FTC Power Ratings: An Optimistic View
Brian C. Wachner and Michael S. Robbins [President and Chief Engineer, respectively, of BGW Systems, Culver City, Calif. 90230, USA]
WIDE DISCREPANCIES in audio amplifier power claims have existed for years. Conservative manufacturers cringed when they saw brown goods manufacturers advertising "$199 complete" systems as having 350 watts of power. This credibility gap between power claims has finally been narrowed by the Federal Trade Commission, however, a fight by parties reluctant to change seems inevitable.
The so-called hi-fi manufacturers adopted a somewhat more meaningful set of standards in 1966-through the IHF [1]. However, this voluntary standard was antiquated with the coming of the solid-state era. Flagrant abuses of the term "continuous power" and the lack of meaningful standards has forced the Federal government to legislate a new standard for the protection of the consumer.
The ruling (see complete text on accompanying page) establishes a stringent set of guide lines and procedures to which all manufacturers must adhere if they wish to disclose power ratings. It is about time that all companies tested their amplifiers using the same set of standards and procedures, and that only realistic steady-state output levels be advertised.
The following power specifications are in the format required by the FTC under its new ruling:
BGW Systems Stereo/Mono Power Amplifier Model 750A
8-ohm Power Output: 200-watts average continuous power per channel; Power Band, 15 Hz to 15 kHz; Total Harmonic Distortion, 0.2%. 4-ohm Power Output: 300-watts average continuous power per channel; Power Band, 20 Hz to 10 kHz; Total Harmonic Distortion, 0.25%. Monaural Operation (Bridge-Connected Mode) 8-ohm Power Output: 600-watts average continuous power; Power Band, 20 Hz to 10 kHz; Total Harmonic Distortion, 0.25%. The most stringent aspect of the ruling, as applied to medium and large-size power amplifiers, is the pre-conditioning test and the five-minute, full-power test. Before measurements may be made at all, the unit must be subjected to a one-hour burn-in. This is accomplished with all channels driven to one-third rated output at the lowest impedance at which the manufacturer wishes to disclose power. The test signal used must be a continuous sine wave.
The power line voltage must be kept at 120 volts for domestic equipment, and the minimum ambient temperature must be 77° Fahrenheit. After one hour of continuous operation (the unit must have the ability to remain on without thermal cycling), the remainder of the tests are performed.
To electronic engineers, this pre-conditioning sequence appears to be very straightforward, however, some noted hi-fi writers are already thinking up ways around this part of the ruling [2,3]. The reason for their imaginative thinking is that a typical solid-state amplifier dissipates a tremendous amount of heat at the prescribed 331/3% power level. In fact, a class-B amplifier exhibits its lowest efficiency at approximately 40% of maximum output power.
The great majority of the medium-to-large power amplifiers on the market today will not pass the pre-conditioning test even at 8-ohms. Amplifiers with built-in cooling and those designed for 2-ohm use comprise the largest number of existing designs which may continue to carry their old power-rating figures now that the law has become effective.
======================
FTC TRADE REGULATION RULE ON AMPLIFIER POWER-OUTPUT SPECIFICATIONS
Power Output Claims for Amplifiers Utilized in Home Entertainment Products
Section 1. Scope.
(a) Except as provided in paragraph (b) of this section, this Rule shall apply whenever any power output (in watts or otherwise), power band or power frequency response, or distortion capability or characteristic is represented, either expressly or by implication, in connection with the advertising, sale, or offering for sale, in commerce as "commerce" is defined in the Federal Trade Commission Act, of sound power amplification equipment manufactured or sold for home entertainment purposes, such as, for example, radios, record and tape players, radio-phonograph and/or tape combinations, component audio amplifiers and the like.
(b) Representations shall be exempt from this Rule if all representations of performance characteristics referred to in paragraph (a) of this section clearly and conspicuously disclose a manufacturer's rated power output and that rated output does not exceed two (2) watts (per channel or total). (c) It is an unfair method of competition and an unfair or deceptive act or practice within the meaning of Section 5(a)(1) of the Federal Trade Commission Act [15 U.S.C. 45(a)(1)1 to violate any applicable provision of this Rule.
Section 2. Required disclosures.
Whenever any direct or indirect representation is made of the power output, power band or power frequency response, or distortion characteristics of sound power amplification equipment, the following disclosures shall be made clearly, conspicuously, and more prominently than any other representations or disclosures permitted under this Rule: (a) the manufacturer's rated minimum sine wave continuous average power output, in watts, per channel (if the equipment is designed to amplify two or more channels simultaneously); (I) for each load impedance required to be disclosed in paragraph (b) of this section, when measured with resistive load or loads equal to such (nominal) load impedance or impedances, and (Il) measured with all associated channels fully driven to rated per channel power; (b) the load impedance or impedances, in ohms, for which the manufacturer designs the equipment to be used by the consumer; (c) the manufacturer's rated power band or power frequency response, in hertz (Hz), for each rated power output required to be disclosed in paragraph (a)(1) of this section; and (d) the manufacturer's rated percentage of maximum total harmonic distortion at any power level from 250 mW to the rated power output for each such rated power output and its corresponding rated power band or power frequency response.
Section 3. Standard test conditions.
For purposes of performing the tests necessary to make the disclosures required under Section 2 of this Rule: (a) the power-line voltage shall be 120 volts AC (230 volts when the equipment is made for foreign sale or use, unless a different nameplate rating is permanently affixed to the product by the manufacturer, in which event the latter figure would control), RMS, using a sinusoidal wave containing less than 2 per cent total harmonic content. In the case of equipment designed for battery operation only, tests shall be made with the battery-power supply for which the particular equipment is designed and such test voltage must be disclosed under the required disclosures of Section 2 of this Rule. If capable of both AC and DC battery operation, testing shall be with AC line operation; (b) the AC power-line frequency for domestic equipment shall be 60 Hz, and 50 Hz for equipment made for foreign sale or use; (c) the amplifier shall be preconditioned by simultaneously operating all channels at one-third of rated power output for one hour using a sinusoidal wave at a frequency of 1,000 Hz; (d) the preconditioning and testing shall be in still air and an ambient temperature of at least 77° F (25° C); (e) rated power shall be obtainable at all frequencies within the rated power band without exceeding the rated maximum percentage of total harmonic distortion after input signals at said frequencies have been continuously applied at full rated power for not less than five (5) minutes at the amplifier's auxiliary input, or if not provided, at the phono input; (f) at all times during warm-up and testing tone, loudness-contour and other controls shall be preset for the flattest response.
Section 4. Optional disclosures.
Other operating characteristics and technical specifications not required in Section 2 of this Rule may be disclosed, provided: (a) that any other power output is rated by the manufacturer, is expressed in minimum watts per channel, and such power output representation(s) complies with the provisions of Section 2; except that if a peak or other instantaneous power rating, such as music power or peak power, is represented under this Section, the maximum percentage of total harmonic distortion [see Section 2(d)] may be disclosed only at such rated output; and provided further, (b) that all disclosures or representations made under this Section are less conspicuously and prominently made than the disclosures required in Section 2 of this Rule; and (c) the rating and testing methods or standards used in determining such representations are disclosed, and well known and generally recognized by the industry, at the time the representations or disclosures are made, are neither intended nor likely to deceive or confuse the consumers, and are not otherwise likely to frustrate the purpose of this Rule.
(NOTE 1: For the purpose of paragraph (b) of this section, optional disclosures will not be considered less prominent if they are either bold faced or are more than two-thirds the height of the disclosures required by Section 2.) (NOTE 2: Use of the asterisk in effecting any of the disclosures required by Section 2 and permitted by Section 4 of this Rule shall not be deemed conspicuous disclosure.) Section 5. Prohibited disclosures.
No performance characteristics to which this Rule applies shall be represented or disclosed if they are not obtainable as represented or disclosed when the equipment is operated by the consumer in the usual and normal manner without the use of extraneous aids.
Section 6. Liability for violation.
If the manufacturer or, in the case of foreign-made products, the importer or domestic sales representative of a foreign manufacturer of any product covered by this Rule furnishes the information required or permitted under this Rule, then any other seller of the product shall not be deemed to be in violation of Section 5 of this Rule due to his reliance upon or transmittal of the written representations of the manufacturer or importer if such seller has been furnished by the manufacturer, importer, or sales representative a written certification attesting to the accuracy of the representations to which this Rule applies, and, provided further, that such seller is without actual knowledge of the violation contained in said written certification.
Promulgated: May 3, 1974
Effective: November 4,1974
Charles A. Tobin By the Commission. Secretary
=======================
There are many good reasons why the law should be enforced without change. Without the new law, it would remain impossible for the audiophile to meaningfully compare specifications from one manufacturer to the next without guidance from a graduate EE. Many existing designs suffer from inadequate heat sinking and will be unreliable for that reason. The consumer will wind up with a higher quality product and a better investment because of the design conservatism now required for compliance.
Technically, there are very good reasons for keeping heat sink temperature down, and this is basically what is required by the pre-conditioning test. Semiconductors are very rugged devices when compared to tubes, however, they can easily be destroyed by allowing the silicon junctions to be exposed to temperatures approaching 200° Centigrade.
Power transistors, in fact, must be de-rated with temperature to insure safe operation. The curve in Fig. 1 illustrates the de-rating which applies to most silicon power transistors operated within their continuous capabilities. [4]. Thermal cycling is also a problem which causes power transistors to fail after extended operation. Any power transistor can be made to fail within a finite number of cycles by repeatedly heating it up to high temperature and cooling it off under power-applied conditions. Some semiconductor manufacturers have established data showing the number of thermal cycles to be expected from a device under repeated cycling conditions. Such a graph is presented for the type 2N3773 device (see Fig. 2), which consists of a 0.250 –x 0.250-in. silicon die mounted on a molybdenum or copper graphically in Fig. 7. By inspection, we see that maximum heating occurs at about 150 watts or 40% of maximum power.
Looking back at the heat-sink thermal resistance of 0.2° C. per watt with 208 watts being dissipated under pre-conditioning, we can expect a temperature rise of: (0.2° C/watt) (208 watts) = 41.6° C. The power band attainable from a given type of output transistor is directly related to its switching speed. It is well known that the faster we make a power device, the more fragile it becomes. If the designer is most interested in reliability and ruggedness, he will choose the slowest devices for the application. Generally speaking, amplifiers used for audio applications are required to deliver greater power below 500 Hz than in any other part of the power spectrum.
A chart of power spectrum density is shown in Fig. 8. All speaker systems have far greater power-handling capacity at low frequencies than they do at high frequencies, and most program material is compatible with this parameter [6]. It is, therefore, unnecessary to require a high-power amplifier to deliver its full rated power at frequencies above 10 or 12 kHz. A controlled power bandwidth, along with a sufficiently wide frequency response, is the most rational choice, although contrary to the audiophile's never ending search for specification improvements. Our listening tests on the same amplifier circuit with fast double- or triple-diffused output devices versus the slower single-diffused types reveal no audible differences. These single-diffused devices have the largest safe operating areas and can handle tremendous surges of power without failing. This peak power capability allows the practical design of amplifiers without the use of current limiting circuits.
The Model 500D is a slightly smaller version of the 750A and is built on the same chassis. Its power output into 8 ohms is 150 watts per channel with a maximum THD of 0.2% over a power bandwidth of 20 Hz to 20 kHz. In order to achieve this additional bandwidth, transistors having smaller dice with the same power dissipation rating are employed.
The 2N6262 used in the 500D measures 0.180 X 0.180 in. across the silicon die and can dissipate 150 watts. The mechanical construction is identical to the 2N3773. A slightly smaller power supply is used in the Model 500D to protect the faster power transistors.
By building an amplifier with enough muscle to pass the FTC preconditioning test, it is easy to overcome another fairly common problem. Most present high-power amplifiers (with greater than 100 watts) use some form of load line limiting to protect the output stage from short-circuit conditions. These limiters simply remove the drive to the output stage if excessive current flow exists. When this happens, the signal is grossly distorted, and spikes, rattles, chirps or buzzes may come from the speaker. Simple limiters act immediately upon application of the overload. More complex circuits have time delays of a few milliseconds before they are activated. A basic problem with any form of load line limiting is that speakers are reactive and can actually push power back into the amplifier and falsely turn on the limiters. The best solution is to design the output stage with enough strength to handle all the energy the power supply can deliver under short circuit conditions and to eliminate all forms of current limiting. Listening tests show dramatic improvements in the transient performance of an amplifier when the limiters are removed [7].
Fig. 6-Power-supply regulation and output-stage efficiency versus power
output for a typical Model 750A.
Fig. 7-Power dissipation versus power output for an actual Model 750A output
stage.
Fig. 8-Normalized average of peak energy levels of symphony orchestra. (From
Electro-Voice data sheet No. 535430.)
Fig. 9-Crowbar sensing and trigger circuit generates gate pulses for an
SCR which discharges the power supply.
Our approach is to use a fast-acting magnetic circuit breaker to disconnect the power from the amplifier under short circuit conditions. Line fuses are not used, as improper fusing could damage the amplifier. Additionally, a fail-safe SCR Crowbar circuit is incorporated to quickly shut down the amplifier if a malfunction occurs. This circuit (Fig. 9) samples the output of each channel and triggers a large, high-current SCR which discharges the power supply and shuts the unit off, if necessary.
The schematic of one complete channel of the 750A is shown in Fig. 10. All of the components shown are mounted on the heat-sink module. The input circuit consists of an LM318H, wide-bandwidth, operational amplifier, which can handle the total audio spectrum with minimal phase shift.
The LM318 has a slew rate 100 times that of the popular 741 type of amp found in many audio designs and 50-70 times higher than the 739 type used in other designs. Complementary transistors are used to source and sink the current required by the output stage pre-drivers and to provide the required voltage swing.
We have outlined here some of the approaches we have taken to produce rugged, reliable amplifier designs-approaches other designers may only now be forced to follow by the FTC ruling. While these are certainly not the only ways the FTC measurement standards can be met, whatever an amplifier designer does to meet and fulfill the regulations will benefit the consumer. At the very least, the FTC rules will narrow the credibility gap between power claims.
References:
1. I.H.F. Standards.
2. Feldman, L., "New FTC Ratings For Audio Amplifier Power," Radio-Electronics, Vol. 45, No. 11, November, 1974, pp. 61-63.
3. Klein, L., "Amplifier Power-Output Ratings: A New FTC Trade Regulation Rule," Stereo Review, Vol. 33, No. 5, November, 1974, pp. 79-83.
4. Solid-State Power Circuits. RCA Corp., Sommerville, N.J., 1971.
5. Wachner, B.G., "Fail-Safe Audio Amplifier Design," AUDIO, Vol. 57, No. 2, February, 1973, pp. 38-46.
6. Overly, J.P., "Energy Distribution in Music," IRE Transactions on Audio, Vol. AU-4, No. 5, Sept.-Oct., 1956, pp. 120123.
7. Carver, R., "A 700-watt Amplifier Design," AUDIO, Vol. 56, No. 2, Feb., 1972, pp. 24-34.
Fig. 10-Schematic diagram of one channel of the Model 750A.
==========================
An Alternative View
Robert H. Tucker [Director of Public Relations; Dynaco, Inc., Blackwood, N.J. 08012 ]
THERE IS NO denying the need for and the essential validity of the basic Rule provisions in establishing a universal reference base for evaluating this one important aspect of power amplifier design. Let us not, however, lose sight of other significant parameters, only some of which are readily reducible to cold type, which help to define the sonic differences which may still be apparent between amplifiers of nearly identical specifications. Too, the past proliferation of amplifier test clinics, each asserting the singular importance of a different specification (which just happened to be the one which the promoting brand handled so competently) should give pause, in their conflicting claims, to those who seek numbers ad absurdum.
That caution stated, on to the Rule. Dynaco is on record, as far back as 1970, as actively supporting only "rms" or average continuous power ratings in the hearings before the FTC, to the exclusion of any short term or "music power" figures, because of the consumer confusion they engender.
Further, two decades of Dynaco amplifiers have always been rated first and foremost by their average continuous power output. Thus, we were gratified with the early drafts of the Rule, when after years of FTC work, they were presented for industry comment. These were essentially identical to the Rule as promulgated, with the exception of the later addition of sections 3c and 3e, relating to preconditioning. These sections were instituted just prior to the last hearing dates, and apparently escaped much industry notice, despite the fact that any engineer would recognize that they radically altered the conditions for test. They will also have the effect of sharply increasing consumer prices with negligible audio benefit. Further, they will have the corollary result of reduced product information, and may thus engender equipment failure through inadvertent misapplication. This is widely at variance with the announced objective to provide consumers with meaningful comparisons of power output ratings, and to prevent fraud.
Mindful of the onus which attends challenges to a consumer-protection authority, and which has thus precluded cohesive, overt support from other amplifier manufacturers who also find serious fault with the preconditioning clauses of the Rule, Dynaco still does not intend to shrink from its responsibility to the consumer to attempt correction. Accordingly, as of this writing, Dynaco has filed a petition before the Federal Trade Commission for a delay in implementation of the Rule, and for a reopening of hearings to permit modification of the Rule. To date, no official action has been confirmed.
The purpose for seeking a rehearing is to obtain a true industry consensus on those aspects which, through industry lack of awareness and intransigence, were not responded to at the appropriate time. Only later were they found to require alteration or redesign of existing models. Naturally, those manufacturers who regularly institute model changes were less affected than those whose designs normally stay current-for extended periods. In the latter case, a de-rating of a model previously thought conservative by prior standards poses a serious credibility problem in a hobbyist field fraught with never-ending concern for "who is best?" As a practical matter, the alternative of subverting the efficacy of existing protective systems has been the unfortunate corrective approach in some instances. We deplore such a crass commercial solution, which can only adversely affect the consumer, yet we must recognize the economic implications which impel such tactics.
That the Rule was not corrected prior to its implementation should not now preclude revision, now that its impact is becoming visible. In this small industry, the arbitrarily severe preconditioning requirements were not recognized until they filtered back to the engineering departments, long after the hearings ended. Indeed, the FTC staff has indicated that the 1/3 power point was purely a convenient "middle ground" between one proposal for a 10% preconditioning level, and another recommendation for full power operation. That it closely approached the "worst case" operating condition was entirely accidental. It appears that these late insertions were simply overlooked by all of us, and thus no substantive data was submitted within the allotted time to encourage change. The FTC staff members have indicated that such a change could easily have been accomplished to a more sensible preconditioning base, if the facts were presented. It is the fault of the high-fidelity industry, not of the FTC, that this Rule was inappropriately drawn.
Mr. Wachner's thesis is not contestable on technical grounds, and if it were a fact that such (high-power) amplifiers were normally used with sustained signal levels which approximated average outputs of 20% or more of rated power, we believe his viewpoint would be far more defensible from the consumer's point of view: A ludicrous aspect of the whole argument, however, is that just those applications which justify such a stringent test condition musical instrument amplifiers, public address equipment, instrumentation and industrial use, etc. are excluded from the Rule's rating requirements because these are not "home entertainment purposes." It is a very real concern that a music amplifier be able to sustain an organ note, for instance, for its full duration. For that reason, "music power", or worse, "peak power" figures are meaningless bases of comparison, especially when they are established by connection of an external power supply.
What the Rule is now asking, though, is that the amplifier sustain such a note for one hour into any designated impedance at a level which is worse (in terms of heat output) than full-power operation, and then continue for five minutes at any frequency within the specified bandwidth at full power, without exceeding rated distortion. Quite commendable, but wholly unnecessary, for even the loudest-fi listener. The amplifier which can do this is to be complimented (and Mr. Wachner's is a good example over most of the audio band) but must you build a tank to attack every target, when a Jeep-mounted recoilless rifle can do the job with greater versatility? The desirability of high-power amplifiers is not predicated, in music listening, on operation at average power levels which are substantially greater than the average power levels employed with lower power amplifiers. Rather, [...]
(Audio magazine, Feb. 1975; various authors)
Also see:
Transient IM Distortion in Power Amplifiers (Feb. 1975)
Buying Watts and Other Things (Feb. 1973)
= = = =