Watts manufacturers doings to us - RMS vs Peak - it’s time to take a stand

[Cuzzie]

Someone wrote on it with a sharpie

[Chienmortbb]

But how may watts RMS was the Sharpie?

[JapanAxe]

2 hours ago, Muzz said:

Ahhhh, but how did you know they were 50w and 800w? 😁

[tauzero]

On 17/11/2018 at 23:41, Phil Starr said:

Fair enough.

If we assume the Bugera is a well designed class D amp with an efficient power supply and draws 880W from the mains and can run continuously at that level then it is probably around 80% efficiency and can supply about 704W. That's with an undistorted sinewave within the passband of the amp. That's the value of what is normally described as the RMS or continuous value. If you look at section 3.2 of the Wikipedia article then you see the peak power of a sine wave is twice the average power. So if it is a 704W amplifier than the peak power is 1408W. If you look at a range of amp and speaker specs you'll see that there's a fairly consistent description of things described as 500W continuous 1000W peak, 200Wrms 400W peak and so on. It's a simple mathematical ratio which enables those who want to boast that their system is 1000W but allows for fair comparison. Actually it is meaningless as it is always double so no extra information is being given. If any other figure is given then as the Wiki article says about PMPO there is no accepted way of calculating the figure and it is without merit. Basically the sums don't add up.

Isn't there another complexity - continuous vs music power, as opposed to RMS vs peak? Music power is the transient power that can briefly be attained, which generally historically has been twice the continuous output, so peak music power = 4 x continuous RMS. As you say, PMPO is a random number - a quick check on That Ebay (looking for "pmpo computer speaker") shows some items with PMPO = RMS x 2, and one pair which has 80W PMPO = 6W RMS (incidentally, the ones with PMPO = 2 x RMS seem to have got there by thinking of a big number, saying that's PMPO, then dividing it by 2 to get RMS).

[Bill Fitzmaurice]

There are two ways of rating power, RMS watts and everything else. RMS is what's used when you want to give your customer accurate information, anything else is what's used when you don't. RMS watts might not be the best description, as RMS really only properly refers to voltage, but it's worked well enough for nigh on a century, so there's no reason to change it now.

[Phil Starr]

At the risk of stirring up another hornets nest I'd like another go at this. I don't think it's difficult to understand and it's a helpful concept.

The first thing to say is that watts are real and that they do matter. An amp with more watts will ultimately be louder than one with fewer watts through any given speaker. If you want it louder then you can increase the watts, or improve the speaker or a bit of both. If you are buying an amp (a head rather than a combo) then the wattage is one of the things you should know. One of many, but one of the big ones.

Watts can be measured (this is Wikipedia) 'The watt (symbol: W) is a unit of power. In the International System of Units (SI) it is defined as a derived unit of 1 joule per second,[1] and is used to quantify the rate of energy transfer.' So long as you all stick to the same system of measurement then it's a simple measurement and most secondary school students could do it. You can convert watts into any other measurement of power, so 745W is one horsepower. (That's brilliant, my Mark Bass Tube is 0.67hp, should all valve amps be rated in horsepower or British thermal units )

That's where the rms measurement comes from. if you have a direct current running a motor or a heater than measuring the power is simple if you know the current or the voltage and the resistance. When someone had the bright idea of using an amp to control a motor then you need a way of converting. The average voltage of an amps output is zero as it goes up and down with the waveform, but mathematically if you square the voltage then all the minus's become plus's and you can work out the average or mean. Then if you find the square root of the that average you can compare ac watts and dc watts. Its the root of the mean of the square... RMS. Of course if you just use the amp to heat an 8ohm resistor it will do the same thing, you can compare the energy being transferred, just less accurately.

In the start of amp design that's how it was done. All controlled by engineers you see, honest, measurable, repeatable. then the ad men got involved.

To be fair the engineering method could reasonably be said to be over-doing things for music. Music has loud bits and quiet bits, bits where the amp isn't making heat into 8ohms and the amp can cool down. That means you can make an amp with less cooling built in and use a smaller power supply saving weight and cost. It makes engineering sense to use an amp that can make it's power for a limited duty cycle, say 20% of the time. That's where the confusion comes in you have to have a standard that everyone agrees on or people cheat. Rms all the time, 20% of the time or just for 0.01 of a second at a time? In Europe the DIN 45-500 standard held sway for a long while it is basically an rms rating which the amp has to produce for 10 minutes with a pink noise signal filtered to contain a similar spectrum of energy to real music. EIA RS-426 and IEC 268 used outside Europe are broadly similar. It's a fair test for audio amps and it ought to be used. Behringer used to give rms ratings based on IEC/EIA measurements.

It's a bit like cars and their fuel consumption figures. You probably won't get the consumption they advertise and some will be closer than others depending upon the detailed design of the car but it's a good start and cars with similar figures will give broadly similar performance. You know the test isn't perfect but that it is fair and good enough to be a big help when choosing what to buy. When somebody like Volkswagen cheat eventually they get caught and everyone knows they are cheats, people lose their jobs and fines and compensation come into play. The same with food standards. Why should consumer electronics be any different?
 

[chris_b]

5 hours ago, Bill Fitzmaurice said:

 RMS really only properly refers to voltage, but it's worked well enough for nigh on a century, so there's no reason to change it now.

Except the Standard of the Audio Engineering Society of North America (AES) seems to be the coming thing.

[Count Bassy]

6 hours ago, Bill Fitzmaurice said:

There are two ways of rating power, RMS watts and everything else. RMS is what's used when you want to give your customer accurate information, anything else is what's used when you don't. RMS watts might not be the best description, as RMS really only properly refers to voltage, but it's worked well enough for nigh on a century, so there's no reason to change it now.

RMS can also be used for current, and can be used for power as long as you take account of the  power factor.

Hopefully not patronising note to those who don't know: Power factor takes account of the phase difference between the voltage and the current. In a pure resitive load load this is '1', but when you get inductance and capacitance involved it is usually less than one. Fairly easy to calculate, or measure, with sine waves, but very complicated when you have a 'random' signal as in audio. It can never be more than 1.

 

[Stylon Pilson]

10 hours ago, Phil Starr said:

An amp with more watts will ultimately be louder than one with fewer watts through any given speaker.

In an exclusively technical sense, yes. But perceived volume isn't always the same as actual volume - that's why we have the whole "valve watts vs solid state watts" discussion. When it comes down to choosing an amp for its suitability when gigging, it's the perceived volume that we care about, not the actual volume.

S.P.

[Phil Starr]

Just now, Stylon Pilson said:

In an exclusively technical sense, yes. But perceived volume isn't always the same as actual volume - that's why we have the whole "valve watts vs solid state watts" discussion. When it comes down to choosing an amp for its suitability when gigging, it's the perceived volume that we care about, not the actual volume.

S.P.

Not wanting to get dragged off the point here but wherever there is a difference between perceived volume and actual volume there is a technical reason for the difference, that's the reason I keep posting to try and explain the technical side. I simplify things a little sometimes but in this case I specifically said through the same speakers, so same efficiency and same frequency response. In which case the same loudness.

Valves are a special case, they compress the signal at high levels. If you used the same amount of artificial compression you could get the same sound levels out of a SS amp. I love the sound of a valve amp, they sound magic but there isn't magic inside, just engineering. 

People do, and should, buy amps and rigs based on sound alone. In the end that's the only important thing but some people are curious about the technical side of things and it can help you home in on what you are seeking having a little knowledge.

I don't think anything you said was wrong but the statement you quoted is correct. A 300W valve amp will be louder through the same speaker as a 100W valve amp.

[Guest gazzatriumph]

A 300W valve ampwill be louder through the same speaker as a 100Wvalve amp.

Would that not depend on the speaker. 

[chris_b]

3 minutes ago, gazzatriumph said:

Would that not depend on the speaker. 

I'd have thought so.

If, for instance, you are using a "standard" 112 then a 300 watt amp will sound as loud as a 100 watt amp, because you can probably only put 100 watts through it before it starts to crack up and sound bad.

[Cuzzie]

When my speaker farts out, it not only sounds bad it also smells bad

[Bill Fitzmaurice]

18 minutes ago, gazzatriumph said:

A 300W valve ampwill be louder through the same speaker as a 100Wvalve amp.

Would that not depend on the speaker. 

It depends on a lot of factors, and no, a 300w valve amp will not necessarily be louder through the same speaker as a 100w valve amp. And there's the rub, there is no direct correlation between watts and perceived loudness.

[Guest MoJo]

52 minutes ago, Cuzzie said:

When my speaker farts out, it not only sounds bad it also smells bad

Are you sure that it's not a speaker induced sympathetic bowel movement, a variation on trouser flapping, if you will?

[Cuzzie]

Ahhhhhh well spotted Sympathetic overdrive (as opposed to parasympathetic) is definitely the best for tube emulation

[EBS_freak]

Inverse log pots on the volume. Man this amp is so loud and it's only on 2 of 10.

Of course, it doesn't really get that much louder past that point.

 

Whilst humans are involved, theres always going to be scams.

[Cuzzie]

Back to my Logs and flapping trousers again.....It is Yuletide 

[Phil Starr]

On ‎12‎/‎12‎/‎2018 at 12:32, gazzatriumph said:

A 300W valve ampwill be louder through the same speaker as a 100Wvalve amp.

Would that not depend on the speaker. 

That's why I specified the same speaker. 

 

[Phil Starr]

Before anyone else tries to confuse the issue any further. I'm assuming for the sake of simplicity that we are using the same speaker, that speaker can handle the power adequately and that the same eq is applied to the signal. I'm also assuming that we have turned up the volume control to the point where the same level of distortion (let's say to DIN 45-500 here so 1%dist and the amp is running for 10mins and the temperature has stabilised) All of those things are implicit in my post which would have been far too long otherwise, as well as not being very helpful.

OK which 'expert' thinks the 300W(EIA, IEC,DIN) would not be louder by 4.77dB than the 100W(EIA,IEC,DIN) amp. I'm surprised some people don't know how to calculate that themselves

Level in dB: L = 10 × log (power ratio)

 

Edited December 13, 2018 by Phil Starr

[Bill Fitzmaurice]

My hand is raised. If everything about the two amps transfer function was identical then 300w potentially could be 4.77dB louder than 100w. But finding two amps with identical transfer function other than power would be next to impossible, and even if you managed to find them then you'd need to find a speaker with linear output with respect to power input from 100w to 300w. That would be equally as difficult a quest.

[agedhorse]

On 17/11/2018 at 15:06, itu said:

One thing is this power taken from the mains. If the unit takes less than claimed, the unit should be able to produce this amount of extra power by itself. Got it? It should be able to create power from, yes, where from?

Does this serve us customers? I don't think so.

I think you are misunderstanding what the rated input power number means and why it's required of any amplifier that is sold in the EU, as it must be tested and certified to EN60065. This regulation specifies that the rated power input be based on a minimum of 1/8 rated audio power (in RMS units) at the lowest rated impedance. It can be higher than 1/8-duty cycle if the amp is specifically designed for this operation, and passes the tests accordingly (many do not, or the companies may choose not to test to extended duty cycle beyond the legally required 1/8-power which is 12.5%) Designers have no choice but to follow these regulations, they are regulations that the European countries have specified that we designers follow. If you don't like it, take it up with the authorities having jurisdiction in your region.

The reason for this extends all the way through the audio world, in pro audio for example, this number is used to assist in sizing branch and feeder mains circuits, environmental heating loads and such for installations where many amps may be installed. The same certification standard applies to bass amps.  This is a commonly misunderstood bit of information, if you wish to denigrate any manufacturer for using it you might want to understand more about what the number legally means first.

Once you understand what this number means on a product that carries an NRTL agency approval, you can then calculate backwards by using efficiency to discover the maximum rated power in RMS terms. 

I normally don't post here, but am available at TalkBass as many of you here already know.

[stevie]

It's very nice to see you on here, agedhorse. I hope you'll stick around.

[agedhorse]

On 18/11/2018 at 00:01, Passinwind said:

In the US the unit's power consumption from the wall  figure is typically spec'ed at 1/8th duty cycle, although 1/4-1/3 is generally more realistic for rock or hard rock with some signal compression. You can easily confirm the relevance of this by recording some actual bass guitar playing and analyzing the crest factor.

Thanks Charlie!

On 19/11/2018 at 05:12, Count Bassy said:

With industial servo amplifiers (to drive motrors) you tend to get a specification such as  6 amps continuous, 18 amps absolute peak for 2 seconds, 10 amps for 10 seconds etc. and generally get some sort of graph showing current again how long it it will deliver it for. However it is all very complicated, for example if you've just used 18 amps for 2 seconds, how long before you can do it again? 

If after the first 18 amps for 2 seconds it goes back to zero then you might be able to repeat it after 4 seconds (i.e. the average over 6 seconds is the continuous rating of 6 amps).

If after the 18 amps for 2 seconds it drops back to 5 amps then you might have to wait 26 seconds before you could do it again.

It all comes down to "Duty Cycle", but it's all very complicated. With servo drives you tend to work out what you think it needs, add a bit for luck, and see how it goes. This is why a lot of industial motors and drives are over specified - i.e. it's cheaper to over spec it than to underspec it and have to replace it.

 

This is very much the same process that we use when designing amplifiers, developing the combination of load profile, driving signal and thermal conditions, integrating these factors over time. This was even more the case when designing linear amps when we had to worry (a lot) more about SOA (safe operating area). Class D amps are more forgiving in this area than linear amps due to cut-off/saturation operation.

On 19/11/2018 at 06:28, Chienmortbb said:

Peak of a sine wave is 1.414  x RMS and Peak to Peak is 2 x Peak.

 

 

While this is correct for voltage, it is NOT correct for power as you would be ignoring the squared factor in the power equation (P=V^^2/R)

On 10/12/2018 at 09:00, Chienmortbb said:

The only way to test and amp is into a resistive load and quote the continuous power (RMA amps x RMS voltage). That give s Average watts. Sadly some use RMS Watts. If you audition an amp or speaker it has to be against another amp/cab in the same room at the same time as our ears have no memory.  You may still get it wrong though as the room is a major component of your sound.

When RMS is used in describing power, all it means is that the voltage and current are based on RMS units, not that the power itself is in RMS units. It's a short-hand notation that's understood by most professionals in the industry and has been used for years.

Edited December 13, 2018 by agedhorse

[itu]

3 hours ago, agedhorse said:

I think you are misunderstanding what the rated input power number means and why it's required of any amplifier that is sold in the EU, as it must be tested and certified to EN60065. This regulation specifies that the rated power input be based on a minimum of 1/8 rated audio power (in RMS units) at the lowest rated impedance.

OK, so if the amp has linear power, the situation is totally different than with switching powers? My understanding may be outdated, because SMPS has pretty much different behavior compared to linear, transformer-based supplys. I would say, that linear needs quite a lot of power to produce the claimed numbers - in RMS.

I have an old HiFi amp from the 80's that can push 2 x 110 W RMS continuous to 8 ohms (tested) and its maximum input power is 650 W. I understand losses and the power consumption of the other components, so the amount of input power makes sense. But if an amp says 1 kW and the input power is 200 W, I feel like I have fallen from the train. Sir, please, open up this a bit more; I do not have reasonable access to that standard, as it costs 300 € + VAT. Can this SMPS class-D amp push continuously 1 kW RMS to 8 ohms or not?

Edited December 13, 2018 by itu