Talk:Audio power

From Wikipedia, the free encyclopedia

 It is requested that a photograph or photographs be included in this article to improve its quality.
The Free Image Search Tool (FIST) may be able to locate suitable images on Flickr and other web sites.

I deleted the term Class A in a paragraph with the example of the 100% efficient amplifier since a Class A amplifier can never be 100% efficient. I was temped to remove the 100% efficient term too since it really isn't needed for the example. —Preceding unsigned comment added by 69.253.87.22 (talk) 03:54, 4 May 2008 (UTC)

Pictures of advertised output powers on product boxes and the like.

I went to a dollar store, Kmart, and Target, intending to take pictures of misleading specifications, but found surprisingly few. I only found one system that listed "peak power", but it was merely twice the RMS power, which is arguably accurate (peak instantaneous power). I saw one other that listed RMS power as 20 W and "FTC power" as 12 W. Neither really warranted a picture, though. — Omegatron 02:21, 1 November 2007 (UTC)

Contents

[edit] Bad Science

No audio amplifier testing spcification used in the last 25 years specifies using a resistive load.
The inductive reactence of the speaker and the power factor are taken into consideration when testing.


Wikipedia should not declare time honored scientific measurements and standards to be of no value simply because you do not like how they are used in the audio equipment industry. Peak - Peak to Peak - Continous - average - momentary peak and many other power measurements have served science well for years.

RMS is Root Mean Squared and would not be equal to mean power with a sine wave. We must ask Root Mean Squared of what? Voice? Audio and Music all have different RMS values for the same amount of energy. RMS was created to match the thermal energy received by a resistor (Dummy Load) with an electrical measurement and reflects accurately the amount of signal power required to create that thermal energy. To make calorimeters and power meters coincide.
If you are trying to cook hot dogs with your 500 watt stereo these numbers are functional. The complex speech waveform and vastly complex music waveforms have very high peek to average ratios that make peek or impulse power formerly referred to as head room important to prevent clipping the peaks of signal when many frequencies come into phase momentarily durring a complex musical piece.
Digitally synthesized music is particularly bad at occasionally creating peaks that are up to 20 times the voltage of the average power. The peak capacity of an amplifier can be increased in many cases buy adding a larger capacitor to the power supply or lower internal resistance in the output stage semiconductors. Higher continuous RMS capacity requires increased continuous power supply capacity and output stage heat dissipation. A very high RMS Public address amplifier may perform poorly with digital music at a fraction of the loudness due to these differences.
While manufacturers may lie or publish false data and Music Power and PMPO ratings are not fully standardized the ratings are not usually meaningless. Speakers with better magnets and low resistance voice coils make more sound with the same electrical energy and may be factored into manufacturers rating in an attempt to give the shopper an idea of the comparable loudness one should expect.

i suggest removing all of the RMSes that are misnomers and using mean power. yeah it's a valid abbreviation, but it's still misleading. just mention it at the beginning and then use the real thing. - Omegatron 19:22, Dec 31, 2004 (UTC)

like it or lump it RMS is the term everyone in the buisness uses despite the fact it is a misnomer. Therefore i think it should stay. It is clearly stated at the begining of the article why the term is a misnomer and it is in quotes later on to try and reinforce this i think thats enough. Plugwash 19:37, 31 Dec 2004 (UTC)
Just because everyone's in the habit of doing something wrong doesn't mean we should encourage them. Besides, lots of manufacturers don't use it, because they know it's wrong. I believe the normal wikipedia response would be to explain the incorrect usage and then use the proper term everywhere else. Since the article is basically about "watts rms" i will change what i think should be changed and you can revert it if you want. - Omegatron 22:47, Dec 31, 2004 (UTC)
Eh, you're right. It's totally appropriate in this article because it's about marketing literature. - Omegatron 22:57, Dec 31, 2004 (UTC)
RMS is not a misnomer if you consider that (power)amps are normally rated for power output using a sine-wave at a nice middle-of-the-bandwidth frequency, in this case RMS means RMS, unless you want to nit pick over distortion and such caused by the amp itself. Secondly does anyone object to a section being added on peak-to-peak power (just a UK thing?) and PMPO being expanded to include peak music program/power output?--Pypex 23:01, 14 September 2005 (UTC)
No, that is simply mistaken, without any recourse to nit picking. If you were to actually measure the RMS power of an amplifier on a sinewave signal you would get a figure higher than the true power figure (the one you think you would get) which is actually the mean of the instantaneous power at all points over one cycle. Squaring the power is wrong, and gives more weight to the power at peak voltage than is correct. We square the voltage, to arrive at power, but that is a separate issue. A correct term, if you want to refer to RMS would be 'power based on RMS voltage measurement', but this is silly because there is no other valid way to measure power on an AC signal. The only correct term is 'mean sinewave power', since without a waveform specification you can chose any figure you want. I would object to any other method being described here, other than as in terms of explaining why it is not valid, unless it is backed by a standard. We have standards, notably an IEC one which specifies precise times for test duration and method. These should take priority over 'folklore'. Unless you can show me otherwise, I maintain that the term PMPO has nowhere ever been defined, and certainly not by a standards body. UK thing? As a professional audio measurement expert in the UK I can only say that the term PMPO is hated and despised by all fellow professionals. It's marketing speak, based only on a desire for high numbers by people trying to sell, should be exposed as such in any article on audio. --Lindosland 11:43, 21 March 2006 (UTC)
I do not like the statement that peak music power is usually twice the sinewave power. It confuses three issues - power supply sag, peak to mean ratio, and thermal limitation. Most power amplifiers will only give a little more true power short term, perhaps 10%, by virtue of having the full power supply voltage. Using peak power, which is twice the normal, is simply cheating on figures, and to be rejected. Thermal limitation is a complex business. Few modern amps are thermally limited in practice. Integrated amps chips have thermal protectin built in, but they are usually provided with sufficient heat sinking to prevent them cutting out (a sudden muting). Some high power amps do have thermal protection coupled with gain reduction. The common supposition that testing power amps at full output on sinewaves overstresses them is mistaken, since calculations show that maximum dissipation in the output devices of a class B amp occurs at about 60% of full output on sinewave drive, where the voltage across the devices is high. At 100% output the amp will run cooler, and at 100% flat out and clipping it will run cooler still (because the devices are almost switching, with no voltage across them for mush of the time). I believe there is actually a case for designing an amp with real music power capability, with say 10 times the steady sinewave power available on peaks, and sophisticated protection so that this was only available for handling brief music peaks (see Programme levels). Only with such a large ratio does 'peak music power' take on a useful meaning, but to my knowledge this has never been done. On modern compressed material it would have no advantage, but on properly recorded material it would be a revelation. --Lindosland 11:43, 21 March 2006 (UTC)

I think the external link, "An explanation of why RMS is a misnomer when applied to power" is not useful. The author of the article misidentifies the calculation for RMS: "...RMS (root mean square) power, would have to be defined as the square root of the time average of the square of the instantaneous power, since this is what 'RMS' means." RMS voltage is the square root of the time average of the square of the instantaneous voltage, and RMS power is the RMS voltage squared divided by the load resistance. The author is correct that the term "RMS power" is a bit of a misnomer and that the measurement he describes would be meaningless, but in doing so the article only confuses the point. Flagmichael (talk) 18:41, 10 February 2008 (UTC)

[edit] Incorrect equation at the end.

The last section claims that P_\mathrm{avg} = V_\mathrm{rms} \cdot I_\mathrm{rms} = \sqrt{\frac{1}{T}\int_{0}^{T} v^2(t)\, dt} \cdot \sqrt{\frac{1}{T}\int_{0}^{T} i^2(t)\, dt}

but it seems to me that it should be something like

P_\mathrm{avg} = \frac{1}{T}\int_{0}^{T} v(t)i(t)\, dt

Could someone show how this is derived? It looks like the magniture of a phasor, but that is only true of a sinusoidal system, not power in general. --njh 11:48, 16 May 2006 (UTC)

Well, the P_\mathrm{avg} = V_\mathrm{rms} \cdot I_\mathrm{rms} is correct (for a resistive load only), and the square root integrals are the equations for RMS, so it looks correct. Your version agrees with Power_(physics)#Average_electrical_power_for_AC and [1], though. Are these really just the same equation for the case of the resistive load? — Omegatron 14:27, 16 May 2006 (UTC)
I think that is correct, that purely resistive loads have ave power = rms voltage * rms current, but speakers are not very resistive (they have lots of inertia). --njh 02:45, 17 May 2006 (UTC)
I've updated it. — Omegatron 11:25, 18 May 2006 (UTC)

[edit] Example

"An amplifier labeled "500 W PMPO" but fitted with a 5-amp fuse can therefore deliver an average power of 5 A × 14.4 V × 60%, or about 43 watts."

Why are they assuming a 14.4 V supply? Is this a car amplifier? Cars have 12 V supplies. I don't know which type of amp this is assuming. — Omegatron 13:41, 17 May 2006 (UTC)
Car batteries are only nominally "12 volt". At rest, a healthy car battery is actually 12.6 volts. While the engine is running, the alternator always maintains the cars electrical system at a higher voltage (in order to recharge the battery as necessary), ranging from about 13.5 to 15 volts. 14.4 volts is very common - much lower and the charging doesn't work very well, much higher and devices on board may be damaged. human 00:06, 29 March 2007 (UTC)

[edit] H-bridge

I don't understand NathanHurst's edit. An H-bridge seems to be a switching network for running DC motors. I assume he means a bridged amplifier, which is a very similar idea, but which I have never heard called an "H bridge". Also, if car amps are commonly bridged, then you could get 12 Vpeak out of them instead of 6 Vpeak, as the example in the article describes. Also, I don't know what any of this has to do with his removal of the fused plug example. — Omegatron 13:05, 18 May 2006 (UTC)

H-bridge is a topology to get full supply p-p voltage. most car amps provide this. But the example was using 6v p-p and so was presumably not using h-bridge topology, instead a single class ab amp. I just clarified this. You could talk about a bridged amp, and then multiply everything by 2 or 4, but that would distract from the point. I removed the remark about the fuse as it is simply a non-sequetar - a fuse blows based on average power, pmpo, whatever it means, is presumably a short duration thing. Perhaps this particular model has a flywheel it charges up in quiet sections. (Yes, obviously it doesn't, but it is not _proof_ that PMPO is wrong) --njh 22:33, 18 May 2006 (UTC)
(A bridged amplifier is, indeed, an "H-bridge". It looks like an H if you draw the loudspeaker horizontally between two facing push-pull output stages running off the same supply.) — Omegatron 04:40, 26 July 2007 (UTC)

[edit] To RMS or not to RMS, that is the question!

Dear engineer types,

I appreciate your discussion and follow most of it. But I believe this academic level is useful only to BSEE types (design engineers, most often) and not nearly as useful to we mere hands on technicians and especially the laity of the general public who might read this. For most just comparing the RMS power number is useful and real. The ins & outs of analyzing complex waveforms goes beyond the need for this article. Feel free to publish all you want in your engineering forums or in formal technical papers, but I request that we keep Wikipedia articles like this one user friendly for the common man. Don't get me wrong, your discussions certainly have their place, but blowing away people with the sophistication of your knowledge does not help them understand the basic question being asked.

(In other words, RMS power is plenty good enough in practice if you are not a design engineer)

sjb

137.164.224.92 23:35, 12 January 2007 (UTC)

We aren't going to dumb down our articles by removing the technical aspects. We should make technical articles accessible, though.
This article does seem to suffer from some overly-zealous conspiracy theory debunking bias, though, implying that the peak output power is a meaningless number. It's perfectly useful for comparing amplifiers with each other in a general way. A guitar amp that outputs 10 W peak is obviously not as powerful as an amp that outputs 100 W peak. Comparing two 100 W amplifiers would require more detailed specifications, but depending on the expected use, it might be a perfectly adequate number.
And totalling the power outputs of 5 channels of a surround sound amplifier is not completely misleading, either, since the audio power is all going to the listeners' ears. I'm not sure about the acoustic implications though. How does the perceived sound intensity compare between a 2×50 W stereo loudspeaker system and a 5×20 W surround system? How do the signals sum in the ear, assuming that they are highly correlated to each other? Of course it depends on placement, too...
Since any normal audio signal is going to have a high peak-to-average ratio, expecting an amp to run at continuous sine wave power forever is unrealistic. To imply that an amp is of poor quality because it can't maintain its peak sine wave power infinitely is silly. An amp playing music in normal use (or even heavily clipping) will never approach its peak sine wave power. Amps are designed for normal music signals, and heatsinks and fans and such are tailored accordingly; to remove a realistic amount of heat per unit time. It would be wasteful to overdesign the amp for conditions it would never see except on a test bench. (And the amount of heat dissipated in the output transistors is not directly related to the output power anyway. A full-scale square wave could easily generate less heat inside the amp than a normal music signal, depending on the topology and efficiency of the amp in different conditions.)
The article's trying to talk about a few different things at once, and not really getting them right. — Omegatron 04:59, 26 July 2007 (UTC)

_____

I agree with sjb.

I could care less about just about all the complicated engineering stuff. I think there are better resources online and offline to find out what RMS means than to look at Wikipedia. Just keep it simple. Why is RMS important when buying speakers? And what does it translate to when I plug the speaker in?

Is that so hard Omegatrom? —Preceding unsigned comment added by 146.232.75.208 (talk) 15:12, 21 May 2008 (UTC)

M. Shane —Preceding unsigned comment added by 146.232.75.208 (talk) 15:13, 21 May 2008 (UTC)

[edit] RMS is not a misnomer

It is my understanding that Audio Output Power (RMS) as opposed to Audio Output Power (PMPO) is an indication of the system used to calculate the power produced by a particular amplifier so that users can compare one amplifier with another. I believe that the RMS label simply indicates to the user that the amplifier is capable of generating a stated amount of power in a standard (resistive) load for a considerable amount of time with a continuous sine wave input. All measurements of the sine wave refer to its RMS value rather than any arbitrary instantaneous value.

Saying that Audio Output Power is measured in RMS is much the same as saying that the frame size of a bicycle is 21.5 inches (c-t) Any cycle mechanic would interpret this as a measurement from the centre of the bottom bracket to the top of the top tube. (c-t) doesn't actually state that directly, but by convention this is what it means.

Surely an article such as this should explain conventions used in measurement rather than nit pick at the semantics of the words taken at their face value? TDuckmanton 21:17, 19 May 2007 (UTC)

Yes it is. There is no such thing as RMS power, only power - and its unit is the watt. If measuring power delivered by sinusoidal V or A waveform we use RMS units as measurements of V and A for calculation. If measuring the power of a square wave we use the peak value of the waveform.
Try this article on the subject, IMO it does a better job than the Wiki one <wink>.
http://sound.westhost.com/power.htm
RichardJ Christie 10:33, 6 September 2007 (UTC)
Of course there's such a thing as "RMS power"; it's just the RMS value of the power over time. The reason this is a misnomer is because that's not the measurement that people are actually using or referring to.
The actual measurement is correctly referred to as "average power", and is derived from an RMS voltage, hence the confusion. There is such a thing as RMS power, but no one uses it. — Omegatron 15:36, 11 October 2007 (UTC)
That's a bit of a specious argument, as no one uses it because there is no point in calculating it. We could have RMS value for almost any variable whatsoever if you wish. In that sense you are perfectly correct. It exists as much as RMS values of stock market fluctuations exist. RichardJ Christie 08:34, 17 October 2007 (UTC)

We could have RMS value for almost any variable whatsoever if you wish

Exactly.  :-) — Omegatron 04:45, 25 October 2007 (UTC)

[edit] PMPO and shared resources

It is important to remember that a 5 channel sound system that shares one power supply will only have the designated peek impulse power on a given channel and only if that channel is the only channel experiencing the peek impulse waveform. This is usually but not always the case. If the same monaural signal is applied to all channels, each could not achieve the the rated peek at the same time and the system should be de-rated for such use.

[edit] Quit Making Lying Statements

Audio power is usualy not about speakers and is usually measured in db.
There is no technicaly regognized measurement called sine power.
All sine wave measurements are for measuring sine waves - voice measurements are for measuring voice signals
Music program power measurements are for measuring the complex waveforms and content of music programming
- All major audio amplifiers use push pull circuitry the distortion percentages go down as power goes up until clipping is reached.
- Distortion levels are never speced at maximum power output.
- Measurement of harmonic and inter modulation distortion requires at least two frequencies to be present and can never be done with a single sine wave.
- There is no manufacturer in the US today quoting maximum power capacity at high distortion levels or at clipping.
Wikipedia should not make unsubstantiated allegations without references
—Preceding unsigned comment added by 71.113.213.199 (talk • contribs)

[edit] dBs and sensitivity

I'm not sure what is meant in the section "Power and loudness in the real world" where it states a higher sensitivity speaker gives greater loudness. The way that it is worded suggests that with fixed power input, one can get infinite power output by making the speaker arbitrarily more sensitive. I though the dBs would be talking about signal to noise ratio, in which case a speaker rated at 84dBs would sound just as loud as a speaker rated at 90dBs, its just that a greater portion of the sound energy in the 84dB speaker would be distorted waveform. Or is it talking about efficiency of electrical to sound energy conversion, in which case the dB would be negative and the more efficient speaker would sound louder? What does "dB sensitivity" mean? It should be described better in the article to help people like me. —Preceding unsigned comment added by 203.94.186.226 (talk) 07:59, 28 September 2007 (UTC)


The way that it is worded suggests that with fixed power input, one can get infinite power output by making the speaker arbitrarily more sensitive.

No. Think of it as the efficiency with which the speaker turns electrical power into sound power. The speaker can only do as much work as you give it energy to do, but no speaker will ever actually do that well; much of the electrical energy is wasted as heat instead of turned into sound. Sensitivity is used to compare the relative efficiencies. — Omegatron 15:42, 11 October 2007 (UTC)
Actually, efficiency is only somewhat related to "sensitivity", since sensitivity is measured at a single point in front of the cone, and directionality of the speaker plays into it as well. But they are closely related. — Omegatron 04:49, 25 October 2007 (UTC)

[edit] Speakers vs amps

One thing this article does poorly is differentiate between speaker and amp ratings. Both are given "RMS" and "PMPO" ratings, but the conditions are different. The peak power of an amps is directly limited by their voltage rails and the minimum impedance of the loudspeaker. It is impossible to have a higher peak instantaneous power than this (unless due to reactance?) But for loudspeakers, the peak instantaneous power is not as clearly defined, and has to do with destruction of the speaker. — Omegatron 04:48, 25 October 2007 (UTC)