Radio Broadcast (May 1928-Apr 1929)

LOUD SPEAKERS— A DEBATE Dynamic Type Versus gnetic Type <*©('! AT THE end of an average day, the wastepaper baskets in the Editorial Office of Radio Broadcast are almost filled with press releases written about every known kind of a radio equipment. It is a rare release which contains information which can be used in this magazine. The truth is, that it is a rare release which has any information of use to anyone. Occasionally, however, the mail clerk brings one which promises a great deal. The following statements are taken from a release dated August 29th, titled "Little-known Facts About Well-known Cone Speakers," and signed by the Director of Research of a cone-type loud speaker corporation. The replies to these statements were made at our request by an engineer who has no connection with the loudspeaker industry, but who, in our opinion, is equipped with as much unbiased data as any engineer in the country. His interest, as shown in his statements below, is the interest, of engineering truth. We believe this little debate may give our readers more information than a much longer article written in conventional style and form. The release: "It is freely claimed that the efficiency of the usual rocker-type driving unit for cone speakers is from 4 to 7 per cent., and that its power rating is strictly limited to that for which it is set. The Western Electric unit, for example, is set for 1 watt, the BBL for 2.5. watts, while the Stevens unit will take 3 watts and still maintain an efficiency of 7 per cent., although at 0.5 watt its efficiency is only 2 per cent." Our engineer replies: "The so-called wattage that can be put into a speaker varies with the frequency. The Western Electric cone handles about 1 5 watts at the lower end of its response without striking the pole pieces. Because of its rather greater efficiency than most cones this gives as much sound output as a cone 3 TU less efficient capable of handling 3 watts." We should like to put in a word here. The average listener need not worry about the power-handling ability of his loud speaker. What difference would it make if his speaker could handle 100 watts of energy without smoking, blasting, or hitting the pole pieces? What he should be interested in is the amount of sound he gets out of the speaker with a given elec trical input. We have stated already that, if a speaker could be made 10 TU more efficient than the best of our present devices, we could obtain sufficient volume of sound from it without using expensive power apparatus. The fact that the efficiency of the unit which the release mentions differs at various power levels indicates, according to the engineer, "con RADIO folk feel keenly that, out of the mass of nearaccurate statements and downright mis-statements about topics which may at the moment he uppermost , there must be some generally accepted facts to which they may cling. This is usually so, but the search for simple and real truth is not easy. In the comparative merits of different types of loud speakers, the flow of inaccuracies has recently been especially strong. The short discussion here of the comparative merits of the magnetic vs. the movingcoil loud speaker is as technically accurate as we know how to make it, without ascending to the rarefied atmosphere of higher physics. We hope it proves interesting and valuable to our readers. — The Editor. siderable lack of linearity in response. When the ratio of input to output is not linear, the harmonics are very bad." EFFICIENCY OF DYNAMIC UNITS CPEAKING of moving-coil loud speakers, the ^ release states, "The very best design available develops an electrical efficiency of only 50 per cent., and this figure is attained only by employing a push-pull transformer which has been designed specially and constructed of very expensive material. "The electrical efficiency of some of the most popular makes of electrodynamic speakers is a FIG. 1. FREQUENCY RESPONSE CURVE OF A TYPICAL MOVING-COIL SPEAKER 109 scant 30 per cent., when actuated by push-pull amplifiers. When, however, they are actuated by ordinary transformers, and from a single power tube, the efficiency drops to as low as 20 per cent, with some distortion. In some of the cheaper makes, the efficiency is further impaired by mechanical losses due to the cone suspension." Our engineer says: "The push-pull element has nothing to do with it. Equal efficiencies ions can be obtained with either single or pushpull amplifiers and transformers. An efficiency of 30 per cent., is probably high for even the best of the dynamic speakers. Any speaker worked from improper conditions will perform unfavorably." The release: "These figures, let it be understood, are electrical only, and to obtain the overall efficiency, we must subtract the cone losses. It is quite true that in cones of the same size — the 18-inch size, for instance — the efficiency of a cone actuated by a dynamic unit is greater than that of one actuated by an electromagnetic unit. This is due to the greater power available for flexing the material, provided, of course, that the source is capable of furnishing the energy. On the other hand, it is well known that the amplitude of the apex of the cone varies inversely as the square of its diameter, plus a logarithmic constant for air slippage." POWER VS. EFFICIENCY '"PHE engineer: "The power available has *■ nothing to do with efficiency; the losses are what reduce it. Dynamic speakers may be made to have less eddy current and hysteresis losses than the rocking-armature units. In the latter type the armature itself is saturated rather heavily. In addition, mechanical masses are supported better and distributed so that impedances, both mechanical and the resultant electrical, are more uniform and permit of better matching in the dynamic type. "This is true for a constant radiating power at a particular frequency, and provided the whole moves as a unit (piston) without interference between the front and back faces. This latter is not the case for the Western Electric type of cone, and is true for the dynamic only when the baffleboard is sufficiently large (diameter), massive and damped." The release: "In practice this works out so that if an 18-inch cone requires 0.008-inch amplitude down to 100 cycles,