Apparently I need the floor, otherwise I get a dip long before 200Hz, which is the desired bandpass. I use many pillows and a blanket between the Umik-1 mic and the driver at 50cm distance (bright green line). I also tested with no pillows and distances 50cm, 100cm and 180cm. The Dark Green line is with no pillows, 50 cm away and closed port - so smooth! I also have the same settings but 180cm distance in the pink line.
At 180cm when it is closed there is very little bass below 70Hz actually, but it could be at just that spot in the room.
I ran a sine wave at 43Hz when open and it was super loud at some spots and much less at some spots. The difference was less for the peak at 72Hz when walking around in the room. The 550Hz dip is not noticable as it varies where this floor bounce cancellation, or whatever it is, occurs.
All have 1/6 smoothing.
At 180cm when it is closed there is very little bass below 70Hz actually, but it could be at just that spot in the room.
I ran a sine wave at 43Hz when open and it was super loud at some spots and much less at some spots. The difference was less for the peak at 72Hz when walking around in the room. The 550Hz dip is not noticable as it varies where this floor bounce cancellation, or whatever it is, occurs.
All have 1/6 smoothing.
Last edited:
Hmmm... That is why I asked for measurements done right at the usual listening place (50 or 100 cm, even 180 cm is too close). Than you may move some foot or two forward or backward (speaker, or listening place) and notice if there is some improvement. I made such measurements at several places and it was very tedious and time-consuming, but in every occasion I managed to find ideal balance, almost without dips and peaks.
I plan to listen to my system a bit everywhere in the room, so maybe I should just stop obsessing about these peaks.
Forget what I wrote. The dips and peaks are pretty uniform in the room. I ran sinus signals at certain bass frequencis and walked around in the room and listened for them.
Of course they are uniform (more or less) - they are the fixed room modes. But it is possible to find the optimum position where their negative impact is minimal. As I wrote - it is tedious and time consuming measurement job.
Back to the thread theme: woofer efficiency... and how to exploit it to our needs (if possible).
After many, many ill-defined, confusing and contradicting (and some blatantly wrong!) comments, arguments and answers by the OP (partially maybe as a result of non-native English... although I am not much better either), I concluded, hopefully – correctly, that his main idea is to:
Use very high efficiency 21” or 18” professional woofer (high BL, low Qts) in a vented enclosure with deliberately non-flat alignment (lower tuning frequency), and after that – boosting (with an equalizer) manifested depression in the frequency response, to bring back flat frequency response. Boosted frequency band should be in the higher efficiency region of the woofer, preferably without much energy penalty caused by that boosting. Very appealing, environmentally green agenda “less Watts – less heath”! So far so good.
But... big and super-efficient 18” pro woofers have much bigger magnets, which use much more energy and raw material to manufacture, which results in more environmental pollution! All of sudden – hidden nasty variable is rising its ugly head... and it is not alone!
In my post #38:
As I already mentioned, there are no commercial subwoofers using this design, although it might find use in small battery-powered portable bluetooth loudspeakers.
When we try to apply this old idea to vented loudspeakers, as OP tried to (emphasizing the “green” agenda), we are facing more hurdles – impedance peak (the second one) is moved higher up in the frequency compared to the single impedance peak from the same woofer mounted in a closed box (i.e. max efficiency region moves up in frequency). So, we must use bigger EQ boost (to flatten the frequency response) applied at lower impedance value than in a closed box example – and that means more Watts, more heating, more greenhouse gases... Faced with this unexpected situation, OP later (his post #124: “... It doesn’t require any aqualisation.”) negate the need for EQ boosting – contradicting himself! Previously, OP was mesmerized by the efficiency graph only, blindly overlooking consequences of the loudspeaker frequency response which he deliberately made non-flat, in contrast to maximally flat alignment!
But the first and the most important mistake by the OP was comparing randomly chosen woofers slapped in arbitrarily picked enclosure volume. Correct procedure is to choose/design the optimal woofer TS parameters (i.e. max efficiency) for a given enclosure volume and given F3 (rings a bell? Yes – it is the all-mighty Hofmann’s Iron Law!). After that, we can try to design/choose new TS parameters for a new woofer best suited to the new “max efficiency, non-flat alignment tuning, with EQ boost to flat” agenda, keeping the same enclosure volume and same F3. Only after fulfilling these requirements, we can find whether this idea implemented in vented loudspeakers produces higher overall efficiency. The big problem is how to evaluate this “higher overall efficiency”, because boosted lower-efficiency frequency band draws more Watts than other higher-efficiency non-boosted frequency bands. For example - two different woofers optimized for 100 liters enclosure volume:
Woofer 1 (optimized for conventional TS maximum flat alignment): F3=35Hz, SPL=90dB/2.83V/1m.
Woofer 2 (optimized for maximum efficiency non-flat alignment, plus EQ boost): F3=60Hz, SPL=93dB/2.83V/1m, so we need +6dB boost (4 times more Watts!) at 45Hz to produce the same F3=35Hz maximally flat frequency response as the previous example with woofer 1.
The big question is – which subwoofer design is more “green”? Obviously, woofer 2 is 3dB more efficient above 60 Hz, but 3dB less efficient (2 times more Watts!) at 45Hz (and below) than woofer 1. So, not so easy to decide, isn’t it?
Professional loudspeaker manufacturers already decided - all passive pro subwoofers are of the conventional TS maximally flat alignment type. Even if they recommend some boost (with the same model sub), it is always at, or slightly above, the tuning frequency - so no lower-tuning non-flat alignments turned-flat-with-boost pro subwoofers per OP idea, sorry! Including OP's beloved Powersoft/IPAL subwoofers!
Even if we concentrate on the efficiency graph alone:
we can see the randomly chosen, wimpy 10” medium-efficiency woofer (in grey) is equal in efficiency in the (approx) 38Hz-45Hz frequency band compared to the ultra-expensive super-efficient 18’” IPAL woofer (in black). Imagine what can be done with optimized 10” woofer!
As we have seen (measurements by @Rewind – thanks!), in some/most domestic rooms we don't need super-efficient super-expensive pro 21" or 18" IPAL woofers: humble medium-efficiency 10" hi-fi woofer is more than enough in a real world room – we even have to use equalizer to attenuate the overexciting SPL at bass range and save the Earth with successful green “less Watts – less heating, think globally – act locally” deed. Attenuation means less Watts, so – less heating, less carbon pollution, more pandas in the forests etc.
After many, many ill-defined, confusing and contradicting (and some blatantly wrong!) comments, arguments and answers by the OP (partially maybe as a result of non-native English... although I am not much better either), I concluded, hopefully – correctly, that his main idea is to:
Use very high efficiency 21” or 18” professional woofer (high BL, low Qts) in a vented enclosure with deliberately non-flat alignment (lower tuning frequency), and after that – boosting (with an equalizer) manifested depression in the frequency response, to bring back flat frequency response. Boosted frequency band should be in the higher efficiency region of the woofer, preferably without much energy penalty caused by that boosting. Very appealing, environmentally green agenda “less Watts – less heath”! So far so good.
But... big and super-efficient 18” pro woofers have much bigger magnets, which use much more energy and raw material to manufacture, which results in more environmental pollution! All of sudden – hidden nasty variable is rising its ugly head... and it is not alone!
In my post #38:
I gave a link to AES paper from 2005 which studied in details the same idea, but applied to closed box loudspeakers. Naturally, frequency band of interest is where the loudspeaker impedance is higher than average/nominal – around the loudspeaker resonance, which automatically gives more efficiency. Optimal TS parameters for maximum efficiency and maximum usable (subwoofer) bandwidth were found.
As I already mentioned, there are no commercial subwoofers using this design, although it might find use in small battery-powered portable bluetooth loudspeakers.
When we try to apply this old idea to vented loudspeakers, as OP tried to (emphasizing the “green” agenda), we are facing more hurdles – impedance peak (the second one) is moved higher up in the frequency compared to the single impedance peak from the same woofer mounted in a closed box (i.e. max efficiency region moves up in frequency). So, we must use bigger EQ boost (to flatten the frequency response) applied at lower impedance value than in a closed box example – and that means more Watts, more heating, more greenhouse gases... Faced with this unexpected situation, OP later (his post #124: “... It doesn’t require any aqualisation.”) negate the need for EQ boosting – contradicting himself! Previously, OP was mesmerized by the efficiency graph only, blindly overlooking consequences of the loudspeaker frequency response which he deliberately made non-flat, in contrast to maximally flat alignment!
But the first and the most important mistake by the OP was comparing randomly chosen woofers slapped in arbitrarily picked enclosure volume. Correct procedure is to choose/design the optimal woofer TS parameters (i.e. max efficiency) for a given enclosure volume and given F3 (rings a bell? Yes – it is the all-mighty Hofmann’s Iron Law!). After that, we can try to design/choose new TS parameters for a new woofer best suited to the new “max efficiency, non-flat alignment tuning, with EQ boost to flat” agenda, keeping the same enclosure volume and same F3. Only after fulfilling these requirements, we can find whether this idea implemented in vented loudspeakers produces higher overall efficiency. The big problem is how to evaluate this “higher overall efficiency”, because boosted lower-efficiency frequency band draws more Watts than other higher-efficiency non-boosted frequency bands. For example - two different woofers optimized for 100 liters enclosure volume:
Woofer 1 (optimized for conventional TS maximum flat alignment): F3=35Hz, SPL=90dB/2.83V/1m.
Woofer 2 (optimized for maximum efficiency non-flat alignment, plus EQ boost): F3=60Hz, SPL=93dB/2.83V/1m, so we need +6dB boost (4 times more Watts!) at 45Hz to produce the same F3=35Hz maximally flat frequency response as the previous example with woofer 1.
The big question is – which subwoofer design is more “green”? Obviously, woofer 2 is 3dB more efficient above 60 Hz, but 3dB less efficient (2 times more Watts!) at 45Hz (and below) than woofer 1. So, not so easy to decide, isn’t it?
Professional loudspeaker manufacturers already decided - all passive pro subwoofers are of the conventional TS maximally flat alignment type. Even if they recommend some boost (with the same model sub), it is always at, or slightly above, the tuning frequency - so no lower-tuning non-flat alignments turned-flat-with-boost pro subwoofers per OP idea, sorry! Including OP's beloved Powersoft/IPAL subwoofers!
Even if we concentrate on the efficiency graph alone:
we can see the randomly chosen, wimpy 10” medium-efficiency woofer (in grey) is equal in efficiency in the (approx) 38Hz-45Hz frequency band compared to the ultra-expensive super-efficient 18’” IPAL woofer (in black). Imagine what can be done with optimized 10” woofer!
As we have seen (measurements by @Rewind – thanks!), in some/most domestic rooms we don't need super-efficient super-expensive pro 21" or 18" IPAL woofers: humble medium-efficiency 10" hi-fi woofer is more than enough in a real world room – we even have to use equalizer to attenuate the overexciting SPL at bass range and save the Earth with successful green “less Watts – less heating, think globally – act locally” deed. Attenuation means less Watts, so – less heating, less carbon pollution, more pandas in the forests etc.
Last edited:
Sheesh! 🙂
No, that is wrong, as one gets not only more SPL per Watt, but multiples of SPL.
If we look at the solution needing particular SPL, we see that this solution will take much less bins, maybe less amps, definitely less trucks, fumes, less people to setup the rig, less space, and so on. This is yet another mistake by trying the "economics and environment shot".
They use the design of higher impedance. They just do not rely on that resonant part.
You know perfectly well that this is not what I did.
Compared to other solutions with particular SPL goal, it does not mean more watts.
False statement. I wrote an answer for false claim about efficiency. The driver does not need any EQ to be more efficient. It only needs EQ for flat response, which was not what you wrote in the first place.
Never did that. I actively pushed that frequency response of the driver or box absolutely does not have to be flat, and it is encouraged to hunt for flat response. Never overlooked that. This is clearly a false statement, if not a lie.
Not exactly. The volume was set rather low to demonstrate SPL density of powerful speakers.
Nonsense. There are more important goals.
As mentioned, efficiency is not a thing at resonant frequency (impedance peak) of the driver, as there is no issue if it eats 10Watt or 20Watt. In the less efficient band though, it becomes an issue if it eats 1000Watts or 2000Watts. You still did not get a grasp on the approach.
This is rather imaginary case, not a proof, and the math is wrong. You need 4x more Watts to get THAT woofer in line with the first one. But what would probably happen is that such woofer would take as much times less Watts to begin with, so the boost would not result in more power taken.
All the graphs are showing it to you in some milder way, yet you do not understand that.
Sensitive, not efficient. You again chose wrong elements for counting efficiency. Sensitivity is not efficiency.
If we put the IPAL driver in flat alignment, it will play "no bass", and will burn down for lack of cone excursion and cooling at expected program power input. You have no idea in a way that it is really not funny.
We can see that weak woofer has high efficiency at "single point" due to weak suspension. Please by all means, do find some more optimized woofer that can outdo IPAL, and best, outdo it at comparable reachable SPL.
Of course we do not, that is okay. But what does it do in this thread?
That happens in most rooms, that is no badge of strong output of the driver. My 4" Daytons TCP115s need that treatment too.
Yes, drivers need that in poor/poorly treated room generally. That has nothing to do with woofer efficiency.
This is all but ontopic discussion.
False claims.
Lies about my responses and context.
Offtopic.
Name calling.
Attempt to damage reputation.
I am kind of stunned admins do not care.
Maybe bad platform of choice. Might be happier elsewhere, indeed. And Yes, I don´t let the door hit my a$$ if needed.
The constant bombardment of fallacious comments against Crash is getting annoying. Can we stop with the straw-man arguments and the blatant misrepresentation of his positions please?
This is great work @Crashpc. I've been doing some modeling for a new subwoofer and I can really see the advantages of these high motor strength drivers. It's pretty easy to model them out in WinISD with filters to get an congruent response against popular hifi drivers and look at the apparent power. I think one downside is we may need amplifiers that can output a lot of voltage so that EQ can take effect.
This is great work @Crashpc. I've been doing some modeling for a new subwoofer and I can really see the advantages of these high motor strength drivers. It's pretty easy to model them out in WinISD with filters to get an congruent response against popular hifi drivers and look at the apparent power. I think one downside is we may need amplifiers that can output a lot of voltage so that EQ can take effect.
WinISD can help, but as it is showing only apparent power and not real power, the evaluation might end up in wrong conclusions still. A phase angle calculation can give you about 50% less power at certain points of the power response. That is too much to ignore.
I think the development of WinISD was halted, and so there is probably no point in asking if the owner/developer yould add real power graph. That would be way too cool too.
I think the development of WinISD was halted, and so there is probably no point in asking if the owner/developer yould add real power graph. That would be way too cool too.
Why EQ anything when you can build with a woofer that is designed for a healthy amount of musical bass that can be tamed by attenution?
A concern with all this talk of EQ bass boost is the question - what are you amplifying? If the woofer is only whispering bass, and you raise the bass whisper to a bass shout, your end result may not be more than unwanted distortion pretending to be bass. What if the woofer was not whispering bass at all and just could not do good bass? It was not built for it. Its natural state was to produce midbass and a lot of shouty midrange frequencies.
The A26RE4 even has high Le and a four layer voice coil that causes a natural rolloff at 1500Hz. And it sounds good up to 1500Hz. So not need for complex crossovers, maybe one inductor, if you are looking for a 2-way, or 3- way. Just fit and forget.
All I am worried about is presentation - does it sound good? Maybe a higher motor strength would give a faster feel to the bass, and less sloppyness. But it would also lower Qts and reduce the deep bass.
A concern with all this talk of EQ bass boost is the question - what are you amplifying? If the woofer is only whispering bass, and you raise the bass whisper to a bass shout, your end result may not be more than unwanted distortion pretending to be bass. What if the woofer was not whispering bass at all and just could not do good bass? It was not built for it. Its natural state was to produce midbass and a lot of shouty midrange frequencies.
The A26RE4 even has high Le and a four layer voice coil that causes a natural rolloff at 1500Hz. And it sounds good up to 1500Hz. So not need for complex crossovers, maybe one inductor, if you are looking for a 2-way, or 3- way. Just fit and forget.
All I am worried about is presentation - does it sound good? Maybe a higher motor strength would give a faster feel to the bass, and less sloppyness. But it would also lower Qts and reduce the deep bass.
What we typically see in drivers that have strong motors is also less distortion. So even with EQ there could be a distortion advantage.
Subwoofer design with EQ is a perfectly fine thing to do when these things are taken into account.
I also don't think it's a matter of 'its not built for it'. What else is an 18" woofer with a very strong motor meant to do? Yeah, it's VERY loud in upper bass, but that doesn't mean it's poor at making low end bass just because it's damped.
Strong motor with a QTS of 0.2 vs a driver of a QTS 0.8 with a weaker motor. Which one do you think produces less distortion at 20hz equalized?
First, keep in mind that I am not pushing it for your solution. That is another topic. As for why, this has two dimensions. Why EQ anything: Because the desingn is engineered in such a way, that it pursues other factors, such as efficiency and high output", and not flatness. As it can be compensated for in this case, it is not seen as a problem. Not everyones cup of tea, but the objective of higher SPL indeed is reached that way.
The other dimension is, that one cannot reach the healthy amount of bass for PA use any other way, if we restrict the design by size and weight for purposes of logistics. Then the "When you can" stops working, as you can not anymore.
Amplyfiing voltage, not power in comparison to let's say the Seas.
It may be. But it has been measured that it is not so. All the engineering and making sure this works, has been done before starting $$$$$$$ production. Now it is happily used around the world. I myself did some laser measurements of cone movement and extrapolating distortion from that. Vance Dickason in Voicecoil magasine and others like data-bass do that too, and even more. The data is in, for yes to see. There is no need or room for polemics. Independent high power testing is available for your review.
It was absolutely built for it.
See what B&C speakers has to say about their driver:
"This next evolution subwoofer uses a new, longer, four layer aluminum voice coil. The result is more energy in the gap, less heat, higher sensitivity, higher power handling, lower distortion and better overall performance.
this high energy subwoofer is a significant step forward from similar models in the B&C range."
See what RCF has to say about their high energy woofer:
"best control under large excursion conditions."
Great point - how do we set a natural state? How can we say with authority, that natural state of a woofer is, when we supply it with constant voltage? It is the state we were used to, but calling it natural is quite too much.
That'se great. You see what happens when a dtiver is built for certain purpose and used as such. That is absolutely great. Those high power beasts were built for high output between 20-100Hz.
It sounds damn good when used as intended.
I have some personal issues with design decisions made by B&C, I do not like some parts of their solution (very stiff suspension, 2-3mm of not enough usable excursion for my needs), but I buy these anyways, because they sound oh so good.
I have posted short video of bass production few posts ago in my test bin. I did not need to EQ the bass up in my room at all, to get gobs of bass. With more amplification, It started to rattle the dish on my table, while the cone moved only decently, not that much at all.
This really is intended use of such driver. Lot of math, thinking, engineering, veryfying and testing was done before.
No need to push it for your solution. Thus isonly about acknowledging/understanding the physics of it.
Strong motor (high Bl) indeed produces low Qts values, but value of Qts itself has nothing to do with distortion!
Distortion in loudspeakers at low frequencies is caused by:
1. Nonlinearities in Bl (not the value of Bl, but nonlinearities in Bl)
2. Nonlinearities in stiffness Kms (nonlinearities in loudspeaker spider and surround)
Explanations of this are in AES papers by W. Klippel, google them.
Several posters misunderstood/forgot the most important part of the loudspeakers measurements and their subjective impact: Subwoofer frequency response in your room, or in disco club, or in concert hall, or at open stadium must be flat! If it is not flat, it is called linear distortion. Distortion!!!
There is no point in boosting subwoofer output only/primarily in the frequency band where it has the highest efficiency, if it causes linear distortion (non-flat frequency response). Overall frequency response of the subwoofer with EQ (if it is designed that way) must be flat!
There is no point in boosting subwoofer output only/primarily in the frequency band where it has the highest efficiency, if it causes linear distortion (non-flat frequency response). Overall frequency response of the subwoofer with EQ (if it is designed that way) must be flat!
Another misunderstanding is that high subwoofer efficiency as a result of its high impedance at certain frequency band requires "only" special high-voltage low-amperage amplifier. But there are no such amplifiers on the market! All audio amplifiers, whether they are hi-fi or professional, are designed for maximum power output in the 4-8 ohm range - which makes them medium-to-high-voltage high-amperage amplifiers! If you need high voltage amplifier to exploit the potential of high-efficiency high-impedance subwoofer, you have to buy conventional high-voltage high-amperage super-expensive professional amplifier (think LabGruppen, Crown, Powersoft, ...).
The Thomann T.Amp TA2400 MK-X can feed such driver if the program is little on the heavier side. I.E., normal MP3s, Pop, disco, trance, metal and such.
This amplifier will exploit capabilities of 18DS115_8, as it provides ~142VRMS. That is 6,1Ohm at 30Hz in my test bin, and 8,1 at 100Hz. With these voltages, the power input would be 3305VA and 2489VA respectively. This power could easily destroy the driver if the content was too heavy and not limited somehow.
Current price of such amp is 333Eur. I use it in a self-upgraded version with two more pairs of end stage transistors and tuned working point of theese.
I fail to see how it is super expensive. Also with rebates, the driver can be had for 400Eur. So Set of this driver and an amp can be had for 733Eur.
It is not on the cheap side, but for the work it can do, it is nowhere near of super expensive.
With reservations, about flat frequency response is indded needed. But it is still only needed at the output level. The underlying causes of such nonlinearities before output are of no business to ordinary user, especially when he does not know why such approach was chosen, and when EQ mitigates such distortion.
Indeed speaker linearity is very strongly connected to these two points mentioned. But viewed from SPL angle of view, It must be put in context with amount of excursion on the driver, as large cone area driver does not have to move that much, and so nonlinearities caused by change of parameters in X are of lesser amount/consequences, because less excursion happens at the SPL.
I could not find LSI/Klippel data for Seas. It says it has 14mm long coil and 6mm high gap, resulting in 6mm of excursion before the coil starts to leave the gap, where things MUST go awry slower or faster.
1mm out of the gap means that 16,6% of the gap is without coil.
The 18DS115 has 40mm long coil and 14mm long gap, resulting in 13mm of cone movement before the coil leaves the gap, and 1mm out of the gap means that 7,14% of the gap is without coil. The implications of that are known.
On the other hand I know that the suspension becomes exceptionally stiff past 14mm on DS115, and at 17mm, the distortion is very heavy and hearable. Yet they test the driver to 20mm of Xmax. This driver would greatly benefit from the motor if the suspension let the cone move further 2-3mm.
This amplifier will exploit capabilities of 18DS115_8, as it provides ~142VRMS. That is 6,1Ohm at 30Hz in my test bin, and 8,1 at 100Hz. With these voltages, the power input would be 3305VA and 2489VA respectively. This power could easily destroy the driver if the content was too heavy and not limited somehow.
Current price of such amp is 333Eur. I use it in a self-upgraded version with two more pairs of end stage transistors and tuned working point of theese.
I fail to see how it is super expensive. Also with rebates, the driver can be had for 400Eur. So Set of this driver and an amp can be had for 733Eur.
It is not on the cheap side, but for the work it can do, it is nowhere near of super expensive.
With reservations, about flat frequency response is indded needed. But it is still only needed at the output level. The underlying causes of such nonlinearities before output are of no business to ordinary user, especially when he does not know why such approach was chosen, and when EQ mitigates such distortion.
Indeed speaker linearity is very strongly connected to these two points mentioned. But viewed from SPL angle of view, It must be put in context with amount of excursion on the driver, as large cone area driver does not have to move that much, and so nonlinearities caused by change of parameters in X are of lesser amount/consequences, because less excursion happens at the SPL.
I could not find LSI/Klippel data for Seas. It says it has 14mm long coil and 6mm high gap, resulting in 6mm of excursion before the coil starts to leave the gap, where things MUST go awry slower or faster.
1mm out of the gap means that 16,6% of the gap is without coil.
The 18DS115 has 40mm long coil and 14mm long gap, resulting in 13mm of cone movement before the coil leaves the gap, and 1mm out of the gap means that 7,14% of the gap is without coil. The implications of that are known.
On the other hand I know that the suspension becomes exceptionally stiff past 14mm on DS115, and at 17mm, the distortion is very heavy and hearable. Yet they test the driver to 20mm of Xmax. This driver would greatly benefit from the motor if the suspension let the cone move further 2-3mm.