FI122088B - Compensation of rising frequency response in a current-controlled passive speaker - Google Patents

Description

REDUCING FREQUENCY RESPONSE COMPENSATION IN A POWER-CONTROLLED PASSIVE SPEAKER

The present invention relates to electrodynamic moving coil passive loudspeakers designed to be used with power amplifiers. It is directed to a means of coaxialising a free field frequency response characteristic of cone-type loudspeaker elements with current control in the absence of the filtering effect of the inductance of the speech coil. The response correction is achieved in a manner that maintains the speaker element in a substantially current controlled state. Speaker inactivity here means that the speaker does not include amplifier functions.

10

BACKGROUND: QUALITY ADVANCED POWER CONTROL

Today, all commercially available audio amplifier and loudspeaker devices operate on a voltage control principle without significant exceptions. This means that the amplifier acts as a voltage source and thus has a low output impedance. However, both technical considerations and listening experience clearly indicate that voltage control is the wrong choice if sound quality is to be considered important. The reason is that the electric motor forces (SMV) produced by the speaker element interfere with the voltage / current conversion left to the speaker by the voltage control principle.

The driving force (F) of the membrane moving is directly proportional to the current flowing through the speech coil (/) according to F = Bli, where the input B1 is called the force coefficient (B = magnetic flux density; / = wire length in the magnetic field). However, in a voltage-controlled element, this current becomes corrupt in a variety of ways.

The electrical substitution of the moving coil element can be illustrated as shown in Figure 1. Rc represents the resistance of the voice coil, the voltage source em represents the element counter-SMV (motion SMV) calculated as: em = Blv (v = the speed of the voice coil), and the voltage source e represents the inductance SMV derived from the loss inductance of the voice coil .

30 em is generally dominant at the lower end of the element's playback region, and e1 is respectively dominant at the upper end of the playback region. In a typical cone or caliper element, the sum of the absolute values of em and e, over the entire operating frequency range is at least of the order of magnitude of the voltage drop at resistance Rc.

Both em and e are susceptible to many serious interferences, so these voltages do not follow the reproducible signal in a linear fashion, but contain a very wide range of distortion components. By controlling the proxy switch of Fig. 1 with a conventional low-output impedance amplifier 2, which functionally corresponds to a voltage source, sources em and e, and all their indistinct distortion components, modulate the residual voltage Rc and hence the current passing through the circuit.

By contrast, by controlling the proxy switch of Fig. 1 with a high output impedance amplifier operatively corresponding to a power source, sources em and e, and the indeterminate distortion components contained therein, cannot modulate the current in the circuit, now determined directly by the controlling power source following the interference-free signal.

10 The following distractions confuse the power of a voice coil with voltage control, but of course not with power control: 1) The voice coil for microphones reflecting from inside the housing and passing through the membrane 2)

4) Variation of the impedance direction angle due to B1 variation, which causes phase current modulation at the center frequencies.

5) Spindle inductance dependent on deviation position, which causes both amplitude-20 diode and phase modulation to current 6) Spindle inductance depends on signal strength level causing unharmonic distortion 7) Resistance variations due to temperature variations and manufacturing tolerances 8) these serious disturbance mechanisms of voltage control and current control in general are available in: Esa Meriläinen, Current-Driving of Loudspeakers, CreateSpace 2010, USA (ISBN: 1450544002).

THEORY: ACHIEVING CURRENT CONTROL 30 Fully ideal current control as well as perfectly ideal voltage control cannot be realized, so in practice you always have to settle for a compromise between these two extremes. The state in which the element ultimately operates is determined by the relationship between the impedance of the input source and the impedance of the element itself.

In Fig. 2a, the element is supplied from a voltage source (E) via a series impedance (Z). If Z is very small compared to the load impedance Z /, the load current h is determined almost exclusively by Zz, 35, and the load voltage Ui essentially follows the supply source E, so that the element then operates in a voltage-controlled manner. (When talking about the magnitude of impedances, we mean their intrinsic value, that is, the length of the pointer.) 5 If Z, on the other hand, is very large with respect to Z1Mn, ZL represents only a very small part of the total impedance of the circuit. The load current h is then almost independent of ZL \ and its variations, so that the element operates in a current-controlled manner. High serial impedance also requires a high source voltage if the current is to be kept sufficient, so in practice it is not advisable to implement current control according to the principle of Figure 2a.

10

If Z is of the same order as Z /, it can be said that the current is determined by half of each impedance and the variations of ZL are reflected by the current at half intensity compared to the case where Z = 0. The operating state can then be considered as half voltage and half current controlled.

15

In Fig. 2b, the voltage source is replaced by the current source and the received impedance by the parallel impedance. From the load point of view, the connections of Figures 2a and 2b function completely identically if I = E / Z. Thus, when looking at the state of the element, either model can be used as needed.

20

If Z in Fig. 2b is very small compared to the load impedance Zi, the voltage Ui is almost exclusively determined by Z, since the load current is very small. The operation of the element is thus essentially voltage controlled. If, on the other hand, Z is very large with respect to Zz, the meaning of Z will be neglected, and the element will be substantially current-controlled.

25

In both models, therefore, a high value of Z means current control and a low value means voltage control. By contrast, the size of the sources is irrelevant to the control mode, so they can be thought of as zero. A zero-voltage power supply corresponds to a short-circuit and a zero-rated power supply to a power failure, so whichever model is used, the impedance seen by the element, i.e. the control impedance of 30 control states, is always Z (Fig. 2c).

The models of Figure 2 can also be used to describe a network of linear components in general, not just one source and one source impedance. According to the so-called Thevenin's theorem, any network containing resistances, capacitances, inductances, and voltage and current sources, when viewed between any two poles, can always be replaced by a single voltage source and one impedance junction coupling, termed i.

4 network Thevenin. Alternatively, a parallel connection between a power supply and an impedance (or admittance), also called a Norton compensation, can be used.

In order to keep Z high, in addition to the amplifier, a possible splitter filter must be designed to have a high impedance of 5, whereby first order filters are concerned. (Note that similar restrictions would also apply to conventional voltage control systems, provided that they respectively adhere to the requirement of low impedance seen by the element. In practice, however, this matter is hardly overlooked, where the actual control state of the element is often somewhere in between voltage and current control. however closer to 10 voltage controls.)

SMV-ORDINATED FLOWS

What matters most in terms of sound quality is the magnitude of the current components generated by the interfering electric motor motors em and e in the voice coil compared to the signal current flowing through the voice coil without these forces. In practice, the em magnitude of the motion SMV cannot be reduced without decreasing the sensitivity of the element. The inductive SMV e can be reduced to some extent by the design of the magnet and the short circuits, but not decisively. Thus, the only effective way to prevent or minimize the passage of these harmful SMV current components 20 is to maximize the electrical impedance seen by the speaker element, represented by Z in Figure 2c.

To examine this, one can imagine an additional voltage source generated in series with the speaker element, as in Figure 1, and then examine the relative current that this source generates in the voice coil of that element as well as in the voice coils of other elements operating in the same circuit.

GENERAL MISSIONS

30 The introduction of current control may have been constrained by the prejudice that the basic resonance of the bass element could not easily be suppressed by the so-called voltage control. with no electrical suppression. However, the necessary resonance damping is achieved by using a low mechanical Q value element, filling the housing with an effective damping agent, and using a suitable active or passive damping coupling to fine tune the bass-35 response.

According to a widespread but purely wishful thinking, the output impedance of the amplifier should be as low as possible in order for the membrane movement to be "controlled" and for the SMVs produced by the element to somehow disappear. The physical fact is that the motion of the diaphragm (more specifically the acceleration) can only be controlled by the current, and electromotive forces themselves can never be suppressed without suppressing the entire speaker, simply because the em = Blv law can never cease to hold, whatever impedances or circuits are used. . Instead, harmful SMV-derived currents are what can and must be attenuated, and only by careful current control design.

PROBLEM DESCRIPTION

10

Thus, current control requires that the output impedance of the amplifier must be high relative to the impedance of the speaker element. However, other circuits including the junction filter must not violate the current control principle by reducing the total impedance (Theven impedance) seen by the element. Thus, simple parallel-connected RC circuits cannot be used for response correction, since this causes the operating mode to move somewhere between the voltage control and the current control.

On the other hand, a typical cone-type loudspeaker needs current-controlled attenuation at high frequencies because the cone acts as a horn for short-range oscillations from the center, and the inductance of the voice coil does not cause low pass filtering, as occurs in normal use. Another reason that causes the frequency response to increase is the baffle step of the housing front panel. For short wavelengths, radiation will occur in semi-space and is therefore more effective than for large wavelengths at which sound is output from the speaker in all directions.

25 GENERAL PRINCIPLE OF THE INVENTION

The present invention describes a way of correcting the rising frequency response of a power-controlled speaker using the principle of Figure 3. The dual-voice coil speaker element 2 is coupled to a power distribution network 3 which allows both voice coils (4a, 4b) to normally operate at low frequency, but inactivates one or more coils at high frequencies where the other coil is still fully or partially active.

By "low frequencies" is meant such low frequencies (typically less than 100 Hz) that the need for attenuation in element 2 due to the reasons described above is still substantially negligible. By "high frequencies", by contrast, is meant such frequencies (typically a few kilohertz) that the full need for attenuation has already been substantially achieved.

• 6

DETAILED DESCRIPTION OF THE INVENTION

Figure 4 illustrates a simple embodiment of the invention. The voice coils 4a and 4b are connected in series, and coil 5, capacitor 6 and resistor 7 together form said power distribution network. At low frequencies, virtually all current supplied through terminals 8 and 9 passes through coil 5 and voice coils 4a and 4b. The current through the capacitor 6 is then negligible and the speaker operates at full control. At high frequencies, virtually all current passes through capacitor 6, resistor 7 and voice coil 4b, passing by voice coil 4a. Assuming the voice coils are identical, the sound level is thus damped by 6 dB. At intermediate frequencies (in the transition range), the current of the voice 10 wire 4a decreases gradually as the frequency increases, and the desired compensation characteristic can be achieved by optimizing inductance 5 and capacitance 6. If the desired attenuation is less than 6 dB, a resistor can be added.

The resistor 7 increases the impedance seen by the voice coil 4a by the frequency at which the inductance 5 and the kappa-15 dance 6 cancel each other out, thereby helping to keep the voice coil 4a in a current control state. Increasing the resistance 7 also increases the total impedance between terminals 8 and 9 at high frequencies, so the resistance 7 should be selected so that the maximum allowable total impedance is not exceeded.

In Figure 4, the impedance seen by the voice coil 4b is infinite, so this coil operates at all frequencies with current control without compromise. The impedance seen by the voice coil 4a is the impedance of the coupling of inductance 5, capacitance 6 and resistance 7, which can be made quite high at both low and high frequencies. With a resistance of, for example, three times the nominal impedance of the speech coil 4a, the impedance of said saw coupling remains acceptable even at frequencies where the impedances of the coil 5 and the capacitor 6 are relatively low. 25

It is important that the impedance seen by the voice coil 4a is high even at high frequencies, despite the fact that the coil does not conduct the signal current in this area. This is because the electromotive forces resulting from mutual inductance and coil movement are still present and should not be allowed to produce unwanted currents in the loop containing the voice coil 4a.

30

The 6 dB attenuation obtained by the coupling of Fig. 4 is often sufficient for small cone elements. However, with larger speakers, the front-axis sensitivity tends to rise steeper with the frequency, and thus more compensation may be needed. The solution is shown in Figure 5. The resistor 10 and the capacitor 11 form a new current path along which a portion of the signal stream 35 at high frequencies also bypasses the voice coil 4b, thereby providing the required additional attenuation. At the resistance values shown in Figure 5, the total attenuation at high frequencies is 10 dB (ignoring the inductance of the voice coil 4b). The desired total compensation curve can be achieved ♦ 7 by optimizing the resistance 10 and capacitance 11 together with the inductance 5 and capacitance 6, even with a circuit simulation program.

The inclusion of resistor 10 and capacitor 11 has little effect on the visual impedance of voice coil 4a, and the impedance seen on voice coil 4b remains relatively high even at high frequencies, since the sum of resistances 7 and 10 is practically multiple of the nominal impedance of voice coil 4b.

If the system does not use a treble element, it may be necessary to reduce the attenuation at the highest frequencies to widen the frequency range. This can be advantageously done e.g. by adding a coil and capacitor parallel resonant circuit in series with resistor 10 and setting the impedance peak of said resonant circuit to the highest reproducible frequencies. At the same time, the capacitance 11 can be substantially reduced. One significant advantage of current control is that, in the absence of the attenuation produced by the speech coil inductance, it becomes possible to use a single path system in many applications that would normally require the assistance of a treble element.

In Figures 4 and 5, exemplary values for circuit components are given in brackets, assuming 4 Ω: η voice coils. Capacitors should preferably be plastic insulated, but bipolar terminals 20 can be used where only cost is important. Coil 5 is not too big yet to be air-drawn. In the final design, the lossy self-inductance of each voice coil and the mutual inductance between the coils must be accounted for and modeled with appropriate matching circuits.

Figure 6 shows an alternative but non-ideal way of attenuating high frequencies in power-controlled loudspeakers. Here, a common single-speech coil element is fed using an RCL switch which can be tuned to perform a filter function similar to that obtainable with the circuits of Figures 4 and 5.

For comparison, we assume that Figures 4 and 6 use the same speaker element (speech coils in a series of 30). Then, at a given volume level, the electric motor forces induced on the voice coil 4a in Figure 4 are half of what is induced on the element in Figure 6. Further, the coefficient of power of coil 4a is half the power factor in Figure 6. the originating currents must be double that of the SMV currents passing through the element of Figure 6. Thus, the impedance 35 of the voice coil 4a seen in Figure 4 may be less than 1/4 of that seen in the element of Figure 6 to maintain the same sound quality. This advantage is, in practice, sufficient to make the twin runner method functionally more ideal than the one-reel option.

% • 8

The superiority of the dual mode method is particularly clear at high frequencies. With the given example values, the impedance seen by the voice coil 4a increases almost directly proportional to the frequency from about 1 kHz, while in Figure 6 the impedance seen by the element at high frequencies is equalized by the sum of the resistances R1 and R2.

COMPLEMENTARIES AND CLARIFICATIONS

The voice coils 4a and 4b need not be identical. By varying the number of turns of the coils, different attenuations can be realized. However, this is not necessarily very practical. The voice coils can also be connected in series within the element's internal structures, saving one outgoing switching wire. Instead of a single double coil element, several can be used. The voice coil 4a may be a series or parallel combination of coils of multiple double coil elements, and the voice coil 4b may be a combination of other coils of the same elements.

15

Parallel switching of voice coils can also be used with current control provided the elements are similar. When an interference SMV appears in one of the parallel coils, the interference current it produces in the other voice coil travels in the opposite direction, and therefore the audible SMV-derived effects are largely reversed, although the impedance 20 seen by a single voice coil per se is relatively small.

The power distribution network 3 may also include elements for other frequencies as well as distribution filter circuits, in the case of a multipath system. However, the principle of high source impedance limits the preferred split filter solutions mainly to first order filters. (The corresponding limitation of 25 would also apply to a voltage controlled system if the source impedance were to be minimized.) The response of element 2 may also require correction to attenuate basic resonance, which may cause the impedance seen by voice coils 4a and 4b to remain relatively low.

The same compensation can also be achieved by using two single-coil elements instead of one to 30 double coil elements. Under certain conditions, this can achieve even better elimination of SMV currents. However, the use of single-coil elements in this manner is more expensive and bulky and is not within the scope of the present invention.

Speaking of a two-spoke coil element refers to a structure in which the coils are on the same keel-35 frame or move the same vibrating membrane. Thus, e.g. the bass and dis-square unit of the coaxial element are different elements in this sense.

Claims (4)

A passive speaker system (1) preferably suitable for use with a power output amplifier, said system (1) comprising a speaker element (2) containing at least two speech coils, said speaker element having a first speech coil (4a) and a second voice coil (4b), and which system (1) comprises a circuit (3) for dividing signal current into said voice coils (4a, 4b); characterized in that: - a first voice coil (4a) and a second voice coil (4b) function in serial form such that at least part of the current which passes through the first voice coil (4a) also through the second voice coil (4b) - at a frequency of 2 kHz, the ratio between the current passing through the first voice coil (4a) and the current passing through the second voice coil (4b) is substantially less than at a frequency of 200 Hz - at some of frequencies between 1 kHz - 10 kHz, the current passing through the first voice coil (4a) is less than one third of the current passing through the second voice coil (4b) 20 - the impedance that each voice coil (4a, 4b) sees in the speaker system (1), without an external connection , is greater than 1.5 times the native coil resistance of the speech coil (4a, 4b) at all frequencies above 150 Hz which is primarily repeated in the system (1) through the element (2) in question and which frequencies the element (2) can substantially give . 25 Speaker system according to claim 1, characterized in that the power factors of the first voice coil (4a) and the second voice coil (4b) are different. Speaker system according to claim 1 or 2, characterized in that the first voice coil (4a) and the second voice coil (4b) have been connected in series in the internal constructions of the element (2), since only three connection coils are needed in the element. Speaker system according to claim 1, 2 or 3, characterized in that the first voice coil (4a) is a serial or parallel combination of the first coils of several double coil elements, and the second voice coil (4b) is a serial or parallel combination of the second coils of the same double coil element.