audio Development |
| 1-781/826-4953 Voice and FAX |
| [email protected] |

It's good to note that VAF designs their speakers using parameters that have been measured after the drivers have had a bit of run-in, but keep in mind, this run-in, break-in if you will, is nothing like what many people do which is to run their speakers for weeks at a time. As the Clark paper points out, its a pretty quick process if you will
Mark: yes, that's correct. which is in-line with the work performed by Clark, Nousaine, and others. The resonant frequency, Fs, decreased in all the studies that have been presented in this thread. If you buy a VAF kit or any speaker for that matter, you should expect due to shipment, storage conditions, etc.that the drivers will require a momentary realigning. Perhaps you can do some leg work and determine what the typical variation in the Thiele-Small parameters are for drivers and how they vary both within a manufactured lot and between manufacturing lots. This might be helpful for this thread don't you think?
Overall, it seems the break-in thing is much ado about not too much at all. My recommendations still stand. Listen to the speakers long enough to decide if you like them but not so long that you can't return them. If you're unsure about whether they're your cup of tea, take them back. Better the money in your pocket than someone elses.
By the way, for those who might be inclined to make their own speakers and are looking for sources of the felt that Dunlavey, VAF, etc use, you might want to contact the following companies:
Felt Fabricating Pty Ltd
Adelaide,
South Australia
Fax: ++8374-1463
or Link Removed

 

[Frank Zimkas]

Chu,
As usual a well thought out reply. My favorite part is



Thanks to all that have and will continue to contribute to this thread. I hope everyone has learned a thing or two about this subject, I know that I certainly have.

Thanks again to all!!

 

[Tom Brennan]

Larry B---We're talking about subtle differences between speakers of the same model not the gross differences between 2 completely different speakers. And subtle changes thought to be break-in related could certainly actually be because of differences within tolerances or positioning.

 

[Philip Hamm]

I don't know abotu speaker break-in but I know for sure that our brains "get used" to speakers.

 

[Mal P]

G'day Chu,
It's good to see someone from overseas familiar with the work of VAF Research Their top of the line speaker has a frequency response from 21hz to 20khz +/- 1.2dB, with a phase deviation of +/- 5 degrees. Whoever said good sound couldn't be measured.
I think that perhaps we're getting a little lost in the semantics however... I personally would consider the centering/settling of the drivers as break-in, especially since parameters change audibly once the drivers actually start moving. I do agree that there is most likely not much benefit of running in a speaker for hundreds of hours, but there could be exceptions, especially when it comes to ribbons.
For what it's worth, Philip Vafiadis is a strong believer that speakers do break-in, he has measured such details (as indicated by my last post) and he has stated that both him and his staff can pick their speakers "out of the box" as opposed to a well run-in speaker 100% of the time. I have no reason to doubt him, he has prioritised the importance of good measurements, and I respect that.
Cheers,
Mal

 

[Chu Gai]

Well Mal, this may really amount to quibbling over trivial things, hair-splitting if you will. If you've read any of the posts here, or on other forums, or even had the 'displeasure' to deal with salespeople, you wind up hearing about how LONGGGGGGGGGGGGGGGGGG it takes to break in a speaker. The work done by VAF and others suggests the break-in is really a pretty quick thing and really doesn't require this week upon week thing. Put them on for an hour or so or however long it takes to drive down to the pizzeria and order a pie, pick up some beer, kick back and decide if you like them or not. If the return period is 2 weeks, make your decision in 10 days. Sounds like enough time to weigh the pros and cons, no? Maybe its different in Australia as to the pizza...the beer I think is a universal given.
Now if you really want to read some funny break-in comments, head on over to odyssey's site and read what they theorize about amps. They seem to have a relatively reasonably priced amp considering they're doing a 20 year warranty, but their break-in theories...well just read them but be careful not to spill the beer

 

[aldamon]

Like I've said previously, if nothing else, this information is good for the consumer and can prevent one from being snowed by the 'occasional' misinformed or desparate salesperson.

Nice work Chu. Thanks for the posts.

 

[Michael R Price]

How about some tests of THD or IMD in the midrange and whether that changes over time? I would expect that if the suspension is stiff, that might cause some nonlinearities. My speakers' woofers (not tweeters) sounded very harsh and fatiguing at the beginning... not to mention having a lack of bass, which the above measurements can explain.

 

[Richard Greene]

This is such a silly subject to debate because the only way to prevent break-in is to never play your speakers.

From a general engineering point of view:

All drivers "break in".
Whether break-in is audible is the only question.

Long-term "break in" takes roughly 10 to 40 years until the materials have deteriorated to such an extent that the performance is bad news. Some foam surround drivers in the past didn't last 10 years but I had one 12" subwoofer driver that lasted 20 years.

Short-term break-in, if audible, seems to last less than ten minutes if the drivers are driven hard, in my experience with the drivers I've purchased (certainly only a very small sample of what's available). Over 95% of drivers I've used in DIY speakers since the 1960's have had no obvious sound quality changes after the first 10 minutes of use.

There are new materials used for spiders and surrounds these days (some huge surrounds in high xmax woofers were unknown decades ago) so you can only generalize based on specific drivers you've used.

If a speaker was going to sound awful when first played the manufacturer would most likely perform a break-in procedure before they were boxed -- let's call that "quality control" as you know real men don't read owners manuals so any warning about break-in would remain unread ... and a small percentage of the speakers would be returned by irate owners. This would never happen to women because they do read owners manuals.


I've been building speakers since the 1960's and have found break-in effects to be much more common these days -- especially with thick, stiff foam surrounds used in some recent high xmax subwoofer drivers I've purchased.
However the audible effect of break-in (driver's are audibly bass shy and seem to have considerably lower than rated efficiency at first) seems to last only minutes if they are driven very hard (to at least half of xmax) ... but I suppose break-in could take a lot longer listening to chamber music! Any changes after 10 minutes were not audible and that makes sense: if there were still audible changes after many hours of use (or even hundreds of hours, as some golden ears insist), how would a driver know when to stop breaking in ... before it broke down. It would be pretty unusual for an engineer to specify materials that kept changing that much in use.

In the Loudspeaker Design Cookbook Vance Dickason explains the main reason for driver break-in before measuring T/S specs is to make sure the driver has no defects before it's installed in the enclosure. He writes (agreeing with Tom Nousaine) that minor changes to the T/S specs are offsetting -- so they would have a small effect on the speaker enclosure. There may be some drivers in the world for which break in would have more than a small effect on enclosure design, but we have to generalize here. Often the changes to T/S specs attributed to "driver break-in" are really due to voice coil "warm up" rather than "break-in" ... and that would be repeated every time you turned on your system to play music through "cold" voice coils.

There is so much disagreement on the subject of break-in that I've become concerned that no one knows what they are talking about, including me ... or they are simply extrapolating experiences with a handfull of drivers they have experience with to all the speaker drivers available for sale in the entire world.
In fact, I am so concerned about the effects of break-in (who tells the drivers to stop breaking in ... before they break down?) that I am scared to use the new drivers I just purchased. So I will use them as paperweights at work.
Sorry it's impossible to take this subject seriously -- how about a debate on new shoes break-in theories?

Rich Greene
Court Jester & Harem Interviewer
(Formerly Village Idiot)