JBL Technical Note - Vol.1, No.8 电路原理图.pdf

Technical Notes Volume 1, Number 8 Characteristics of High-Frequency Compression Drivers 1 Introduction: When a professional sound contractor or acoustical consultant specifies a JBL 2445 driver over a JBL 2425 driver, he is paying 1.5 times more. For his extra expendi- ture, he is getting two very important advantages: about 2 dB more level in power handling and significantly lower second and third harmonic distortion for a given acousti- cal power output. It is JBLs contention that the reduced distortion is the greater of these two benefits, and we believe that careful systems designers will, after digesting the information contained in this Technical Note, be more inclined toward specifying the larger diameter drivers, especially when their relatively small price impact on the total system cost is taken into account. In this Technical Note, we will describe the operation of compression drivers in detail, focussing on those aspects which are of interest to consultants and systems designers. There is nothing mysterious about these devices, and they can be accurately described by a number of mathematical equations. These equations and the calculations made with them are included in the Appendices of this Technical Note. 2. Some Basics: A. Physical Description: Figure 1 shows a detailed drawing of a compression driver. A cutaway view is shown at A, and an end view of the phasing plug with diaphragm removed is shown at B. The significant physical and acoustical parameters are indicated on the drawings, and they are defined below: R E = voice coil resistance in ohms I = voice coil length in meters B = magnetic flux density in the gap in Teslas ST = area of annular slits in phasing plug in square meters SD = area of diaphragm in square meters ST/ SD = loading factor (normally equal to 0.1) M MS = mass of moving system (diaphragm-voice coil assembly) in kilograms Figure 1: Details of a Compression Driver A: Cross-Section View B: End View of Phasing Plug, Diaphragm Removed A compression driver differs fundamentally from other loudspeakers in that the diaphragm does not radiate directly into the air. It is placed fairly closely to a solid structure known as a phasing plug. The phasing plug has a number of openings in it, and the area of these open- ings is usually about one-tenth that of the diaphragm itself. Moving Mass MMS (diaphragm plus voice coil) Magnet .Magnetic Gap (flux density B) Phasing Plug Diaphragm Pole Piece Top Plate Voice Coil (length, I, in meters; resistance, RE, in ohms) Area of Annular Slits on Phasing Plug, ST, in square meters ST Area of Diaphragm, SD, in square meters When the diaphragm is actuated by current through the voice coil, high pressures are developed in the space between the diaphragm and phasing plug because of the relatively constricted openings in the phasing plug. Such high pressures are suitable for driving horns, since the high acoustical impedance at the throat of the driver is a good match for that encountered in the throat of the horn. B. Common Sizes of High-frequency Compression Drivers: For compression drivers intended for high frequency use, there are three main sizes: Small: 44 - 50 mm (1.75 - 2 in) diaphragm diameters. Examples: JBL 2425H/J: Altec 902,908: TAD TD-2001 ; EV DH1506; Renkus-Hesnz SSD1800; 1801; Emilar ECH175 Intermediate: 75 mm (3 in) diaphragm diameter. Examples: Altec 288,290,291; EV DH1012, DH2012; Emilar EC314, 320; Renkus-Heinz SSD 3300,3301 Large: 100 mm (4 in) diaphragm diameter. Examples: JBL2445J;TAD 4001 In general, the larger a compresssion driver, the greater input power it can handle, because of the larger voice coil and greater heat sinking. Contrary to what many people believe, the smaller drivers do not neces- sarily have more extended high frequency response than the larger models. Other pertinent characteristics of a compression driver are its efficiency (which translates directly into sensitivity), extension and smoothness of high frequency response, and distortion as a function of acoustical output. 3. Efficiency: The efficiency of a driver determines how much of the electrical input power is converted into acoustical output power. The theoretical maximum efficiency of a compression driver is 50%. While the parameters of most drivers are selected with this efficiency goal in mind, practical drivers fall short of the goal by two or three dB. Thus, we commonly see drivers with 25 or possible 30% efficiency. For example, the JBL 2445J has a midband efficiency of 30%, while the 2425H/J design has a mid- band efficiency of 25%. The shortfall in efficiency is for the most part quite negligible, and it is largely due to eddy current losses in the top plate and pole piece of the driver. Figure 2 shows the electrical equivalent circuit of a compression driver operating in its midband range. Maximum efficiency occurs when the value of radiation resistance, RE T, is made equal to the voice coil resistance, RE. The significance of radiation resistance is that it is the useful acoustical load reflected back into the electrical part of the system, representing a resistive load on the power amplifier in series with the voice coil resistance. See Appendix 1 for calculations of radiation resistance and efficiency. Figure 2: Equivalent Electrical Circuit, as seen at Terminals, of a Compression Driver at Mid-Frequencies 4. High-Frequency Response: A. Mass Break-point: All high-frequency drivers begin a roll-off in their out- put above what is called the mass break-point frequency. Obviously, the mass of the larger diaphragm assemblies is greater than that of the smaller devices, as is its voice coil resistance. However, the larger magnet structure and the increased length of wire in the voice coil provide more driving force, and this enables the larger driver to main- tain its mass break-point substantially in the same fre- quency range as the smaller driver. For most drivers intended for high quality sound or music reinforcement, the mass break-point is in the 3500 Hz range. Above that frequency, the response falls off at 6 dB/octave. In many applications, the fall-off can be ignored, since it may correspond, more or less, to accepted system equalization practice. But in the cases of studio monitoring and music reinforcement, the inherent roll-off of the driver will have to be compensated for. JBLs most recent passive dividing network designs have provision for such compensation. Inasmuch as the high- frequency portion of a system is always padded down relative to the low-frequency portion, there is power to spare, thus allowing the compensation to be made with- out additional power input. Figure 3 shows the equivalent circuit of the driver at high frequencies. Note that there are three reactive elements in the circuit. The most important of these three is the shunt capacitance, which governs the mass break- point. The element L E is the inductance of the voice coil, and it can result in another high-frequency break point in response. In JBLs drivers, an electrical shorted turn is plated directly onto the pole piece, and this effectively nulls out the voice coil inductance. The circuit element labeled LC E C is proportional to the volume of air between the diaphragm and the phasing plug. If this spacing is too large, then there will be another break point in the fre- quency response. Typically, in most driver designs, this Figure 3: Equivalent Electrical Circuit, as seen at Terminals, of a Compression Driver at High-Frequencies R E RET R E L E i-CEC C M E S RET 2 Figure 4: On-Axis Response of JBL 2440,2441 and TAD 4001 Drivers on a JBL 2350 Horn Figure 6: Response of JBL 2425 Driver on 25.4mm (1 in) PWT with 1 Milliwatt Input additional break-point is quite far out in frequency and is of no consequence. B. Secondary Resonances: Our neat theoretical model does little to explain the role of secondary resonances. These are controlled break- up modes of the diaphragm and surround structure which can be used to shape high frequency response in a beneficial way. Figure 4 shows the response of three 100 mm diaphragm drivers mounted on the same horn, a JBL 2350. The old JBL 2440 driver went out to about 9 kHz without showing any apparent mass break point. The reason for this was the role of a secondary resonance in the surround which kept the response rising. Recall that the 2440 had an aluminum diaphragm with a half-roll surround. The TAD 4001 driver has basically the same kind of surround treatment, but its beryllium material, because of its increased stiffness, exhibits the same characteristic moved out about an octave. The JBL 2441 driver, with its unique surround treatment, controls secondary resonances in a different way, producing an extended peak-free, but slightly rolled off, response. For further discussion please see the reference at the end of this Technical Note. 5. Measurement and Specification of Compression Drivers: In order to eliminate the variable loading effects of horns, compression drivers are usually measured, for the sake of comparisons, on a device known as a plane wave tube (PWT). An example is shown in Figure 5. The tube is cylindrical, with a measurement microphone placed close to the end where the driver is attached. Progressively, as sound is propagated down the tube, it encounters a care- fully tapered acoustical resistance wedge. The sound has Figure 5: Details of a Plane Wave Tube Tube Accoustical Damping Microphone been effectively attenuated by the time it reaches the end of the tube, and there is no reflection back toward the driver. The loading of the tube on the driver is thus quite smooth over a relatively wide frequency range. A typical PWT curve is shown in Figure 6 for a JBL 2425 driver. Note the location of the mass break point. When comparing data from different manufacturers, be careful in noting the diameter of the PWT used for the measurements. All JBL drivers are referred to a diameter of 25.4 mm, even if the actual measurement was made on a PWT of different diameter. Some manufacturers use a PWT with a diameter of 19 mm. This smaller diameter produces a level which is 2.5 dB greater than on a 25.4 mm tube, and that difference should be noted. See Appendix III for sample calculations involving plane wave tubes. Cross-sectional Area 6. Non-Linearities in Compression Drivers: High frequency distortion in compression drivers is basically a thermodynamic phenomenon. It results from the existance of high acoustical pressures at the diaphragm-phasing plug interface rather than from mechanical non-linearities in the moving system itself. It usually comes as a surprise to many sound contractors to find out just how little displacement there is in a com- pression driver operating under normal conditions. Referring to Frank Massas Acoustical Design Charts (Blakiston Company, Philadelphia, 1942), we calculate from chart 78 that a JBL 2445J driver, with 10 acoustic watts output at 1 kHz, will produce a peak diaphragm dis- placement of 0.2 mm, or 0.4 mm peak to peak. For each doubling of frequency, the displacement will drop in half, so it can be seen that normal usage results in quite small excursions of diaphragms in compression drivers. Over such small ranges of motion, the moving system itself is quite linear; that is, displacement is a direct function of the driving force. For normal acoustical outputs in the one to five watt range, it is clear that excursions are indeed quite small. Figure 7 shows the basis non-linearity of the volume- pressure relationship in gasses. At the bottom of the graph, we represent a linear, sinusoidal driving function as produced by the diaphragm. The pressure produced by this is non-linear for large changes in volume. In a horn- driver combination, such high pressures develop even Frequency in Hz Relative Response in dB dB (SPL) Frequency in Hz Impedance in ohms E3EZD TAD 4001 JBL 2440 Response Diaphragm Mass Breakpoint, IHM Impedance Curve Region of Secondary Resonances H F driver 3 Figure 7: Non-Linear Relationship Between Volume and Pressure in a Gas Figure 9A: 2445J and 2360. Distortion raised 20 dB. 6.3 Volts at 1 kHz Non-linear Output Input Volume greater distortion as the wave propagates down the horn. The more rapidly the horn flares outward, the lower the pressures become, and the lower the distortion will be. In order to compare high-frequency drivers, it is necessary to mount them both on the same horn, drive them to the same output level, and measure them at the same distance. If necessary, their frequency response curves may be equalized differently for their outputs to track at high frequencies. Using the test set-up shown in Figure 8, we compared the second and third harmonic distortion components of a JBL 2445/2360 combination and a JBL 2425/2327/2360 combination. Figure 8: Test Set-up for Driver Comparison 1.5m 2445 or 2425+2327 JBL 2360 Both Drivers Equalized for Same Response and Level at Microphone Both horn/driver combinations were driven to the same output level and were equalized with a 6 dB/ocatve boost above 3 kHz in order to maintain flat power response and flat axial response on the 2360 horn. The results of these measurements are shown in Figures 9 and 10. Distortion components in the graphs are raised 20 dB. In the graph of Figure 9A, we observe that the value of second harmonic distortion at 10 kHz is some 12 dB below the fundamental, and this corresponds to a distor- tion of 25%. In the graph of 10A, the second harmonic distortion at 10 kHz is some 8 dB below the fundamental, corresponding to 40% distortion. In Appendix IV, we present calculations of distortion. Note that the calculated values are in good agreement with the measured values. Frequency in kHz Figure 9B: 2445J and 2360. Distortion raised 20 dB. 2 Volts at 1 kHz Frequency in kHz The drive levels used in these measurements are typical of horn/driver combinations which have been equalized for flat power response. Second harmonic components rise fairly gradually and are apparent throughout the midband. For a given intensity, second harmonic distortion doubles with each doubling of driving frequency. Third harmonic components become quite apparent at higher drive levels, and they are most pronounced in the range above 5 kHz. For a given intensity, third har- monic distortion rises as the square of the driving frequency. We do not hear the harmonics of high frequencies, but the same non-linearities which give rise to second and third harmonics will also cause intermodulation dis- tortion of frequency combinations in midband, and these will be quite audible. JBL does not manufacture an intermediate size 75 mm (3 in) diaphragm driver, but it should take only a little extension of the theory verified here to show that the intermediate design would exhibit harmoinc distortion at some point between th