What is the most powerful audio output tube?

"RMS" is short for "root means square", or the square-root of the "average" of the values squared. Power is voltage-squared times resistance. The power produced by a voltage source is the effective voltage times the resistance. For DC, the effective voltage is the same as the voltage. In the case of a sine wave, the "effective" voltage is the peak to peak voltage divided by SQRT(2). For other wave shapes the difference between the effective (or RMS) voltage and the peak voltage is different but the RMS voltage is the effective voltage.

Now go back to power, (RMS)Voltage * (RMS) voltage / resistance = power. It makes no sense to say RMS * RMS = RMS. It's not, it's just "power". RMS has no physical meaning for power.

Thanks for the useful info. Is there any method to determine the ACTUAL power output from an amplifier? Yes, I know that with a sine wave, there will be bursts of power, with most being from the bass or low end. But there has to be a way to state the actual Maximum power any amp can produce. Not only to know the abilities of an amp, but also to choose a speaker(s) that can handle the maximum power.

If this is all based on mathematics, I am a not very good with math... (I'm being honest about that).

Furthermore, if you calculate the power during a cycle of the sine wave voltage input and plot it instantaneously you get a new sine wave (well its starting value may make it a cosine wave - but same difference) at twice the frequency, and not crossing the zero line. So the average power is the *average* value of that sine wave, which is half its peak, and the same answer as expression above gives using RMS voltage. If you measured the RMS amplitude instead of the average amplitude of that new sine wave you would get totally the wrong answer. So there *is* such a thing as RMS power but it is a physically meaningless and useless figure.

--

Roger Hayter

"Maximum" power is another misnomer, in this sense anyway. Since the problem is heat, you want to find the average power over some, perhaps small, interval. There are a few ways of measuring this. One can put the speakers in water and measure the temperature rise. ;-) Or, measure the voltage and current waveforms and multiply the two and average over the interval of interest.

Simple arithmetic is all that's needed, at least at this level.

Sure. You can take the RMS value of the national debt or what Hillary spent on losing but the result is pretty meaningless.

resistance."

You oopsed that one. It is current squared times resistance, or voltage squ ared divided by resistance. I say oopsed because I am pretty sure you know that and if you could edit posts here you probably would have. Not as bad a s some of my gaffs though, like typing "now" instead of "not" or vice versa . Other than that you wrote a pretty good explanation of it.

However there are other concerns having to do with language that don't real ly fit with real math. Like unemployment, it is NOT 5 %, "inflation" which is supposed to be the inverse of the value loss of the currency is not calc ulated correctly, that's just what they call it. Remember "I can't eat an ipad" ?

They should have probably called it something else but I wouldn't trust any convention of today to do it, like not having a Constitutional convention. But a nutritionalist is now a nutritionist, a croiminologist is now a crim inalist, the fovea in your eye is now called the macula.

If they could simply say that now what was called "RMS power" is now called "average power" or "real power" or something like that it would suit me ju st fine. But once the people who decide such things get busy they get too many bright ideas. They might decide on a slightly different formula at the behest of the manufacturers.

In other words, sometimes you just have to go with it. Sine wave peak volta ge at a certain distortion level also specified times 0.707 times itself di vided by resistance is what they call RMS power. Even if they rename it rig ht you can't go back through a hundred years of owner manuals and change th em all.

And remember the formula only applies to two channel equipment or the front two channels of surround amps etc. They can do whatever they want with the rest. That means pro audio as well, I have seen pro amps that say "450 wat ts + 450 watts that had about 55 volt power rails. You don't even get that into four ohms, and with only one pair of outputs per channel I doubt they will do too well on a two ohm load.

The standards also do not apply to car audio. Even with separate power supp lies and 30 amp fuses to each the limit is around 360 watts per at any impe dance load, yet some claim 1,000.

They lie. The FTC tested some new LED lamps and found out that some don't p ut out near the lumens they claim. The government lets them lie more and mo re. You get an interest bearing bank account and they tell you the APY, not the APR. The number looks better.

Same with MPG figures for cars, it is always YMMV because they lie. Actuall y they put it that way in the US so the numbers look better, much of the re st of the world uses litres per kilometer which means the lower number is b etter.

At least with this mythical "RMS power" you can count on something. Not pea k instantaneous power, not music power whatever that means. Like if the sam e people measure the MPG of a car the same way at least you have a method o f comparison.

Yes, you're correct. s/times/divided by/

You have to look at the definitions. They're not always what you think they should be but they are what they are.

You wouldn't want to eat a MaxiPad, either. ...or a steel wool pad, either.

Sine waves aren't a good measure of audio amp performance. Sound has a much higher crest factor.

(55V/1.4)^2/4 is about 400W. Allow some clipping...

From the above, you must think they use 10mohm speakers. ;-)

The bottom line is that you don't know how audio power is measured. Hint: They don't use sine waves.

** Maker's specs are normally quite correct for any AC powered amplifier.

** That has no comprehensible meaning. ** See the maker's specs.

** There is no easy way to do that and a lot depends on what the purpose of the amplifier and speaker is - plus who is going to operate the system.

Unlike amplifiers, most speakers have no fixed limit on max power input - only a limit on the average power they can handle over a period of time. The problem is that speaker maker's never tell you what that is.

... Phil

Hint: They don't use sine waves. "

They used to. So more has changed. Fine, if I buy anything I'll measure it myself from now on, but that is unlikely. The old Phase Linear can blow many speakers and run me out of money to not be able to pay the electric bill.

4CX25000A would be good! Hmmm, I think the filament runa about 10 V at 170 A, you can run at least 10 KV on the anode at several amps. The output transformer would be about the size of a small car.

I was at a Greatful Dead concert in 1969 at the Fox Theater in St. Louis, and they used the modulator out of an AM broadcast transmitter as an audio amp. The tubes were about a foot in diameter.

Jon

It's 378. Number one, how much clipping do I allow. Ten percent ? That is h ard to take even for regular people let alone anything even close to an aud iophile. Also, that 55 volts was measured with no load. The units have basi c iron, not a tightly regulated SMPS. No way in hell it maintains that.

Watts is volts times amps. The speaker impedance does not matter. I remembe r when Fostgate had not been bought by Rockford and they rated into eight o hms and used higher rail voltages. After the buyout they started going to t he lower impedance load for their power ratings.

You do realize that higher impedance dynamic speakers are more efficient ri ght ? All that lower impedances do is to not make you kick up the voltage a s much, but per watt, with all else equal, lower impedance is less efficien t per watt. NOT per volt.

In fact even house stereo speaker manufacturers have gotten away from 1 wat t/ 1 meter sensitivity ratings because to get the good sound they have to u se that choke in the crossove that makes the load more reactive, and plus l ike in the case of my Boston Acoustics which sounded great, a 3.1 ohm woofe r. And it had inductance in front of it to bring out that bottom octave.

So now many of them use @ 2.83 VRMS/1 meter for that spec. And in the meant ime most consumer amps cannot handle that load. I have had a few pairs of s peakers with good bass that kicked in the current limiting on some amps.

So now you are telling me they don't even use a sine wave ? how the hell do they measure the distortion then ? To have to old rating system it had to be a sine wave because the distortion meter requires a sine wave.

Unless they use the null method but then it has to be phase compensated to be fair, and that is pretty much a bunch of shit.

Generally I take the easy way out. I just use integrated program material, rock usually and crank it into a real load until it clips, and I see the ra ils dropping and the ripple in it in the envelope on the scope. I take that peak value and just say "Your amp clips at ____ watts". Older amps do bett er because they didn't have quite the linearity at the upper ranges of thei r power, and that is what gave us 3dB clipping headroom, because like Pione er, all of them, they did not want to claim that power rating at 8 % distor tion. And it cost them in bigger heatsinks ad all that, but their customers didn't mind the extra watts. Or the extra weight.

boomer#...

** People who see a term being used in trade or commerce etc and think they know what it means from a literal interpretation of the words are called idiots.

A *term* means what the folk USING it intend it to mean.

A bottle of "Steak Sauce" contains no steak, but from the term it ought to be made from steak.

The term "watts RMS" is defined and used to mean the average power output measured with a sine wave signal and specified load.

Only morons reverse the word order and imply that "RMS watts" are a special kind.

.... Phil

The classical method of measuring arbitrary waveform power up into the RF, microwave and light frequencies is to feed the unknown waveform to a resistor.

An other identical resistor is heated by a variable DC source, which is adjusted until the resistors are at the same temperature. Then jus measure the DC source voltage and current and multiply them together.

Of course, this is a slow method, so this is a good way for long time output power measurements.

A more modern method of measuring how much power is actually delivered into a speaker is to put a small (less than 0.1 ohm) in series with the speaker. Using a two channel (stereo) ADC, measure the amplifier output voltage directly from the amplifier terminals on one channel (L) and the voltage drop (current) across the resistor (R).

Put the resistor in the speaker return wire to avoid ADC common mode issues.

Assuming the ADC runs at 48 kHz, for each sample first scale both samples so that they represents Volts resp. Amperes. Then multiple the L and R channel scaled samples with each other (representing the instant ious energy for that sample) and add the product to an accumulator. After 1 second divide the accumulated value by 48000 to get directly watts. If you want 1 minute average, after one minute, divide by 2,880,000. Then reset accumulator for next period.

That may be true in the industrial amplifier market, but in the consumer market I don't think it is even vaguely true. Such meaningless expressions as "peak audio power" are common.

--

Roger Hayter

Maths may be needed to understand the measurements, but very little is needed to do the actual measurement. The power output of an audio amplifier is defined as the power produced into a defined load (often four or eight ohms) as an audio sine wave at a given frequency, or range of frequencies, with a given degree of distortion. The power in an audio sine wave can be at most about a quarter of maximum voltage peak the amplifier can achieve into a suitable load resistance sqared and divided by that resistance. Because by defiinition we are talking about power in an audio sine wave. Any power above that would be very distorted. In theory an amplifier could produce about four times as much power as a square wave as it could as a sine wave. So to measure it you need a standard resistive load, a voltmeter and a distortion measuring meter of some kind. Oh, and a signal generator to drive the amplifier with an audio signal of the desired frequency.

For a solid state amplifier the supply voltage to the output stage and for a valve amplifier the designed anode dissipation of the output valves can give a very good idea of the maximum possible power an audio amplifier could produce. It is an interesting exercise to compare that with the claimed power output of an amplifier. At the very least, comparing this with the claimed power output can give one an idea of what class of people the makers are trying to sell the amplifier to.

--

Roger Hayter

** No, it is generally true of audio amplifiers for hi-fi and pro audio.

** Think again.

** In car audio and portables.

Not "AC powered ".

.... Phil

** Correct.

** Wrong - and how very dumb.

** Wrong again, same basic error.

The power ratio is 2:1.

** Wrong about valve amplifiers.

A pair of output valves can be run in class A, class AB, class B and class B split rail modes.

Efficiency varies from 30% to 75% over that range - so the power output varies by a large factor.

.... Phil

Noted.

You are of course correct. I was confused by the rail to rail voltage, as indeed some advertisers seem to be!

I was saying that the maximum possible can be predicted - if the designer decides to use class A then even this won't be achieved. Just a rule of thumb to exclude ridiculous claims about "peak music power" or whatever.

--

Roger Hayter

The speaker coil resistance increases with temperature (and hence of previous power input). The power limit can be calculated by knowing the coil maximum temperature at a specific power (actually voltage output for a typical amplifier).

The power limits for a closed speaker is more predictable, but for bass reflex enclosures, you can easily blow the cone out at 20 Hz (not to mention 3 Hz, which is typically the frequency, at which the final amplifier drops to 1, due to the feedback loop) amplifier.