bowtie427ss
Lunatic Member
Agree. I think your post, #2 defined it very well.
-
Feeling lucky? Check out the AK 24th Anniversary Giveaway thread and join in the fun!
Amplifiers - rated watts vs. watts per channel
- Thread starter Thumpy
- Start date
JoeESP9
ESL's & tubes since 83
Where do you think that other 35-40% goes? The correct answer is heat.
Class A amplifiers are even less efficient. Class D amplifiers are probably the most efficient. Even so, they are not 100% efficient. They get warm to the touch. The metal work on a 100% efficient amplifier will stay at room temperature. However, there are no 100% efficient amplifiers.
Thumpy
Well-Known Member
You might want to re-read my original post. I already know that no amp is 100% efficient and there are losses due to heat. I mentioned it TWICE in my original post. You are telling me nothing I didn't already know. My original question was how can an amp draw only 170 watts, but produce 200?
Xenon-Codex
Active Member
The output is false. If it's measured with 10% distortion like many things are today, it's useless value. It's possible to get high output, but it destroys the speakers with the huge distortion. No sine wave anymore, more like square wave or even DC to the speakers. Real hifi amps are measured with 0.01% distortion...
swechsler
Frog Whisperer
Not really. Any THD spec <= ~0.5% is essentially irrelevant; they're just bragging rights by the manufacturer. One of the most highly regarded vintage amps ever, the Harman/Kardon Citation II, is rated at 60 watts RMS per channel with 0.5% THD.
Nah, for the back panel rating, they roll frog bones in the dirt and get the rating from there.
whoaru99
Epic Member
There is a method, but only the mfg and UL (where listed) seem to know for sure what it is on any given unit.
That said, some companies like the QSC example I gave earlier, do provide. Here's a small excerpt from a longer list. The 4 ohm, 1/8 power rating is typically what you find on the back panel of those.
Last edited:
The power consumption for most equipment I've come across is based on the unit being driven to rated power and does NOT include draw from additional accessories connected to auxiliary outlets. Usually the consumption is stated with language "MAX" or "driven to rated power". Some manufacturers also provide consumption at "idle".
I cannot conceive why anything under than consumption at "rated" or "full" output would be listed for the information to be meaningful. I would almost think with would be required for UL certification.
I cannot conceive why anything under than consumption at "rated" or "full" output would be listed for the information to be meaningful. I would almost think with would be required for UL certification.
whoaru99
Epic Member
It is not the requirement for UL. I more often, but not exclusively, see "max" ratings on gear that's not UL listed and that's likely for CYA since they haven't undergone a UL certification.
My Levinson No. 23 doesn't even list power consumption on the back panel. I guess it doesn't draw any power.
There is no UL marking on the amp.The general point of UL not requiring power consumption for audio amps at continuous rated power is because that's not how amps work when playing music/normal use. That's unlike, say, a 1500W heater that draws 1500 watts all the time it's in normal operation.
Look back at the QSC info I posted. I have several amps on that list. Look specifically at the max draw for PL3.4 (one of which I have). If it had to be listed at max draw that would be 69 amps. It would have to be sold with a power cord and plug bigger than a typical welder. But it isn't because that's not how the amp works in its intended function. In its intended function the 4 ohm, 1/8 power figure applies and that's what's on the back panel.
Last edited:
It is not the requirement for UL. I more often, but not exclusively, see "max" ratings on gear that's not UL listed and that's likely for CYA since they haven't undergone a UL certification.
My Levinson No. 23 doesn't even list power consumption on the back panel. I guess it doesn't draw any power.
The general point of UL not requiring power consumption for audio amps at continuous rated power is because that's not how amps work when playing music/normal use. That's unlike, say, a 1500W heater that draws 1500 watts all the time it's in normal operation.
Look back at the QSC info I posted. I have several amps on that list. Look specifically at the max draw for PL3.4 (one of which I have). If it had to be listed at max draw that would be 69 amps. It would have to be sold with a power cord and plug bigger than a typical welder. But it isn't because that's not how the amp works in its intended function. In its intended function the 4 ohm, 1/8 power figure applies and that's what's on the back panel.
But how would the consumer be aware of potential maximum consumption in order to determine whether overloading issues may result - particularly in situations where A/C power outlets may be shared with other electrical appliances? Wouldn't it make more sense to err on the side of caution?
whoaru99
Epic Member
Putting true max, as in worst case, power draw on the back of an amplifer as standard doesn't make any sense to me. That sort of draw only happens on a full power sine wave bench test.
If by max you really mean listed similarly to the way QSC does it*, I'd be on board with that. That is a practical "max" rating, albeit higher than the typical home use draw (meaning usually people don't play their amps WFO most of the time).
*Current draw at 1/8 of full power is measured with pink noise as a signal. It approximates operating with music or voice with light clipping and repesents the amplifier's typical "clean" maximum level, without audible clipping. Use these figures for typical maximum level operation.
Last edited:
Equipment that plugs in is rated to make sure the supplied mains power and wiring is sufficient for expected operation.
But only Class A amps draw full power at idle. For all others, the ratings imply a model of operation, usually in the form of a power profile. A sharp but short peak that exceeds the rating won’t overheat the mains wiring and risk a fire, and that’s what organizations like UL care about. (Underwriters=risk assessors for insurance companies.) Heat in the mains wiring will be more driven by the usage profile.
And that is applied on top of amplifier efficiency.
Some folks have it a bit wrong. A typical Class B amp is theoretically no more than 78% efficient when amplifying a sine wave. But nobody uses it because of crossover distortion, so they always bias the output devices to run partly in Class A—Class AB. How much they do governs efficiency. A Class A amp is the least efficient—it consumes full power even when idle. Typically these are no more than 25% efficient. So AB is somewhere in between, usually 50-60%. Class D is much more efficient.
If you run the Pioneer at its full rated average (RMS) power, playing pink noise, or maybe even just the 1-KHz sine wave they used to determine the power rating, at 200 wpc, it will fry or shutdown after not too long. That it is capable of building a waveform with the average amplitude and area that represents 200 wpc doesn’t mean it can do it all day. A few amps have kind of cooling and power supply capacity, but not many. The reason they don’t is that they don’t expect people to play sine waves or pink noise at full rated power indefinitely—it’s not a requirement most amp designers feel the need to fulfill. Thus, they build it to withstand the sort of usage profile some large majority of users might actually demand. The usage profile therefore varies as a result of spectrum and as a result of changes in volume.
So, superimpose variations in efficiency over usage profiles, and you get a model of power consumption. The average of that, plus some margin, is what will inform the input power rating so that you don’t overheat mains wiring.
But there is a reason B&K, who made my amps, suggest separate 15-amp 120VAC circuits for my 125-wpc Reference 125.2 amps. If each channel produces 125 watts, and there are two channels, and the amps are 50% efficient, the draw from the mains would only be 500 watts. (That’s where your typical factor of four comes from, I suspect.) A separate circuit can supply 1800 watts. That tells me that any given program may briefly demand much more than the rated output, and they don’t want supplied power to run out of steam in supplying it, which would cause the power supply to clip. It does NOT mean there is risk of overheating mains wiring that can only supply the actual rated input power. (The Reference 125.2 is actually rated for 680 watts of mains power.)
An amp with an output power rating higher than the input power rating tells me the output rating is probably for peak power, not RMS, and the input rating is built around an expected lower duty cycle.
Rick “the two ratings are answering different questions” Denney
But only Class A amps draw full power at idle. For all others, the ratings imply a model of operation, usually in the form of a power profile. A sharp but short peak that exceeds the rating won’t overheat the mains wiring and risk a fire, and that’s what organizations like UL care about. (Underwriters=risk assessors for insurance companies.) Heat in the mains wiring will be more driven by the usage profile.
And that is applied on top of amplifier efficiency.
Some folks have it a bit wrong. A typical Class B amp is theoretically no more than 78% efficient when amplifying a sine wave. But nobody uses it because of crossover distortion, so they always bias the output devices to run partly in Class A—Class AB. How much they do governs efficiency. A Class A amp is the least efficient—it consumes full power even when idle. Typically these are no more than 25% efficient. So AB is somewhere in between, usually 50-60%. Class D is much more efficient.
If you run the Pioneer at its full rated average (RMS) power, playing pink noise, or maybe even just the 1-KHz sine wave they used to determine the power rating, at 200 wpc, it will fry or shutdown after not too long. That it is capable of building a waveform with the average amplitude and area that represents 200 wpc doesn’t mean it can do it all day. A few amps have kind of cooling and power supply capacity, but not many. The reason they don’t is that they don’t expect people to play sine waves or pink noise at full rated power indefinitely—it’s not a requirement most amp designers feel the need to fulfill. Thus, they build it to withstand the sort of usage profile some large majority of users might actually demand. The usage profile therefore varies as a result of spectrum and as a result of changes in volume.
So, superimpose variations in efficiency over usage profiles, and you get a model of power consumption. The average of that, plus some margin, is what will inform the input power rating so that you don’t overheat mains wiring.
But there is a reason B&K, who made my amps, suggest separate 15-amp 120VAC circuits for my 125-wpc Reference 125.2 amps. If each channel produces 125 watts, and there are two channels, and the amps are 50% efficient, the draw from the mains would only be 500 watts. (That’s where your typical factor of four comes from, I suspect.) A separate circuit can supply 1800 watts. That tells me that any given program may briefly demand much more than the rated output, and they don’t want supplied power to run out of steam in supplying it, which would cause the power supply to clip. It does NOT mean there is risk of overheating mains wiring that can only supply the actual rated input power. (The Reference 125.2 is actually rated for 680 watts of mains power.)
An amp with an output power rating higher than the input power rating tells me the output rating is probably for peak power, not RMS, and the input rating is built around an expected lower duty cycle.
Rick “the two ratings are answering different questions” Denney
whoaru99
Epic Member
Yeah, peak of sorts, since it would be rather absurd to list back panel at full continuous rated power since that's not how it works playing music, which is the typical intended function of an audio amp.
That's it in a nutshell. In most cases we don't know what that duty cycle/test point is. In the case of better documentation such as the QSC examples, we do.
That's what I used to see back in the day and would test out with test equipment... metered variac, dummy loads, scope and dual channel volt meters. Reading it from the service manual, sometimes the owners manual, the number would correspond to those in the manual stated at idle, at max output both channels driven and, good pro stuff like the QSC, at 1/8 or 1/3 power. If you had it hooked up to the bench you ran it across the range and monitored the three instruments to confirm, quick and easy, just a matter of a few switches on the test panel... did this on some units to confirm a problem or after repair... usually did it with anything that had the power amp repaired.
I too sometimes check the power draw on the back of AVRs for sale and compare to the watts per channel and wonder... wonder what they are doing with the specs now days... find a way to inflate the output rating for those looking at that and find a way to deflate the consumption for those worried about resource management. Most of them don't weigh much... okay, light weight digital supply but comparatively small heat sink and cooling for outputs passing that much power... they probably don't. I've never had one on a full service bench like I used to use everyday but would like to... just to see what I think I already know. Back to the days of instantaneous peak power and other nonsense... I'm thinking that or something trending in that direction.
whoaru99
Epic Member
Well, you know, that's one of the beauties of switching amps. There isn't a ton of dissipation, so lots of heatsinking isn't necessary, for an on/off switch. That said, afaik, not too many AVRs are switching amps.
But, lots of pro amps are these days. The 10lb Crown XLS1000 mentioned earlier, 175W back panel rating, yet >200 wpc @ 8 measured output, steady, and nearly 10X the back panel rating at a max output test. Enough to overload my Kill-A-Watt, LOL.
Last edited:
Switching amps for the output... switching the power to the speakers. I know the power supply is a switcher and the signal is... or can be switched, digital that is, where we always considered line level.. but at the power transistors of the final stage to the speakers...?... those are the heat sinks and cooling fins I was thinking about. I have a QSC GX3 that is quite light in weight that has a variable speed fan with flow through design and has appropriate heat sink cooling for the outputs and some of those AVRs, if they are to be believed, all channels combined into an equivalent, have as much output power. Something isn't right somewhere unless some how or another those AVRs have digitized the power output section. Does the Crown mentioned have digital output stages? I understand the qualified back panel power rating but I'm thinking of cooling for the outputs... has that been digitized as well? If so it has me wondering what they are doing to make it acceptable which it must be or Crown wouldn't be selling it. I'm getting behind on things...
I don't use my QSC anymore as I no longer have the situation but it is a good versatile solid performing power amplifier that pretty much doesn't care what you ask it to do. I assume Crown is the same. By the way... do you mean the Crown weighs ten pounds... if so that is amazing. I am getting behind on things.
whoaru99
Epic Member
Yes, a few AVR have switching/Class D outputs. But, I'm not getting my undies in a bunch over their ratings. To me the all channels driven thing is a red herring for at least a couple reasons, but essentially all goes back to intended use and duty cycle.
Far as the Crown XLS1000, yes, weighs ~10lbs. It has switching power supply and switching/Class D output stages.
XLS1000 "bucket test" there's a little more in post #99 after that one.
Far as the Crown XLS1000, yes, weighs ~10lbs. It has switching power supply and switching/Class D output stages.
XLS1000 "bucket test" there's a little more in post #99 after that one.
Last edited:
