This is an area of debate so some will probably disagree with me...
I think much of this "driver break-in" may have to do with the ears and brain acclimatizing to the sound. Some drivers take 200 hrs and some 1000 hrs. All drivers that I have tested, except for A7.3 and A7P were fresh out of the box and they seem to perform just fine. They are what they are. Sure suspensions loosen up, but that impacts the lower octaves for bass extension, not midrange and HF's. If it squawks, I am not sure that goes away with time other than your ears are not as bothered with more time.
You'd want to measure them first to make sure that consistency is good enough so that driver to driver issues aren't going to have an effect on the sound. I would also encourage the comparison to be made against the driver with thousands of hours on it vs a driver that has otherwise been run in for a few hours in the correct way to loosen its suspensions and to let the soft parts settle/reach equilibrium.
Considering we're dealing with mechanical things it seems obvious to me that stuff will change with use. Both due to the flexing of the suspension items and vibration of the cone. What the sonic changes are will depend on materials etc...
In my cold workshop (around 12degrees C) the P830983 has a 7-8 dB peak centered around 12K which gradually disappears as the temperature of the surround rises, I have noticed similar frequency response fluctuations due to temperature changes in many different speakers such as Vifa P13/17/21 drivers, some Wharfdales and even a JBL 6" monitor driver which had some sort of PVC type surround which at low temperatures stopped the cone from doing any bass whatsoever, so I would not discount the effect of any temperature variations on the surrounds of drivers that are composed of a material that changes compliance and dampening characteristics due to temperature variations.
X,
Here are links to burst files that I used for IMD testing of TC9FD:
https://www.dropbox.com/s/mm9sjboc9ncihsi/BP200-350 and walking bursts.wav?dl=0
https://www.dropbox.com/s/j2ec779em0nbcz0/LP350 and walking bursts.wav?dl=0
Bursts are at 1 second intervals. The LP350 sequence shows off severe limitations of full range drivers.
Start at low volume, and repeat tracks with increasing volume; audible distortion will become apparent long before mechanical or thermal damage is anywhere close to happening.
--------------------------------------------------
Method of stereo from mono speaker:
Stereo sound clip of 30-45 seconds may be rearranged into sequential mono track of left and right channels. This is played back through test speaker and recorded. Then based on known sample length a stereo track is created that is used for demo.
Here are links to burst files that I used for IMD testing of TC9FD:
https://www.dropbox.com/s/mm9sjboc9ncihsi/BP200-350 and walking bursts.wav?dl=0
https://www.dropbox.com/s/j2ec779em0nbcz0/LP350 and walking bursts.wav?dl=0
Bursts are at 1 second intervals. The LP350 sequence shows off severe limitations of full range drivers.
Start at low volume, and repeat tracks with increasing volume; audible distortion will become apparent long before mechanical or thermal damage is anywhere close to happening.
--------------------------------------------------
Method of stereo from mono speaker:
Stereo sound clip of 30-45 seconds may be rearranged into sequential mono track of left and right channels. This is played back through test speaker and recorded. Then based on known sample length a stereo track is created that is used for demo.
Well said, Mondo. When the advice on break-in time for A7P's went up several times, I inquired for a procedure to speed up the process. Due to the reactions I got, the feeling crept on me that I had asked a high priest and his followers for a condensed version of a sacred ritual. Of course, nothing useful came out of it. Although this is a faux pas in a church, it shouldn't be one over here.
After the manufacturer recommended break-in period times three and then some, I can confirm that the character of the A7P had not changed appreciably when compared to the before break-in character. The impression they left on me after break-in was the same as the impression they gave before break-in, and the measurements backed up that impression. As I remember, the only difference was at the onset of the 16 kHz resonance. So any change is at most subtle, but the overall character of the driver does not change. I did the break-in out of ear's reach, so there was no opportunity to get used to their sound.
Although a single case is not enough to base a sweeping statement on, I have more than just the A7P's on my shelf, and I think I can safely say that the out-of-the box performance of a driver is sufficiently representative for its performance after some use, to safely base a verdict on. Vance Dickason's comments apply here: break-in before evaluation mostly serves to cull out any possible duds before the test.
@X, I think it is easy to underestimate the significance of your work here. With more and more stuff coming available to the average DIYer, while at the same time the amount of brick-and-mortar shops that carry a decent selection is dwindling, there needs to be some other way to evaluate drivers before a purchase is made. It would be wonderful if a couple of large DIY suppliers would pick this up, and provide a form of "remote auditioning" for the stuff they sell. Preferably in a (standardized) way that has support from a large part of the DIY community, and even more preferably, strongly based (and in collaboration with) the pioneering being done here.
The Thiele/Small Qts parameter only describes the behaviour of the driver around and below it's fundamental mechanical resonance. The frequency content of the overshoot and ringing due to Qts is only around the mechanical resonance of the driver. If you place a high pass sufficiently above the drivers mechanical resonance those effects will disappear. A lot of sealed midrange and tweeter drivers will have a Qts above 1 because the size of the rear chamber is dictated by what works best for higher frequencies, not what works best around Fs where the driver isn't intended to be used.
To evaluate the linear response of a driver above Fs you need to consult a frequency response or cumulative spectral decay plot. Ridges in the CSD plot (from cone resonances) translate into ringing in the impulse response.
Last edited:
Usually, when an audio sounds good in a few seconds of listening, it could sound worse/horrible after a few hours or days. If it sounds horrible in a few seconds of listening, it will sound horrible "forever".
But IME there is this "breakin" phenomenon with drivers. This is related with surround/spider/cone suspension mechanism that will affect the most critical cause of non-linear distortion, which is a situation where the coil rubs the pole piece or its surrounding due to non-centered pistonic movement. This cannot be seen from FR measurement.
Rubber surround doesn't need "breakin" as do cloth surrounds, but spiders are mostly made of cloth.
I modify drivers a lot. For better flux (and no cavity resonance), the gap should be small, but then to make the coil centered (so the movement doesn't rub the pole or cyllinder) is a difficult job. The magnet should also be centered, which doesn't always happen even with expensive drivers.
First, do not confuse system (driver+box) Q and driver Q. Second, what you hear may have no direct correlation with Q but other parameters or simply IMPLEMENTATION.
Once a driver is in a box, driver's Q is almost irrelevant.
System's Q is decided by a lot of factors, mostly target box size and target lowest bass extension. Driver's Q will affect this decision.
If we ignore box size or frequency, actually lower Q driver is the most dynamics. Driver's Q (Qtc) has inverse relation with Qms. Lower Q driver will usually have higher Qms (and strong magnet), meaning a better suspension and can respond very quickly to signal input. This is DYNAMICS!
Low Q woofers are difficult choice if you expect a bookshelf with low frequency extension. But for a midrange in a multi-way, if you know what to do, it's very yummy. My midrange has Qts=0.26 and I don't believe a 10F can sound like it. Very dynamic and impactfull, no boxy sound at all. Even the box disappears.
If the Qts is lower than that, I will have no option but building a big horn for it.
High Q system will sound "boomy" because there's peak before roll-off. But whatever the Q of the drivers, (in general) you can choose to have a boomy response or not. Box size and low frequency extension is the limit.
Last edited:
BW,
Thanks for the test sound clips. These are not meant for human enjoyment but rather a mini torture test like water-boarding a driver - make it feel like it's gonna burst at the voice coil seams but no permanent harm done.
This will require me to get my big guns amp out (200w IRS2092 with +/-54v power supply).
As I had only one driver to work with for the A7.3, I had no remaining driver left to swap and compare. All I had instead of two drivers to compare before and after were measurements before, during, and after. Neither the measurements, not my ears detected a significant difference at 0, 40, 80, or 200hrs. But that's just me and one example of a silver coned A7.3.
It would indeed be great if manufacturers would have a standardized way to make sound clips for auditioning. Sort of like an IEC baffle standard or an AES power rating standard. But doing this first hand I see the labor involved and the difficulty in getting people to agree on a standard. Maybe it should be a 15 second long clip of pink noise played at 2.83v and 1m while mounted in same IEC baffle used for frequency measurement. Then a series of swept square waves from 20Hz to 20kHz, then a series of impulsive sounds a kick drum, tom tom, rim shots, high hats, fortissimo piano key strikes, plucked stand up bass, a wooden clack, maybe some wire brush drum or snare drums. It should be an industry standard IEC test sounds CD - so everyone always uses the same.
>>> It would be wonderful if a couple of large DIY suppliers would pick this up, and provide a form of "remote auditioning" for the stuff they sell. Preferably in a (standardized) way...
I can see manufacturers doing this for the DIY community. For example Fostex can provide sound clips of their drivers under the same exact conditions for us to compare. That would be nice.
I can see manufacturers doing this for the DIY community. For example Fostex can provide sound clips of their drivers under the same exact conditions for us to compare. That would be nice.
Nice idea, but the problem is that most if not all manufacturers would want to present their drivers in the most favorable light, and a standardized format might not be flattering to many drivers (and so might not be the best marketing strategy for some drivers).
Anyone with a financial stake in getting drivers sold will not want to provide this or even think it a good idea for the obvious reason that it may thwart potential sales based on anecdotal recommendations and an aura of mystery and the conviction of belief, both known to be powerful sales catalysts.
It will have to be someone independent from financial gains related to selling the drivers under test to give an unbiased treatment.
Last edited:
godzilla, would you mind sharing your personal favorites that you use or enjoy? I am also really curious of some of your "closet" drivers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.