I've recently underwent the process of getting a subwoofer to replace the countless unused subwoofers I have that would trump them all. While reading about the subwoofer I decided to get (Definitive Technology Supercube I) I read about 2nd & 3rd order harmonics and how they can become prevalent below the subwoofer's ability to perform in a linear fashion. I'm assuming this has something to do with THD (total harmonic distortion) but I want to know how significantly noticeable such distortion would be below 30Hz.
Now before someone jumps all over my thread and tells me how big a mistake I made in choosing this subwoofer, please understand I have chosen this subwoofer for several reasons. I purchased it used. If I don't like how it sounds/works then I can sell it.
So I have the following queries.
1 - Can someone explain how these distortion patterns work (2nd order harmonics, 3rd order harmonics, THD, anything else that I'm missing)?
2 - It seems most people that have this subwoofer and one's like it get the impression that it goes much lower than 30Hz with substantial output. Could it be that what they're hearing is distortion without realizing it?
3 - Is it possible that this distortion at such low frequencies would be difficult to differentiate from similar frequencies that are not distorted?
4 - Is the distortion in any part due to mechanical or electrical stress of the subwoofer?
Thank you all for any helpful information you are able to provide.
You have asked a very complex set of questions, so let us try and make it as simple as we can.
First all sounds are built on a series of sine waves no matter how ugly they look.
Now lets take an amplifier and drive it to clipping, because it runs out of voltage to put nice round waves on the scope. Since all tones are sine waves, then that picture going from pretty to ugly is generated by adding sine waves to the original tone. Now frequencies twice the original tone are the second harmonic, those three times the third harmonic and so on. Psycho acoustically even harmonics are much less unpleasant than odd ones, unless you are a rock fanatic and you for reasons unknown to me will find odd harmonics pleasing and go out of your way to produce them.
All musical instruments produce harmonics, and most the even harmonics. The intensity of the harmonic content drops off with increasing frequency. A violin string and an open organ pipe are rich in even harmonic content. A stopped organ pipe (Gedeckt) produces only odd harmonics.
Now speakers. These are mechanical devices and have both mechanical non linearity and resonant modes, both mechanical and electrical. However these ills manifest themselves they can only do via the generation of sine waves which we call harmonic distortion.
If a driver is pushed beyond it mechanical limits (xmax) then it won't produce a symmetrical sine wave, so if the output is different from the input there has to be added sines waves or harmonic distortion.
Now all speaker systems have resonance. For one every driver, because it is springy has resonance the Fs. There is also the box resonance in a sealed enclosure and the box tuning resonance in ported and ABR tupe enclosures like your Supercube I (Fb). Pipes like TLs have a pipe resonance (Fp).
Now every resonance has a spread. This spread is defined by Q, or quality of the resonances. Low Q resonances are narrow and high Q resonances wide. Resonances also have amplitude, which is their output contribution.
Now the worst resonances are high amplitude and high Q.
Now one fact to keep in mind about speakers. The Q of the total system can NEVER be lower than the Qts of the driver, which is derived from the driver mechanical and electrical resonances. Qt of the whole system will always be quite a bit higher than the Qts of the driver. That is why I personally avoid using high Qts drivers in designs.
The next issue is that Fs pretty much sets the lowest useful bass extension of the driver, and for most it will be a bit above Fs and for some quite a bit above. Some low Qts drivers can be driven below Fs, because the Q of the resonance is narrow and therefore not objectionable.
Now in a sealed enclosure the distortion is dependent on the drivers behavior, and in general distortion will rise quickly below Fs and in general it is a good idea to have a high pass filter engaged around Fs. All of the sound is produced by the movement of the driver cone or cones.
Now in ported or ABR enclosure, there is a deliberate intent to create a resonance to control cone motion and and augment the driver output with this tuning with optimal matching of Fb and Fs. In this region, the cone of the driver is pretty much stationary and almost all the output comes from the port or the ABR. The output is usually 180 degrees out of phase with the driver, so there is cancellation, but port or ABR output is so dominant it is not significant. This waveform has a huge harmonic content, but fortunately at these frequencies the ear is not very sensitive. There is because of this, significant time delay between driver and port or ABR.
Below the F3 of the system the driver decouples from the box and output falls fourth order for ports and nearer sixth order for ABRs. As output falls harmonic distortion rises off the clock.
In a TL only odd harmonics are generated and can be controlled by driver placement at the node of the dominant third harmonic and by damping. With driver Q in the 0.3 to 0.35 range and correct damping to suppress the impedance peak of tuning, the Qt of the system can be brought to the 0.5 range, and still have useful port output to reinforce the driver with comparatively very low distortion. Since the pressure in the pipe at the driver is very high, the driver excursion is very well controlled and mechanical distortion minimized. In low bass long lines however time delays are highly significant due to the time it takes the sound to traverse the line.
In my view the quality of the bass really comes down to system Qt more than anything. I find that the vast majority of systems have too high a Qt. This leads to bass blooming and excess excitation of room bass modes. The bloom and distortion from this poorly controlled resonance Qt is easily recognizable. It destroys the definition of double basses, cello strings, bass piano strings deep organ notes, bass drums and I could go on and on.
This problem is so prevalent that listeners have come to expect it, especially long time owners of audio systems. In fact a period of readjustment I find is almost always required when encountering a properly damped low Qt system. However almost always after a period of reorientation, they return to their systems for ever dissatisfied.
So low Q systems can only be built around sealed enclosures, TLs, open baffles, infinite baffles and with limitations coupled cavity systems. In the later two, Q is set by the drivers. However the driver is unconstrained and so linear driver design becomes of paramount importance. Multiple large drivers are required for infinite baffle and open baffle systems really can't achieve the dynamic range required for all sources.
So that leaves the sealed enclosure. Unfortunately when selecting a low Q driver cut off will be quite a bit higher than Fs. Therefore large amplifier power and robust drivers, with huge linear xmax numbers are required, to tolerate the huge bass boost Eq required.
Horns can get you there as well, but the enclosure are huge and distortion in horns is high.
If you have the space, that leaves the TL as my favored solution. I think they do sound better in room than sealed enclosures, because of the phenomenon of encirclement from pipes, where you have much more even distribution of sound throughout the space.
The coupled cavity has some attractions, as the designer usually was wide control over Q. However all output is from the port and relatively high and rises with the order of the design. As Q is lowered obviously the operating bandwidth narrows. If you don't believe this just go and listen to the boomy bass of the wide bandwidth of the Bose Acoustimass band pass bass modules! So I do use this, but just as augmentation to a speaker with already decent bass performance, where narrow bandwidth augmentation of an octave or less is required.
One other issue which gets talked about is generation of of beat harmonics below cut off. I think this is a huge red herring. I have never seen it, or heard it, and it would take more than one speaker. I suppose it is possible.
It would happen if you generated two frequencies very very close together. Then you get alternating reinforcement and cancellation, the the generation of very low frequency beat tones. This has long been perfected by Tibetan monks, and in their hands sounds quite incredible. It requires years of practice and patience to perfect.
As far as your speakers are concerned, as is usual with manufacturers there is very little useful information. The manufacturer claims response to 13 Hz, with no relative level parameters. I would say any output from that device at 13 Hz will be miniscule and 100% THD.
It is a resonant system and will be in the higher Qt category, how high I do not know.
I hope this helps you.
