Believe it or not everybody is a believer. Some believe in god, other in science, some in experiment, other in themself.
Those who don't believe (say in god) will ask the god believer to prove themself. It's always the same.
Those who don't believed in speaker break in will argue that the believer simply get used to the sound.
When i break in my loudspeaker sometimes i use pink noise, sometimes music( most of the time in fact). I first listen to them when i begin (less than 30 minutes) than i go away for many hours. When i "believed" it's done, i listen again.
Talking just about the woofers, there's always more bass after than before. That's what my experience tell me.
p.s. i didn't study in engineering or even in electronics but i "believe" i'm a scientist!
Daniel
Those who don't believe (say in god) will ask the god believer to prove themself. It's always the same.
Those who don't believed in speaker break in will argue that the believer simply get used to the sound.
When i break in my loudspeaker sometimes i use pink noise, sometimes music( most of the time in fact). I first listen to them when i begin (less than 30 minutes) than i go away for many hours. When i "believed" it's done, i listen again.
Talking just about the woofers, there's always more bass after than before. That's what my experience tell me.
p.s. i didn't study in engineering or even in electronics but i "believe" i'm a scientist!
Daniel
This has been done to some level at this forum. I will quote what Pio2001 had to say about the results since I can't seem to link directly to the post:Ron E said:
Room ambient temp, or driver VC temp? I have experienced mild fluctations measuring drivers on different days, which is why I now try to do everything at the same time so that the relative measurements are more useful. But even then, although I haven't tested in extremes, I think that the fluctuations are small relative to the large changes shown by the breaking-in. Next time I would like to do such testing in a more temperature controlled environment...Ron E said:
Please keep us informed with new developments.
For the record, my words were _not_ directed at anyone in particular - least of all those who took offense.
--------------
Vikash,
The analysis that was carried out on the other board wasn't an uncertainty analysis, rather it seems it was a quick analysis of whether it is statistically probable that if a raise in Vas was detected 11 out of 12 times that it was random chance. That said, the general trend of your results agrees with many experiences.
An analysis of uncertainty in measurement would take far more time, and require knowledge of how the measurements were made. One needs to factor in what is being measured and how, considering propagation of error and estimates of uncertainty in each measurement. I once did an uncertainty analysis on the typical equations for solving for T/S parameters and found that accuracies better than 5-10% are probably not acheivable with typical setups.
BTW, I am not trying to say that your results fall within the scope of measurement uncertainty. I would say that it is pretty clear that Cms changed during your test. I merely found the variance in Mms and Bl interesting. I have observed changes in Cms in my tests as well, but they were more transient. In one case, a woofer after being left for several months, had a higher Fs than before..... (Note that I can't tell you the temperatures at which I took the measurements, either)
Regardess, those changes are not the changes I dispute. I have yet to see anything but anecdotal evidence of large tonal balance changes due to driver (or any typical speaker component) breakin. These claims I attribute to accomodation.
--------------
Vikash,
The analysis that was carried out on the other board wasn't an uncertainty analysis, rather it seems it was a quick analysis of whether it is statistically probable that if a raise in Vas was detected 11 out of 12 times that it was random chance. That said, the general trend of your results agrees with many experiences.
An analysis of uncertainty in measurement would take far more time, and require knowledge of how the measurements were made. One needs to factor in what is being measured and how, considering propagation of error and estimates of uncertainty in each measurement. I once did an uncertainty analysis on the typical equations for solving for T/S parameters and found that accuracies better than 5-10% are probably not acheivable with typical setups.
BTW, I am not trying to say that your results fall within the scope of measurement uncertainty. I would say that it is pretty clear that Cms changed during your test. I merely found the variance in Mms and Bl interesting. I have observed changes in Cms in my tests as well, but they were more transient. In one case, a woofer after being left for several months, had a higher Fs than before..... (Note that I can't tell you the temperatures at which I took the measurements, either)
Regardess, those changes are not the changes I dispute. I have yet to see anything but anecdotal evidence of large tonal balance changes due to driver (or any typical speaker component) breakin. These claims I attribute to accomodation.
Ron E said:
Ron E said:
I dispute the fact that since documentation does not exist to satisfy_you_ does not mean that the phenoma does not exist.
I have yet to see any documentation or anything proving that it does not exist.
All you are saying is that it has not been proved to _your_ satisfaction.
You so far have not proved that it does not exist or occur.
You obviously doubt it, but that is not proof that there is no such thing.
I attribute your claim that it does not exist to be behaviour that is common for someone questions everything.
big grin_
regards
Ken L
454Casull said:
Yes, may be you're right about the Cms. But they are exactly the same type and of the same material (I don't know about the inside). But they are from the same lot (ready stock item).
BTW, I got another feeling that there's something in the crossover that has something to do with break-in. It seems like the capacitor needs somekind of "charging" before fully functioning. But that's another issue, that I'm myself not really sure. Just an imagination.
But I have another similar experience...
I usually designed my speaker crossover using only half of the speaker (mono). When it was done I built the other channel, but never been 100% similar (because I need comparison). The right channel were so smooth and have better bass performance. I tried to match the left channel step by step until they are about the same, only the internal cable is the difference. The left channel used copper from Solen inductor. I couldn't believe that the cable had such a big effect, and I blamed the amplifier and the unequal placing of the speaker in the room. I thought there must be some defect in the left channel of the amp.
Until lately, I cranked up the volume and I heard as if something wrong was happened to the right speaker. I opened the driver and checked if there was something causing the resonance. I couldn't find anything but from the driver! I tried another CD, I tried another driver, and yes, it was the driver. So I pulled the left driver and checked, and yes, the right driver had defect that can be heard when driven hard (at bass).
I pushed the cones with my fingers, and yes, the broken one was more flexible...
Ken L said:
Argumentum ad hominem?
While you observe that I have not proven whether or not there are large spectral changes with break in, the same could be said for the proponents of "break-in theory"....
Subjective perception is a tricky thing. I have given examples of where the claims of people who claim to hear things have been proven wrong, yet confusingly they continued to believe them. People confidently trust their perceptions, yet it has been proven time and again that our perceptions are not as reliable as we might think.
Psychoacoustically, people tend to favor the louder of two systems, irrespective of tonal balance. In this case the level differences impair our ability to make a proper value judgement. In the post war era, hi-fi afficionados conducted an experiment and found that people preferred narrower bandwidth systems over the wide band hi-fi systems. This was completely opposite to what they were expecting, and upon further investigation, it was found that people preferred the narrower bandwidth systems over wider bandwidth ones only when the systems showed relatively high levels of distortion. Here an uncontrolled variable interfered with the experiment, leading the researchers to a spurious outcome.
Accomodation is something experienced by all of us, with all of our senses. It is a much simpler explanation. I will continue to accept the accomodation hypothesis as true until it is proven otherwise.
Look for an old Stereophile review of a Koetsu phono cartridge, and Koetsu's response, for a really good laugh about break-in. It is a dramatic example of how a person who has listened to more high end systems than most of us ever will can delude himself embarrassingly.
Of course there is a driver break-in period -- and althou it may level off, it continues until the driver dies... it is a physical device subject to mechanical stress...
I have heard many a time, same speaker, same driver lot, unbroken in speaker, next to broken in speaker. Clearly not the same. And a large enuff sample set that it wasn't a fluke.
It's like a pr of blue jeans. Stiff when you get them, needing some wear & washes to get them comfy.
dave
I have heard many a time, same speaker, same driver lot, unbroken in speaker, next to broken in speaker. Clearly not the same. And a large enuff sample set that it wasn't a fluke.
It's like a pr of blue jeans. Stiff when you get them, needing some wear & washes to get them comfy.
dave
Ron E said:
If you look at my earlier post you will see that some proof that it is not just accomodation has been produced. Unfortunately, the guy I did the work for does not want to get involved here at diyAudio, despite my nagging, (he worries about professional credibility problems!?), so I cannot get his results to post here.
However, I do have a lot of respect for the accomodation theory, and indeed, use it myself in a lot of debates about components, but IMHO we have enough evidence of other factors in loudspeakers to warrant more research...
I have done some tests on Peerless drivers for a local company, and the difference in their small signal impedance vs freq graph between new and worn drivers are minimal in the drivers I have looked at.
Due to the nature of the work and my contract with that company, I'm unable to say more.
\Jens
Due to the nature of the work and my contract with that company, I'm unable to say more.
\Jens
I was only looking at the impedance function at the time, so I did not do any time domain measurements - sorry.
The main difference, I noticed, between a new and a worn driver seemed to be the size of the impedance at resonance, because the rubber surround got a little looser after some time.
\Jens
The main difference, I noticed, between a new and a worn driver seemed to be the size of the impedance at resonance, because the rubber surround got a little looser after some time.
\Jens
Al , Jens,
Please take a look at the experiment carried out by Vikash. Are these results what you would expect to find based on your experiences?
Please take a look at the experiment carried out by Vikash. Are these results what you would expect to find based on your experiences?
Rodd
I was carrying out a similar run in to Vikash with my set of AP100s, the results after 50 hrs are here.
Unfortunately the testing got put to one side after I was burgled, I only just got all my test kit back from the police forensics lab on Friday, (unfortunately not my scope, that and the car DAB radio are the only losses). However, I will get back into action in a week or so.
As for the magnitude of the results, they are all much as I would have thought from previous testing, apart from Qes, I don't remember seeing much more than about 5% variation before.
I was carrying out a similar run in to Vikash with my set of AP100s, the results after 50 hrs are here.
Unfortunately the testing got put to one side after I was burgled, I only just got all my test kit back from the police forensics lab on Friday, (unfortunately not my scope, that and the car DAB radio are the only losses). However, I will get back into action in a week or so.
As for the magnitude of the results, they are all much as I would have thought from previous testing, apart from Qes, I don't remember seeing much more than about 5% variation before.
pinkmouse said:
Durn it. That's a $%^&& shame!
You have my sympathies on that, Pinkmouse.
I live in what used to be a small town 40 years ago We not only never locked our doors back then, my brother and I never end had keys!
Now I have monitored alarm systems at my home and business.
A sad fact of life.
Regards
Ken L
Well here is what I think,
To my knowledge the measured T/S parameters a greatly dependant on the current used when measuring the parameters. This means that you absolutely must use the same test conditions as the manufacturer in order to me able to compare results. Where I was trained, they use a different, but quite similar, model as the T/S. To obtain the parameters in the model you measure the impedance curve of the driver in a known volume cabinet using a known current, and “fit” the parameters to the measured curve.
The Le parameter is too me not very accurate, one of the reasons for this is that the “Le” parameter you measure is completely useless for calculating big drivers LF response, while at the same time the measured impedance and calculated impedance never correspond at higher frequencies. I believe the T/S model in fact disregards this parameter when calculating the impedance for a system.
In any case, the impedance increases has another slope in real life comparing to the T/S model. Some say with a factor of Le and some with Le^2. Or at least that is what I can remember right now.
As for the rest of the parameters, they easily have a variation of 5 – 10%.
So in order to determine what happens there is only one reference – the driver at time 0 – and not the factory specs. I can’t seem to find any information about the conditions where the T/S parameters are measured, but it’s it certain that the driver would need a long time to cool off, as the temperature increase during “burn in” in both electrical and mechanical parts need to be eliminated before measuring the parameters. In order to really get scientific a temperature controlled room and a measure of barometric pressure are also needed – no kidding!
\Jens
To my knowledge the measured T/S parameters a greatly dependant on the current used when measuring the parameters. This means that you absolutely must use the same test conditions as the manufacturer in order to me able to compare results. Where I was trained, they use a different, but quite similar, model as the T/S. To obtain the parameters in the model you measure the impedance curve of the driver in a known volume cabinet using a known current, and “fit” the parameters to the measured curve.
The Le parameter is too me not very accurate, one of the reasons for this is that the “Le” parameter you measure is completely useless for calculating big drivers LF response, while at the same time the measured impedance and calculated impedance never correspond at higher frequencies. I believe the T/S model in fact disregards this parameter when calculating the impedance for a system.
In any case, the impedance increases has another slope in real life comparing to the T/S model. Some say with a factor of Le and some with Le^2. Or at least that is what I can remember right now.
As for the rest of the parameters, they easily have a variation of 5 – 10%.
So in order to determine what happens there is only one reference – the driver at time 0 – and not the factory specs. I can’t seem to find any information about the conditions where the T/S parameters are measured, but it’s it certain that the driver would need a long time to cool off, as the temperature increase during “burn in” in both electrical and mechanical parts need to be eliminated before measuring the parameters. In order to really get scientific a temperature controlled room and a measure of barometric pressure are also needed – no kidding!
\Jens
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.