Real world use of woofer Xmax

I haven´t found anything of substance to this topic. Obviously, different manufacturers set their speakers differently, different design choices are made for different reasons, but it would be nice to have some words about that.

For example, repeatedly I have found that RCF drivers work great at, or even past Xmax. Haven´t seen an issue yet. Though looking at the magnetic gap, I am worried how can they sustain the centering at these excursions. Feels ridiculous by looking at it with "common sense" eye.
On the other hand, although okay by design, I saw with many B&C speakers, that they have stiffer and steeper onset of suspension breaking at Xvar (B&C speakers Xmax marking), and the speaker soon stops responding to more stimuli in greater excursions. Also more power is burnt on suspension at Xmax/Xvar than usually. What would be the reasoning behind this? How it translates to real world usage? (Well, I burned my 21DS115 a bit, working it to Xvar, but there might be more to it). I guess that compared to RCF, additional ~120-150Watts in power compression, does not going to change that much if the speaker is going to burn or not. or would it?

How about long term longetivity of the suspension at, or past Xmax? Do rentals and fully exploited installs/rigs take recones as consumables to be replaced within 3-5 years of use?

LF18X451_Excursion.jpg


Working at 12mm with negligible loss is awesome to see.
Not planning to go past 15mm, but it is nice to see that the driver handles even 3dB momentary overload with grace. No signs of rubbing or too stressed noises, which in turn can be dangerous, because if the speaker does not protest either with noise, sound, smell or anything, then one cannot do anything about it dying.
 
Some companies will design the suspension as an excursion limiting device to protect the driver, usually so it can be used safely in multiple alignment types. Example would be putting one in a ported enclosure without a high pass, the sub will be able to survive being played below tuning without eating itself alive.
When the suspension isn't used as a limiter/protection, xmax is usually just determined by the form factor of the subwoofer and to fit in with specific alignments, usage, and power ranges. Like, modeling a scenario and alignment, the model will show that a sub might only be moving +/- 10mm so the designer will design the motor with around 15-20% margin to have 12-13mm. It's just a big bowl of design compromises to spoon through.
 
@hurrication:
Indeed. I suspected that. It is just that noone specs this on their product. Hard pick then it is, for personal purposes, or for the case where I need to fully exploit the driver capabilities. B&C speakers produces LSI data, where this is somewhat obvious, one can count the force to overcome the suspension at X, and thus the power burnt, but other manufacturers might not provide such data...

@diyuser2010:
That´s great info! I missed that one. And that´s quite some longetivity. Seems I´m set for 10+years then, by which I replace the design anyways. basically "indestructible" when there is no additional screw-up.
 
Last edited:
B&C rep Bennett Prescott has said that 1000 hours at Xmax is considered their useful life in one of the videos posted.
Not sure what video you are talking about, but in the one below Bennett says one of their main considerations is the speaker will behave well for 1000 hours of hard use, not that's considered it's useful life.
I've owned many loudspeakers that are not as well built as B&C that have been driven past Xmax for 10 years, probably 4000 hours of use without needing recone. I've torn up the same loudspeakers in seconds when run past Xmech..

At 32.00:
"And there's a lot of long term trade offs as well made, where consider suspension aging. So you want the speaker to behave well, and you want it to behave well for 1000 hours, which is not unreasonable, 1000 hours of hard use.
And you can do that, but again you may need to pay some price in terms of dollars, or in terms of how difficult it is to assemble, or what other numbers you can get, that maybe you can't get quite the X that you want, you want 16 millimeters but maybe you can only get to 14 millimeters with the materials that you know will last 1000 hours.
Somebody else may offer something that says, hey we'll do 16,18 millimeters, but if you run it for 1000 hours it's going to come apart.
And that's kind of one of our main considerations, is can we make this loudspeaker durable, can we make it consistently for the next 20 years. We've been making the 18TBX100 in it's most modern incarnation for almost 20 years. That's not at all unusual, we have lots of products we've been making for that long, and that's a big part of consistent manufacturing, that's kind of hard to put in a spec sheet, because there's no spec sheet number for that."

 
  • Like
Reactions: Crashpc
The real world use of Xmax is to allow people to calculate the maximum SPL as you approach the low frequency limit of a driver at some nominal level of distortion ( often 10% ). Plug the maximum power and Xmax along with the TS parameters into WinISD and plot the maximum SPL. At some low frequency the SPL limit takes a sharp turn downward when Xmax limits the output rather than the power. It works for any driver: woofers, midrange and tweeters.

The formula isn't too bad
"
"For anyone who wants to know and produce their own little spreadsheet here is the formula for Maximum attainable half-space SPL for a direct radiator driver for any given frequency, this is true whether it is an 18" woofer or a 1" dome tweeter:
SPL= 20 * log(10) (1.18 / 0.00002 * Sd * Xmax / SQRT(2) * 2 * PI * F^2)
Where Sd is in Meters^2
Xmax is in Meters one-way linear travel
And F is the frequency you want to solve for.
Jeff B."
 

Attachments

It is hard to tear up modern PA speakers past Xmech, because the coil is usually completely out the magnetic gap at that point. Not saying impossible, but very ill handling of the speaker has to be done in order to possibly witness immediate damage.

1000 hours or even multiples of that for working the speaker at or past Xmax looks generous to me. Can't imagine going that way for more than an hour before I check on the speakers when I think something is not 100% is right. Well, not everyone can do that, but that was not really my concern, to leave these overstressed for prolonged period of time.

The fact that the speaker only started somewhat mildly protesting at over +3dB past Xmax is very calming. Not gonna happen and speaker handling that gracefully all at the same time is reassuring. The B&C did not protest at all though, it started to smell. Thanks god for that too.

Well, RCF now does these 18mm Xmax speakers, and I am sure they will go past 20mm with EASE. The LF21N551 did, even with shorter coil.

I build my designs excursion limited, because it is easier to check and not immediately so dangerous as testing the power limit. Don't like paying for recones every second gig when the DJ is left unattended and usually likes to kill people with sound with the rig that is not supposed to do that. Even precautions like gains at max and DSP locked, they can find a way to eff things up somehow.
 
^ you gotta take next step in limiter design: microcontroller monitoring SPL, and when limit is reached a spray gun starts to water the DJ. If it still continues as fun activity, replace feed for the DJ with some multitone signal, 5 tones per octave for the whole bandwidth ought to do it. Perhaps better mix in monologue that reminds about cost of continued activity, health of the audience and DJ's wallet.
 
It is hard to tear up modern PA speakers past Xmech, because the coil is usually completely out the magnetic gap at that point. Not saying impossible, but very ill handling of the speaker has to be done in order to possibly witness immediate damage.

I folded the cone of a Beyma 15P1200Nd when testing in a prototype 40Hz ported box. Forgot the HPF, and forgot that REW starts its sweeps at a lower frequency than the selected "start" frequency.
500w of <20Hz (IIRC) resulted in seriously-impressive excursion, and when the sweep finished, there were six folds radially on the cone. I think they lined up with where the basket spokes are, but can't remember for sure.



On-topic, but I haven't really seen it discussed: if we have a 35Hz and 70Hz tone at equal level, the amplifier will clip when the level of each signal is at -6dB.
If we design a speaker so that it reaches Xmax with a 35Hz sine tone at full power, when "real-world" signals are applied, it will almost certainly never reach Xmax when the amplifier is at maximum output. The 35Hz + 70Hz example would reach 0.75*Xmax, which means there's another 2.5dB of mechanically-limited output available.

I'm aware that "real world signals" vary, but I'm yet to see anything with pure sine tones at full level. Even with something like this (caution: bad language from the start, but the bass is fun)

The bass tone isn't pure sine wave, and there's the kick on top.

Of course, as you use a bigger and bigger amplifier to make use of the excursion-limited output available, heat will become a problem. Clever limiting required.


Chris
 
Not sure what video you are talking about, but in the one below Bennett says one of their main considerations is the speaker will behave well for 1000 hours of hard use, not that's considered it's useful life.

My mistake, I was conflating a video with an email from Bennett. Here's a snippet:

"First, in heavy use we expect a transducer to survive about 1000 hours which means running a test 24/7 for six weeks, and second because the real aging is completely dependent on application - both enclosure, signal, and processing."

I don't mean to imply that magically at 1001 hours it dies, or becomes unusable, but 1000 hours is their threshold to where they ( B&C ) feel it's done a lot of work. Obviously, your mileage may vary. Temperature, humidity, excursion and other factors are at work here.

Klippel has a paper on suspensions that may be handy to the OP and others:

https://www.klippel.de/fileadmin/kl...s/Aging of loudspeaker suspension_Klippel.pdf
 
  • Like
Reactions: Crashpc
Not sure what video you are talking about, but in the one below Bennett says one of their main considerations is the speaker will behave well for 1000 hours of hard use, not that's considered it's useful life.
I've owned many loudspeakers that are not as well built as B&C that have been driven past Xmax for 10 years, probably 4000 hours of use without needing recone. I've torn up the same loudspeakers in seconds when run past Xmech..

I have to completely agree! But it depends strongly on what type of enclosure you are using, a FLH poses a 5-20x higher mechanical stress on the driver. Many FLH drivers die because of the high cmpression (equals stress on the membrane, they just tear or fold). The question is not how much the driver can endure, it's the question on how you use it. In PA, you never use the speakers above 80% of their limits and the reward is a 5-20x longer lifetime,, no matter what enclosure you are using. Sometimes even more if you not only look at the mechanical but also the electrical limits. Yes, you can easily kill 2kW drivers with just 500W if they can't do enough excoursion, the ends of the VC can't dissipate the heat and simply burn to a crisp.

Aside from that, you have to have a very close look on how they define the Xmax. If they add 1/4 or even 1/3 of the mathematical Xmax, you likely have to reduce your calculations by the equal ammount or accept your headroom is only for a much shorter time.
 
Well, the loss of 20% performance will have impact on dimensioning, rigging, logistics, crew and such can be found not so feasible, when fair lifetime before renewal investment is 5 years, by which many decide just to replace the system anyways. Also on a driver with true Xmax of 9,5mm, going gracefully to 18mm does not make sense to leave that much performance on the table. Mind you, the sinusoidal peaks are not as peaky and steep, and so the 1,414 ratio migh be good to apply to the raw values, by which we and up at excursion of 13,5mm, which is right near 14mm Xmax specified by the manufacturer hehe.
The pole pieces are designed in such things in mind, otherwise it wouldn´t make sense to go past 12mm in most drivers. No need for 15mm, 18mm or even thicker pole pieces. That thinner pople piece would gain you Xmax for free immediately. Even less material. But it is not so... Two drivers with mechanical Xmax 10mm will behave very differently with pole piece 8mm thick or 18mm thick.
 
Also on a driver with true Xmax of 9,5mm, going gracefully to 18mm does not make sense to leave that much performance on the table.

You don't. That's the headroom of your system.

E: Don't forget, at 100% your system sounds also a lot worse! If you need to go for 100% you don't put enough equipment on the location - and you are paying with sound quality and having to replace the drivers a lot sooner.
 
Last edited:
Yes. This is usage specific. My situation is, that I do not want to lug around with unnecessary gear, and I use it fully about once in three months for one gig, otherwise it goes anywhere between 1% and 30%. It is that "special time" when the speaker needs to be exploited fully, and I think I can demand that from high end product. It is unspoken "obvious thing" that I will not run the system knowingly past specified Xmax and I will not run the system into hearable distortion. That´s given. The additional 2dB I can momentarily extract from the driver would cost $$$$ otherwise. Now I have tried that with RCF LF21N551, and it worked exactly the way I wanted. I was blown away with what it could do. It would be IT. Yet I was afraid to push it that extremely, so I replaced it by double amount of LF18X451 to be on the safe side. Don´t think that I am really extracting the last halfdecibel from it. There is some sense and method to the madness. And reasonability too. In my lifetime I ruined zero speakers to date by input signal. And I hope it will stay that way.
Only smoked the 21DS115 a bit, finding how the suspension breaks the cone "too soon" for my design. RCF seems to not cause trouble in my approach. That needed to be discovered too on the way to successful design.
 
Your calculation misses a vital point, you ignore the power compression. Above 80% input you get ~2,5dB more spl per 3dB power increase and at 100% input power you get ~1,8dB more output per 3dB input power increase. So, no, your calculation doesn't add up in the least, power compression is a thing and no matter what amazing drivers you put there, they have power compression too.
 
Your calculation misses a vital point
Wellll? Does it?
you ignore the power compression.
That´s where my design comes in. Of course any system suffers power compression including mine, but I did wok on that five ways.

1) Only impedance peak is worked. Impedance minimums are cut out. So the heavy power at the port tuning frequency is not applied. Possibly ~25-30% less input where it matters. That´s a decibel sir.
2) As the impedance valley is cut off, there is no cone movement "minima" compared to conventional port design. So better cooling where it matters.
That´s a decibel where it matters.
3) As the port is tuned low, cone excursion is increased across the band compared to conventional design. More cooling, less compression.
4) Port is tuned off band and highly optimized. Last sim for "hard driving" scenario showed ~8m/s particle velocity. Again, low port compression compared to the conventional design. That´s a decibel.
5) Finally, driver that is capable of great displacement with little loss on the suspension was chosen with under 100W loss* at Xmax 14mm. and still fine around 15mm as presented. I did have some B&Cs in my hands, and I believe these measured in the ballpark of 170W and were very steep above that. The same power pushing 21DS115 to 17,5mm pushed RCF to 21mm. In the lowest end, that´s a decibel.

Mind you, I am not saying my system is exactly winning. It is gaining those decibels back from lost sensitivity of such undertuned design. This is trading blows with the conventional design in certain particular aspects, winning 1dB here, losing 1dB there. BUT. This design has it easy thermally, is more linear with power, is compact, can produce deeper bass with limited input too, with 30Hz tuning. I mean, it does have merits, it is not nonsense. I do accept it might not be preferred. The difference shrinks with live and mild music, and shows itself more with synthetic and heavily compressed stuff.
 
That´s where my design comes in. Of course any system suffers power compression including mine, but I did wok on that five ways.

1) Only impedance peak is worked. Impedance minimums are cut out. So the heavy power at the port tuning frequency is not applied. Possibly ~25-30% less input where it matters. That´s a decibel sir.

Well, that's a major fault. At the fb there's the minimum excursion, so applying power there gets you spl increase like nowhere else!
If you want to boost anything, it's the fb you want to push because that's where the power compression applies the least since you've got there the least excursion. If you really, REALLY want to increase the bass output, that's where you have to put the EQ. You don't gain a decibel but probably lose a whopping 3dB!

2) As the impedance valley is cut off, there is no cone movement "minima" compared to conventional port design. So better cooling where it matters.
That´s a decibel where it matters.

No, that's not how you do it. That's where you can boost without exceeding the Xmax. As long as there are other frequencies are going on (hint: they are!), you don't lose any cooling. If you've got a 1-note bass for half an hour on exactly the fb, that's another matter. But for that, there are limiters.

3) As the port is tuned low, cone excursion is increased across the band compared to conventional design. More cooling, less compression.

That's not acutally true. If a sub is tuned very low, the max excursion is usually reached at higher frequencies which means higher distortion at higher frequencies, no matter the cooling.

4) Port is tuned off band and highly optimized. Last sim for "hard driving" scenario showed ~8m/s particle velocity. Again, low port compression compared to the conventional design. That´s a decibel.

That's ofc true and good planning. 👍

5) Finally, driver that is capable of great displacement with little loss on the suspension was chosen with under 100W loss* at Xmax 14mm. and still fine around 15mm as presented. I did have some B&Cs in my hands, and I believe these measured in the ballpark of 170W and were very steep above that. The same power pushing 21DS115 to 17,5mm pushed RCF to 21mm. In the lowest end, that´s a decibel.

That's a misconception. At a that high excursion, the suspension is already in progression. While the driver technically can do that excursion, it's just thought as headroom. You don't want ever to continously reach the Xmax ever. Don't forget, it sounds a lot worse if anything goes in progression.

Mind you, I am not saying my system is exactly winning. It is gaining those decibels back from lost sensitivity of such undertuned design. This is trading blows with the conventional design in certain particular aspects, winning 1dB here, losing 1dB there. BUT. This design has it easy thermally, is more linear with power, is compact, can produce deeper bass with limited input too, with 30Hz tuning. I mean, it does have merits, it is not nonsense. I do accept it might not be preferred. The difference shrinks with live and mild music, and shows itself more with synthetic and heavily compressed stuff.

I can only say, some of your calculations are off, some fit but in the end you spend a lot more money replacing the drivers than placing a few more would cost. If you can get the drivers at a very cheap price, it's nice you can take that hit. For most ppl who need to earn money with the PA, that's a net loss however you see it. If you earn money by that and can live with the losses, go for it! And I'm not the one who decides how to spend your money, I just try to help. If you don't want that, it's fine by me too.
 
Well, that's a major fault. At the fb there's the minimum excursion, so applying power there gets you spl increase like nowhere else!
Uh oh, no. You can't have it both ways. At Fb, the power input and compression is the highest. No free SPL either. If the DJ decides to push steady bass line at that frequency to the listeners intestines, the speaker is a toast. If you ad +EQ in there, it is a toast within 1-2 minutes., unless the limiter kicks in with -10dB cut or about, by which point any EQ you were thinking you were doing is outta window, or even kicking down the rest pf the band with it.
Any push at ob is instant loss for the conventional design. That's the lowest point of the efficiency curve, exaggerating the issue with it.

With lower tuning the peak is also lower.

Similar for the rest. The drivers are not same and some are happier to work near Xmax than others. The LF21N551 went for 21mm like nothing. Mind you, the box stood and worked. Worked adequately against two 18"s in restricted range. That's not a theory. And I will prove it with next build, and hopefully measure all necesary data.
 
Last edited:
Uh oh, no. You can't have it both ways. At Fb, the power input and compression is the highest.
We are taking about Xmax. That is the frequency where you can boost without exceeding the Xmax. It's correct that there's the highest power but that's the only frequency where you can get more spl with power if you don't want to exceed the Xmax. Yes, you get a higher power compression but I already wrote you don't want to drive the subs at 100%. And you don't want the subs doing the compression in sound. It depends on the tuning and the music played what max is reached first, Xmax or power.