Published in June 2006

Unraveling the Mystery
By Doug Wilkens

Creating a credible pro power amp output ratings standard is crucial.


Amplifier under test using low distortion audio signal generator, true RMS meter, oscilloscope and load bank.

     The professional audio industry does not have its own standard(s) for rating power amplifier output.
     Does this statement surprise you? You would think that the professional audio industry would be the group that is most interested in having a meaningful and precise method of comparison between amplifiers in the marketplace today.
     What standards or federal regulations that do exist have been “inherited” from the consumer audio industry and were developed because of outrageous claims made by various consumer manufacturers regarding peak power, peak music power, instantaneous peak music power and other similar phrases. The FTC rule specifically states that the rule is for “sound power amplification equipment manufactured or sold for home entertainment purposes.”
     This report is the result of a number of questions and ensuing tests that were done over the past couple of years by members of the Live Audio Board forum on www.prosoundweb. com and also comments and questions from contractors and consultants associated with the Syn-Aud-Con “listserv” bulletin board.

Output Rating History
     To have a better understanding of the situation, a review of the history of power amplifier output ratings is in order.
     Naturally, a complete listing of every attempt back to the earliest days of audio power amplifiers to make sense of amplifier measurements and ratings would be too much to include here, but some highlights will help give an idea of the major accomplishments in this field.
     Perhaps the first major attempt to make sense of amplifier ratings was from the Radio-Electronics-Television Manufacturers Association (RETMA) in conjunction with the Electronic Industries Association (EIA) in 1949, when they published a standard, “Engineering Specifications for Amplifiers for Sound Equipment” (EIA/RETMA SE-104). Basically, this was a listing of the various specifications required for an amplifier, including output power levels at a distortion of 5%, gain, frequency response, as well as a number of other electronic and physical measurements and ratings.
     In 1958, the Institute of High Fidelity (IHF) expanded on the RETMA Standard, with an enhanced set of conditions and ratings (IHF-A-200). This standard was updated with a more comprehensive document in 1966 (IHF-A-201).
     ANSI (American National Standards Institute) did some work in the ’70s to develop a standard for “Measuring Audio Amplifier Power Output Rating for Institutional Audio Visual Equipment,” but the standard never went into effect. ANSI apparently decided to certify the standards developed by the EIA.
     The Deutsches Institut für Normung (DIN) in the past also developed standards for hi-fi equipment that addressed amplifier power output ratings. The most recent standard for power amplifier ratings is 61305-3:1995 and is available at www.nor mung.din.de. Thanks to international agreements, most of the standards that were developed originally by different standards organizations in various countries are now part of the International Electrotechnical Commission (IEC), so you will find this same standard available as IEC Standard
IEC 61305-3.
     In the ensuing years, the EIA continued to work on standards for testing/measurement of audio amplifiers, including multi-channel amplifiers, and published an interim Standard in 1981 to try to deal with wildly inflated claims about amplifier power ratings in the consumer market (the instantaneous peak music power, et al, mentioned earlier). The EIA documents are available for purchase at http://global.ihs.com. Do a search at “Standard Test Methods of Measurement for Audio Amplifiers” to see the list of documents available for purchase.

Government Got Involved
     Unfortunately, because the consumer-oriented manufacturers continued down their merry path of inflated claims, the government stepped in and finally made it the law for manufacturers to adhere to a clearly defined method of testing and rating audio power amplifiers. It is not only the audio industry that has been forced to accept government intervention into ratings and specifications: Many other industries have been faced with the same situation because industry would not self-regulate!
     So, in 1974, the Federal Trade Commission (FTC) came out with its Trade Regulation
Rule regarding “Power Output Claims for Amplifiers Utilized in Home Entertainment Products.” This FTC rule has been amended over the years, with the most recent version available
for download at the National Archives website, www.access.gpo.gov/nara/cfr/ waisidx_03/
16cfr432_03.html.
     Although various professional amplifier manufacturers have participated in comments and suggestions to the EIA and FTC, it seems strange that, thus far, the professional audio industry (and this certainly is the industry that should be most interested in a professional standard) has not worked with the only organization capable of injecting professionalism into power amplifier ratings for our industry: the Audio Engineering Society (AES). We all should support any effort made to create an amplifier power rating standard that relates to professional products.
     In the course of doing research for this discussion, many amplifier manufacturers were contacted for comments and specifics about how they test and rate their professional power amplifiers. Strangely, it appears that many of them are happy with the status quo, that no standard exists for professional power amplifier testing/rating.

Why?
     Why the apparent lack of interest of the professional audio industry? Why not have a professional amplifier ratings standard?
     It certainly makes sense: If there is a standard, consultants, contractors and customers can compare products easily. Everyone would be better served by having a standard that clearly defines exactly what the specifications are for power amplifiers, the exact testing methods, the parameters for the tests, etc. If it were easy, it probably already would have happened, but some investigation will reveal all is not so simple.
     A large number of power amplifiers now use switch mode power supplies, pulse-width modulation (PWM) techniques or other methods of “digital” power amplification, and many of the manufacturers of these amplifiers do not want their units tested in the same manner as power amplifiers that have traditional linear power supplies and/or analog amplifier circuitry for the output stage.
     Most of these manufacturers would prefer to have a tone-burst signal used as the source material to drive an amplifier under test, rather than continuous sine-wave or pink-noise source signals.
     Using sine waves or pink noise as source material, traditional amplifier designs typically are much more capable of producing continuous power output, compared with many amplifiers that use switch mode power supplies and/or PWM (or switch-mode/digital) output circuitry. In many applications, this does not matter because typical music program material is not composed of single sine waves, or even the equivalent of pink noise (with a 10dB crest factor).

Chart 3

Test Procedure
     An example of how manufacturers of PWM amplifiers test their products is in the following description of the test procedure recommended by Lab.gruppen, a Swedish manufacturer of power amplifiers:
     This document aims to simply point out the main things to be aware of when putting the amplifiers through their paces. First a “Quick-guide” list is presented, followed by more thorough explanations for the individual bullets.
• Max.-output voltage bench testing quick guide:
  1. Use a 1kHz burst signal (33.3ms ON/66.6ms OFF). A standard continuous test signal will
      possibly activate the Automatic Fuse Saver (AFS), which limits the max.-output voltage.
  2. At high-frequencies (more than 10kHz), use a burst signal to avoid engaging the Very High
      Frequency (VHF) protection circuit. Otherwise, you will not be able to verify the true max.-
      output voltage at VHFs.
  3. A 22kHz (min 24dB/oct.) low pass filter should be placed in series with the audio analyzer to
      avoid measuring above audible HF-content in the output signal.
  4. Be aware of the gain-switch setting (recommended 32dB).
  5. Set the MLS switches to 0dB; otherwise, max.-output voltage will not be achieved.
  6. Notice the output polarity when measuring.
  7. Make sure to use balanced connections and cables throughout in the test setup.
• Max.-output voltage bench testing extended explanations:
   When conducting a bench test, there are seven main areas to be cognizant of:
  1. If a sinusoidal signal is used to determine maximum power output, it is most likely that a protection circuit called the Automatic Fuse Saver (AFS) will engage. In fact, the only
instances users have ever reported it engaging are during tests with sine wave generators. The AFS is designed to detect and then limit excessive mains current draw to avoid blowing mains fuses during performance.
     The steady high draw that a full-scale test tone generates is sufficient to engage the protection, which subsequently limits the mains current draw to a safe level and hence a reduced output is seen. This has never been recorded during a music program test. To measure max.-output power, use an oscilloscope together with a burst signal. Recommended is a 1kHz signal with 33.3ms on-time followed by 66.6ms off-time—and so on. This will give you the true max.-output power.
  2. All Lab.gruppen amplifiers are equipped with a VHF (Very High Frequency) protection circuit. This circuit detects and mutes the amplifier if there is too much high frequency energy on the amplifier output. This is a problem that often is caused by self-oscillation in the audio distribution system before the amplifier. During normal operation, with music material, the amplifier will not experience continuous VHF energy, hence causing the amplifier going into protection. However, the amplifier will still amplify transient peak-VHF signals in the music at maximum output power.
     The “attack” frequency at which the protection engages against such continuous sine waves at maximum voltage is model dependent and falls between 10kHz and 20kHz. Above the “attack” frequency, the VHF sensitivity threshold point increases at 6dB/octave. Chart 1 shows the VHF protection operation area.
     There is a delay—or hold time—before VHF-protection sets in, which is dependent on both amplitude and frequency. Higher frequency and/or higher voltage results in a shorter hold-time to trick the VHF protection. This makes it possible for the amplifier to amplify all transients and high frequency bursts that exist in normal music while still protecting the HF drivers from destruction if too much non-musical high frequency energy occurs at the input. The use of an oscilloscope together with a burst signal will stop the VHF protection engaging, and thereby enable the measurement of the amplifier at maximum output voltage across the entire audio frequency spectrum.
  3. Most Lab.gruppen amplifiers have switch mode circuits in the power supply and/or in the amplifier stage. Some fractions of the switching signals can be found on the output of the amplifier. To prevent this HF content from interfering with the meaningful and desired audio measurements, a 22kHz (min 24dB/oct.) low pass filter should always be connected in series with the audio analyzer. Signals above 22kHz are handled according to EMC regulations, and not considered as music-signal.
  4. The Lab.gruppen amplifiers are extremely flexible when integrating into any system. There is a multiple position gain switch on the rear panel that allows the user to define the input to output gain of the amplifier, from 41dB to 20dB. During a bench test, it is critical to be aware of the gain structure of the amplifier. It is suggested that this is set to unity on all amps (32dB).
  5. Many Lab.gruppen models feature the Matching Load System (MLS). Via a set of switches on the rear panel of the amplifier, the user is able to control power transfer into specific speaker loads. Prior to bench testing, these switches should be set in their maximum settings (0dB); otherwise, maximum power output may not be obtained.
  6. Channel B is always polarity reversed on the input, but polarity compensated by feeding the minus pin on the channel B output with the output voltage (output on pin 1 in opposite phase). Channel A is connected in normal polarity mode. By having channel A and B operating in opposite polarity, the energy storage in the power supply is more efficient. This is significant for signals below 100Hz and improves power bandwidth.
  7. Be sure to use balanced inputs on all measurement equipment (also oscilloscope probes).

Comparison Requirements
     To effectively compare amplifier power output, you must have standardized source material and a standardized load, with standardized test parameters, for a specified time period, in order to have a clear comparison. The EIA and FTC test procedure does just that, but it is evident that PWM amplifiers will not perform in the manner that realistically demonstrates their performance in real-world applications, if the EIA/FTC testing is used.
     Additionally, the bandwidth of the test system is a major consideration when comparing traditional analog amplifiers with PWM amplifiers, due to noise outside the audio spectrum (PWM/digital amplifiers typically require a test/measurement system that has a very serious filter to remove signals above/outside the audio spectrum). For additional reading about this topic, go to Audio Precision’s website (www.audioprecision.com) and look for “White Paper: Measuring Switch-mode Power Amplifiers” by Bruce Hofer.

Additional Complication
     An additional complication of power amplifier testing is that the real-world load on an amplifier is quite different from a non-inductive resistor that often is used as an amplifier load in a test laboratory. In reality, the load typically is a loudspeaker system that has impedance that changes significantly over the audio spectrum. When you look at a typical loudspeaker impedance curve, you will see a significant deviation from the “nominal” impedance rating of the device.
     Because of this fact, there have been several attempts over the years to define a standardized test load that emulates the type of load that a loudspeaker would provide to an amplifier, but it seems without general acceptance. Most “lab” amplifier testing is done with non-inductive load resistors. It is difficult to obtain high-power non-inductive load resistors, so most high-power amplifiers are tested with wire-wound resistors of one type or another, including some rather creative loads using hot water heater elements, electric stove elements, pizza ovens, etc. Several examples of this type of creative effort can be found at
http://srforums.prosoundweb.com/index.php/m/0/5094/16/0.
     Many integrators and consultants would like to see a variety of different methods of measurement, testing and ratings used for power amplifiers. For example, some have stated that they would like to see a chart showing the current and voltage output of an amplifier, as opposed to a traditional wattage rating. This (and many other types of data) can be helpful, but most users have become accustomed to being able to compare amplifiers by their output wattage ratings because it is easy.
     Crown is one of the few manufacturers that addresses this issue by including a chart of current/voltage (V/I) output ranges in its description of operating conditions. Chart 2 is one sample of many V/I charts that are included in Crown’s description of operating conditions of the CTs series of amplifiers.

Pink Noise
     In addition to the tests performed at single frequencies (as well as over the entire operating bandwidth of the amplifier), use of pink-noise source signal also has been promoted (because pink noise more closely resembles actual program material with regard to average and peak signal content).
     If our industry decided that rating a power amplifier by output current and voltage was important, and that this information should be part of the specifications, it still would be appropriate to include traditional wattage ratings as well—and Crown does exactly that.
     The measured power output of an amplifier can be significantly higher if higher distortion percentages are permitted, so the standard would have to include what the maximum allowable distortion would be for the maximum output measurements. Shifting from 0.1% to 1% makes a considerable difference in output power ratings; look at various specifications in which the manufacturer provides both ratings to get an idea of the difference!
     To add to the confusion of many users, the AC mains power requirements of an amplifier often are listed in the specifications and frequently are also printed on the rear of the unit. If the amplifier is operating at full output power, certainly the power requirements for the amplifier will be much higher (sometimes significantly) than the output power rating of the amplifier due to the efficiency of the amplifier.
     Fortunately, a few manufacturers go to the trouble of listing the idle power consumption, the 1/8th power (or 1/3rd power) consumption and the maximum power consumption. Unfortunately, non-technical users often mistake the power requirements to be the output power rating; just look at some of the classifieds in your local paper.
     It is an interesting exercise to analyze what information is contained on manufacturers’ amplifier specification sheets right now. QSC includes output ratings according to both EIA and FTC requirements.
     Chart 3 indicates the specifications of the QSC Powerlight Series of amplifiers, showing the EIA and FTC specifications for 8 ohms, 4 ohms and 2 ohms, as well as the Distortion, Current Consumption and other important technical criteria that are useful to consultants and contractors for making comparisons with similar products.
     Yamaha also gives complete specifications, but adds the power consumption at idle (no signal), as well as at maximum output, which is quite helpful. Note that, although QSC uses current consumption, Yamaha uses power consumption. QSC clearly states it is using EIA and FTC standards for the power ratings; Yamaha does not state this, even though it appears the company actually is using the ratings as specified by the EIA and FTC. Yamaha also includes a tone-burst rating (1kHz 20ms), which provides an indication of dynamic headroom. QSC shows the THD distortion at full-rated power; Yamaha shows THD and IMD at half-rated power. Chart 4 shows excerpts from the specifications for the Yamaha PC9500N/PC4800N amplifiers.

‘Unraveling the Mystery…’
     The title mentioned “unraveling the mystery…” but by now it may appear the mystery has deepened. To really remove the mystery, a procedure with the appropriate measurement techniques and agreed-to data format must be developed and turned into a standard.
     One hopes that, in some small measure, this discussion will help spur activity on the part of consultants, integrators, dealers, manufacturers, sound-reinforcement companies and end users in our industry to work together to develop a credible standard for professional power amplifier ratings.


Doug Wilkens has been involved with the audio industry all of his life, including many years as an engineer for major US sound contracting firms, as well as marketing, sales and management positions for a number of audio manufacturers.