Driver 'break-in' period

[roddyama]

If you challenge these audio mystics' beliefs, you are mocked (even by a "so-called" moderator - I suggest you look up the term).

Skeptically yours
Ron

Ron,

My apologies for the pic. Yes, I was mocking you, but it was done so in fun. I hope you can accept this apology as it wasn't meant in offence.

Sincerely Your,
The So-Called Moderator,

 

[eStatic]

As far as Estatic and Pjotr's comments go, read Dick Pierce's words. I am not denying the suspension compliance changes. I am denying the tonal balance changes that people claim for broken in drivers, cables, amplifiers, whatever.... There is nothing but anecdotal evidence for this. If you challenge these audio mystics' beliefs, you are Skeptically yours
Ron

Ummm... Ron E, not sure why I was included at the head of that paragraph. As far as I can tell, I can't hear what can't be measured. In fact I can't hear a lot of what can be measured. I don't think I've ever suggested otherwise.

 

[Ron E]

Ummm... Ron E, not sure why I was included at the head of that paragraph.

It should have been two paragraphs. My inclusion of you and Pjotr had to do with the exchange between you and roddyama in which it appeared you had not read the link I posted to Dick Pierce's text, or perhaps even the first few posts - as you were "otherwised" by information that had already been posted.

In case you are not familiar with him, Dick Pierce is an audio consultant with considerable experience who spends a lot of time on the rec.audio.* forums. He is known for his rather direct manner, but in most cases you cannot fault his knowledge or technical ability.


Changing gears...

characteristics like impact and dynamics improved dramatically.

A statement such as yours sometimes follows from a fundamental misunderstanding of physics. It seems intuitive that a loosening of suspension causes an increase in sensitivity, but this is not so. The reduced Fs reduces the sensitivity by the same amount that the increased Cms increased it... This interpretation of what you have said is based on my interpretation of the words "dynamics" and "impact" as SPL.

 

[roddyama]

A statement such as yours sometimes follows from a fundamental misunderstanding of physics. It seems intuitive that a loosening of suspension causes an increase in sensitivity, but this is not so. The reduced Fs reduces the sensitivity by the same amount that the increased Cms increased it... This interpretation of what you have said is based on my interpretation of the words "dynamics" and "impact" as SPL.

Your take on someone’s fundamental misunderstandings are about as thorough as what the measurements say about the sound of a loudspeaker. The behavior of a real loudspeaker goes far beyond the calculation of fs or Cms. Some of the aspects of the workings of a loudspeaker that tend to defy the usual mathematical analysis are:

- The nonlinearity of the energy transduction from electrical to magnetic to mechanical to acoustical and finally to aural.
- The nonlinearity of the above forms during transient reproduction or for that matter, high-level steady state reproduction.
- The less then ideal behavior of the cone/diaphragm under various transient and non-transient type reproduction.
- The result of something as simple as the concentricity of (or lack of) the voicecoil in the gap.

The list does indeed go on. Small did not deal with these aspects of the loudspeakers mathematically because he dealt with the small signal, steady state model of an ideal loudspeaker and his analysis were based on those assumptions. Even though he did cover some aspects of large signal behavior, the basic model was not used in this case except where he extrapolated under the assumption the small signal model would hold or at most have an insignificant affect if it didn’t. Even at that he only covered general topics like efficiency or power rating.

The suspension of a loudspeaker is anything but linear. The value of Cms is merely a number calculated from measurement of other phenomenon made under very controlled and limited conditions. There is not much there to form a basis for assumptions on the loudspeakers sound or the phyical changes of its suspension in the real world, under real world conditions.

Once you began to deal with matters in the real world you’ll find that the majority of the solutions cannot be found in the textbook. That the idea of a mathematical explanation for every physical phenomenon is just plain teenaged gullibility. It’s the perception that comes first, than the idea. The math is a distant third or forth.

 

[Ron E]

Once you began to deal with matters in the real world you’ll find that the majority of the solutions cannot be found in the textbook. That the idea of a mathematical explanation for every physical phenomenon is just plain teenaged gullibility. It’s the perception that comes first, than the idea. The math is a distant third or forth.

Fourth?

While linear models do not predict 100% of the behavior of loudspeakers, they illustrate general trends better than a more complex model would.

Is there any reason to believe that after a time of exercise a driver operated in a range far above the region where Cms dominates will exhibit changes that make it more or less linear in the short term - in your case weeks or months? Show me some evidence, Rodd. I am confident that if you were able to perceive a difference in sound, you would be able to see it in measurements, whether it be Frequency response, distortion tests, Power Compression, etc...

Hmmm. Isn't there a relation between transient and steady state...something like FFT?

The one place I see room for slight changes in response is the surround termination of cone resonances, although I would expect this to be minimal (I certainly haven't seen it in my measurements or anyone else's) and actually rather difficult to measure under real-world circumstances, unless the measurement setup were not changed in any way between measurements....due to directivity concerns and complex radiation characteristics in the non-pistonic range as well as the effects of temperature and humidity.

That said, there is significantly more uncertainty in human perception. The acclimatization effect is real and there is no getting around it. If you had stated that you compared a source to a stable reference and noticed changes after run-in, I might be more inclined to add that data to my own understanding. Even if there were measurements showing differences, one would need to be skeptical, as there could be a simple explanation, like temperature....

Skepticism is healthy. Belief without proof does not further understanding....

 

[roddyama]

Fourth?

Yeah, I always hated the math.

While linear models do not predict 100% of the behavior of loudspeakers, they illustrate general trends better than a more complex model would.

Linear models are good for design and R&D, not as a basis for real world performance testing. We haven't quite reached that level of technology yet (although in some less complicated systems, we're getting close).

Is there any reason to believe that after a time of exercise a driver operated in a range far above the region where Cms dominates will exhibit changes that make it more or less linear in the short term - in your case weeks or months? Show me some evidence, Rodd. I am confident that if you were able to perceive a difference in sound, you would be able to see it in measurements, whether it be Frequency response, distortion tests, Power Compression, etc...

I don't have the measurements, and I sure that no one else does (or they do and their not making it known). I would be very interested to see the distortion numbers of a driver right off the assembly line vs a driver that has been "broken-in". But even that is limited to single tone THD and 2 tone IMD. Not necessarily representitive of the real world. BTW, I did hear the difference and I think I would see that in the appropriate measurements.

Hmmm. Isn't there a relation between transient and steady state...something like FFT?

Excellent. Yes, but when was the last time you seen a FFT of a real world signal used to test a speakers performance. This is what is needed. A signature analysis of the FFT of a real signal at real levels compared to that of the speakers output (or at least to the movement of the voicecoil). Than you can realistically assess the affects of "breakin". Signature analysis is used extensively in the NVH world.

The one place I see room for slight changes in response is the surround termination of cone resonances, although I would expect this to be minimal (I certainly haven't seen it in my measurements or anyone else's) and actually rather difficult to measure under real-world circumstances, unless the measurement setup were not changed in any way between measurements....due to directivity concerns and complex radiation characteristics in the non-pistonic range as well as the effects of temperature and humidity.

Exactly, just some of the variable that make these type of measurement less than accurate and ineffective.

That said, there is significantly more uncertainty in human perception. The acclimatization effect is real and there is no getting around it. If you had stated that you compared a source to a stable reference and noticed changes after run-in, I might be more inclined to add that data to my own understanding. Even if there were measurements showing differences, one would need to be skeptical, as there could be a simple explanation, like temperature....

The problem with conclusions about perceptions is that they are based on the very perceptions being evaluated. To accept the validity of one without the other is itself invalid. To accept them both is to accept the idea that the perception can be the basis for a conclusion with or without the analysis.

Skepticism is healthy. Belief without proof does not further understanding....

Belief in some basic assumptions is the prerequisite of understanding.

 

[fireman]

I do the same with new speaker units (bass and mids) but use a sine generator tuned to Fs of the speaker. You will be amazed how little power is needed to get excursion to X_max. After a few hours Fs clearly drop with most speakers. It is mainly because the spider becomes more flexible. I usually run in speaker units for 48 hours at Fs and X_max this way before measuring T/S and designing them in.

I have no problems using this method with woofers but what about tweeters ? First, do they also need to be broken in ? If so, Fs is not a problem but X_max is !!! What would be a good way to determine the appropriate power to use with the tweeters ?

 

[Ken L]

Driver break in is quite real

And easily heard.

Some drivers more so than others.

FWIW, it took my FE206E's somewhere around 300 hours to break in.

Definitely, _not_ something I would have gotten used to or would have wanted to get used to.

To a large extent, "driver break in" threads usually degenerate into the same garbage that "can hear the difference in silver cables" threads seem to _big grin_

Y'all have at it.

Regards, Ken L

 

[Ron E]

The "L" in Ken L must stand for Lazarus - the thread that came back from the dead......

 

[fireman]

The "L" in Ken L must stand for Lazarus - the thread that came back from the dead......

I revived this thread as I wanted to know if and how to break tweeters in. So far, none of you have addressed this issue.

Flame suit on

I don't want to start a war but I have notices that too many times, these threads become personnal. It's too bad because there is a ton of knowledge in this forum. If people would restrain from personnal comments/dogma on technical questions, maybe we would spare ourselves a lot of grief.

Flame suit off

We are now returning to our regular programming

Do tweeters need to be broken in ? If so, if I use a sine wave generator tuned to the Fs of the drivers, what would be a good way to determine the appropriate power to use since there is no X_max for these drivers ?

 

[Pjotr]

Hi,

Can report my experience on a pair of Tannoy dual concentrics. New from the box they sounded a bit “blocked” or “dull” and a bit disappointing. After a few weeks service they came much more to life, especially the tweeters.

When you want to run in tweeters artificially I wouldn’t feed them more than 1W to avoid excessive heating. I would suggest pink noise rather than pure sines and fed through the appropriate x-over. Alternatively you can use an appropriate .wav processing program to make pink noise and high-pass filter it as well digitally.

Cheers

 

[jesper]

Just to add on to the list, I recently bought a pair of Fostex FE87E. When I first hooked them up (in a 3l sealed box) they sounded horrible. Little treble, very boxy and "hard" midrange. Now, after a week they are starting to shine, detailed and smooth. It's a VERY noticeable change.

/Jesper

 

[markp]

Tweeters dont have the same suspension as woofers do and as such dont seem to need the break-in. There is no spider on a tweeter and that is where most of the break-in seem to happen.

 

[Pjotr]

There is no spider on a tweeter and that is where most of the break-in seem to happen.

No, there is no spider. But the surround of most tweeters is made from fabric (usually the same as the dome is made of). Anyway the break in of the Tannoy tweeters was quite noticeable. Maybe not all tweeters do need break-in (like pure aluminium ribbons) but with some it is noticable.

Cheers

 

[Ron E]

I have always found it interesting that the more expensive a product, the more likely it is that someone will say it needs break in.

This is a tactic used by Hi-end audio dealers when dealing with people having buyers remorse. The more expensive a speaker, the more it appeals to one's sense of fashion or pride, and people are more than willing to "give it a chance", not realizing that they are just getting used to it...

I especially enjoy the claims that solid state electronics and cables need break-in

"There's a sucker born every minute" - P.T. Barnum

 

[Ken L]

North Creek Music

George Short, of North Creek Music believes in such a thing as break-in of drivers.

He sells "perfect Pair" drivers matched to within .5 DB.

"For perfect pair matching, tweeters are broken in with 24 hours of pink noise, while woofers receive this along with 24 hours of a 25 Hz 1/3 octave warble tone. Perfect pairs are matched to within ± 0.5dB and provided with anechoic 2pi frequency response and free air impedance curves. Woofers are also provided with T/S parameters"

He has an excellent professional resume and that is a well known and respected company.

Further interesting info on his site at

http://www.northcreekmusic.com/Drivers.html

Regards

Ken L

PS the L stands for "Lucky" _grin_

 

[Ron E]

OK, North Creek says so, so it must be true. Maybe he can post measurements showing the changes pre and post break-in

Remember, North Creek makes money appealing to specific class of tweekazoid DIYers - mainly non-technical folks who don't really want to DIY, but just want a sense of involvement.

 

[Ken L]

Re: Driver break in is quite real

... Maybe he can post measurements showing the changes pre and post break-in

Are you kidding? Why would anybody ask him to take the time to do that?

The man is running a business, and a successful one at that. He could care less about this thread. _grin_

To a large extent, "driver break in" threads usually degenerate into the same garbage that "can hear the difference in silver cables" threads seem to _big grin_

You have a nice day _big grin_

Regards

Ken L

 

[pinkmouse]

Just a quick point, but a few years ago whilst working with a PA speaker designer, we did waterfall tests on some Precision Devices 15" drivers, from new to 500 Hrs, and HF breakup definately reduced with run in drivers.

They were run as a pair with opposite phase, sitting in 150l approx monitor wedges, running out of phase with 50w pinknoise and a blanket thrown over them. Testing was done every 50hrs approx, (Every other day ).

 

[markp]

Just a quick point, but a few years ago whilst working with a PA speaker designer, we did waterfall tests on some Precision Devices 15" drivers, from new to 500 Hrs, and HF breakup definately reduced with run in drivers.

They were run as a pair with opposite phase, sitting in 150l approx monitor wedges, running out of phase with 50w pinknoise and a blanket thrown over them. Testing was done every 50hrs approx, (Every other day ).

Well, we now have actual data proving that a driver can be broken in. Thank you Pinkmouse! Now maybe the naysayers will keep quiet!