How to start the "drive unit" of speaker design

Drive unit

The drive unit in the speaker determines the final effect that the speaker can achieve (of course, if the rest of the design is correct), and plays the most important role in the entire HIFI system. As mentioned above, there is no perfect drive unit in the world (it may appear in ten years or more). Because the diaphragm of the perfect drive unit requires the same density as air, there is excellent uniform linear motion in the full frequency range, and there is no various types of distortion-we still have a lot of way to go to achieve this. But don't be discouraged. With the rapid development of materials science, I believe that every two or three years there will be major breakthroughs to promote the development of the driver. Thanks to a large number of computer simulations applied in the fields of material mechanics, aerospace technology, etc., researching high-performance materials to replace the original expensive and heavy materials. Now Kevlar, synthetic carbon fiber and cast aluminum are widely used. We are looking forward to the application of synthetic diamond, ultra-low density silicon glass colloid, new metal and single crystal carbon in the diaphragm.

The problem to be weighed by all types of unit designers is that the diaphragm must maintain good piston movement (requires rigidity) without generating resonance at mid and high frequencies (requires good self-damping), and it also brings two Additional problems: cavity resonance and nonlinear magnetic force.

Piston movement:

The rigid diaphragm means that the acceleration of the voice coil can be quickly and accurately transmitted to the entire surface of the cone and the dome. It has a fast impulse response, low modulation distortion and a very clear and transparent sound. "But usually designers of" objective schools "will question:" After passing through the frequency divider, the inductance blocks the rising speed of the pulse. Why is the bass and midrange unit faster? "This is actually a misunderstanding, they did not say One piece, in fact, "fast" here means that the unit has piston movement in a large range, and its frequency response is straight, it can respond quickly to pulses, and the attenuation of after vibration is fast. So is it hard enough to make a diaphragm? Of course not, such as bronze, which is hard enough and easy to form. It was used in ancient times to make bells and bells. It has a long time of resonance and the attenuation is very slow. The clock itself needs a long enough time for vibration attenuation, so the structure is made special to achieve the purpose; on the other hand, the vibration of the clock is attenuated by air, and its coupling with air is low, so the damping caused by air is very small .

Self-damping:

Similarly, we require the voice coil to brake the cone and dome, and do not want them to vibrate when they pass the voice coil without a signal, but unfortunately, all materials currently have only a little self-damping, and the vibration continues The time is very long (that is, the Q value of the vibration is high). One method is to use a very heavy rubber ring to stick to the diaphragm, but this will make the frequency response uneven and the sensitivity greatly reduced, so it is not feasible.

At present, the best ballistic fiber, carbon fiber and aluminum basin units have at least a high Q resonance point at the high end of their operating frequency. At this point, the sound pressure is very high, and the after vibration is very large, resulting in a large amount of sound staining, so it is sometimes called Split points for pots. And the splitting of the ballistic fiber, carbon fiber, and paper cone units is not gradual, but sudden. This requires a high-order filter with rapid attenuation and some notches to correct the resonance peak, but usually these points are between 3-5KHZ, the human ear is very sensitive to it, and a little staining will be noticeable.

In fact, the current widely used two-way speakers have this problem. These designs usually use 6-8 inch bulletproof fiber, carbon fiber and aluminum basin units combined with tweeters, because their crossover points are usually set at 3-5KHZ. This makes it difficult to take into account both the suppression of its resonance point and the coordinated sound with the tweeter at that point. For example, in order to reduce the influence of the resonance point of the woofer unit, the crossover point is lowered, the tweeter will have a larger power input, and it is easy to exceed its linear range, which will increase the distortion. The self-damping of the tweeter also becomes worse, the treble sound is stained loudly, and the frequency response is also large. If the crossover point is set high, the phenomenon of splitting the woofer basin will appear, making the sound staining very serious. Therefore, a better design usually uses a 24DB / OCT filter.

By the way, I like the sound of bulletproof fiber and carbon fiber. As mentioned above, the rigid cone has its speed advantage, but it is too difficult to control the split phenomenon of the basin. Therefore, some people use high internal friction materials to make cones (formerly plastic, but now gradually replaced by polypropylene and other materials), they have good self-damping. This kind of material usually has a very smooth frequency response curve, and even a 6dB / Oct filter can be used, but most of them I do n’t like, because I feel that the sound is fuzzy when outputting at low and medium sound pressure, and the softer materials are usually It also brings greater distortion.

I think this usually happens in soft dome tweeters. There are basin splits in the entire working frequency band. Although their high self-damping makes this situation impossible to measure with instruments, the ears can hear it. But currently the best soft dome tweeters use some synthetic techniques and coatings to improve hardness, and do not cause deterioration in frequency response. Excellent examples are high-end soft dome tweeters from Dynaudio, Scan-Speak, and Vifa.

Cavity resonance:

Although the dust cap of the mid-bass unit (or the dome of the treble) looks nothing, it forms a small resonance cavity with the central magnetic pole. A well-known example is the B110 unit used on the LS3 / 5A. There is a wide range of frequency response at 1500HZ. There are also 3 very high Q peaks near 4500HZ. This is actually a typical dust cap. Resonated. The 1 inch soft treble of the popular AUDAX in the 1970s also had the above problems at 9 to 16KHZ. Previously, materials such as filled wool were used to suppress the peaks of these points, but the effect was not good, and there were still 1-3DB peaks.

At present, two methods are usually used to solve this problem. One is to use a magnetic pole with a central vent (eg Dynaudio, Scan-Speak, Vifa), and the other is to use a phase plug instead of a dust cap (eg Audax, Focal) , Dynaudio's Esotec D-260, Esotar T-330D, and Scan-Speak D2905 / 9000 tweeter successfully use the magnetic pole of the central vent, so the rear load is similar to the transmission line, which can produce good damping. They are used in the widely praised Sonus Faber Extrema and ProAc Response 3. In contrast, Focal's T120 and T120K use reversed glass fiber and bulletproof fiber domes, which produce many peaks between operating frequencies. Although many people applaud it, I don't like it very much.

Magnetic field nonlinearity:

If the magnetic field is constant, similar to the air core inductance, the inductance value generated by the voice coil is also constant, so the change in impedance with frequency can be cancelled by using a simple RC compensation circuit. However, the voice coil is placed in the magnetic field and moves in the magnetic gap. The magnetic field distribution is nonlinear, so the inductance of the voice coil is also nonlinear, so its impedance changes with frequency is also nonlinear.

This non-linearity brings many problems. First, it affects the high-end frequency response of the unit, and secondly it causes the delay of the sound (relative to the tweeter). The input of high-power signals exacerbates this situation. When the displacement of the voice coil exceeds its linear displacement, the magnetic field changes more. For example, the preferred 8 "Vifa unit P21W0-12-08, its linear displacement is only 8mm (positive and negative 4mm), other typical 8-inch units are only 6mm, most midrange units are only 1-3mm, when they emit At low frequencies, it often exceeds its linear range, and the inductive modulation of the voice coil appears, producing IM and FM modulation distortion in the entire frequency range. In this case, a large number of 2 channels or 3 channels with a lower intermediate frequency crossover point are generated. In the system.

Is there a solution? Of course, Scan-Speak's SD magnetic circuit system and Dynaudio's DTL magnetic circuit system use copper short-circuit rings to reduce the self-inductance of the voice coil. For example, the 8 "Scan-Speak 21W / 8554 is perhaps the best 8 in the world. For the inch unit, its voice coil inductance is only 0.1mH, and the comparison 8 "Vifa P21W0-20-08 is as high as 0.9mH.

In addition, the problem of changing the self-inductance of the voice coil also brings more hidden problems. We know that the high-end frequency roll-off characteristic of the drive unit is caused by the mechanical roll-off characteristic of the unit and the self-inductance of the voice coil (which causes the electric roll-off Characteristics) It is determined that the frequency of the mechanical roll-off characteristic of a good unit is lower than that of the electric roll-off characteristic, making the overall characteristics of the synthesis better. However, the electrical roll-off characteristics of many units are lower than the mechanical roll-off characteristics, which will produce strong modulation and transient characteristics deterioration. var check_allow = "/d/Api/iscantalk.html"; var add_url = '/ d / article / write /'; function CheckLogin () {now_uid = ''; var ElecfansApi_checklogin = '/ webapi / passport / checklogin'; var logout_url = "{: U ('Login / logout')}"; var logout_url = 'http://bbs.elecfans.com/member.php?mod=logging&action=logout&refer=front'; $ .get (ElecfansApi_checklogin, function (data, textStatus) {if (data! = "") {EchoLoginInfo (data); CheckEmailInfo (data); data = $ .parseJSON (data); now_uid = data.uid; / * var login_content = 'write an article