Watt for speaker and Watt for power consumption, Class D and AB

[jst]

For example Aiyima A07 , it's a class D, and default power supply/power adapter is 32v 5A, Amirm measurement is 48-61watt per channel to 8ohm speaker, or 122 watt in total, while the power consumption is 32x5=160 watt. So it's comparable from power consumption on the adapter and watt to speaker output.

But I have CA Azur640A, it's a class AB, power consumption is 600 watt according to spec and speaker output is 65 watt per channel 8ohm. Why does it take so much power if the output to speaker is much less ?

Or power consumption and speaker wattage is calculated differently ? and it's not the right measurement for speaker wattage.

About Class D / AB amplifier, does AB consume more power than D assuming both have more or less same spec ?

Is there a wall socket with info on lcd that show volt amp and watt ? what does it called ? Is this the right one (kwh meter) ?

Spoiler: kwhmeter

 

[kchap]

For example Aiyima A07 , it's a class D, and default power supply/power adapter is 32v 5A, Amirm measurement is 48-61watt per channel to 8ohm speaker, or 122 watt in total, while the power consumption is 32x5=160 watt. So it's comparable from power consumption on the adapter and watt to speaker output.

But I have CA Azur640A, it's a class AB, power consumption is 600 watt according to spec and speaker output is 65 watt per channel 8ohm. Why does it take so much power if the output to speaker is much less ?

Or power consumption and speaker wattage is calculated differently ? and it's not the right measurement for speaker wattage.

About Class D / AB amplifier, does AB consume more power than D assuming both have more or less same spec ?

Is there a wall socket with info on lcd that show volt amp and watt ? what does it called ? Is this the right one (kwh meter) ?

Spoiler: kwhmeter

Let's pretend we have ideal qmplifiers:
The 32V 5A PSU can supply a maximum 160W but it would not be required to supply that continuously; it depends on what the amp is drawing at the time. Yes if the class D amp was perfect it could supply an 80W per channel amp; both channels driven.

• The ideal class A amp can only be 25% efficient; a 100W amp will dissipate at at least 400W overall, even when the speakers are unplugged.
• The ideal class B amp can 78% efficient; a 100W amp running at maximum power will dissipate 127 ~ 130W overall. At least the ideal class B amp does not dissipate any power when the speakers are unplugged.
• Class AB runs in class A at low power and class B at high power. Efficiency is hard to calculate, I had a DIY mosfet amp I built 80's which was biased to 200mA on a +-50V supply, say it was 60 ~ 70% efficient.
• The ideal Class D amp is 100% efficient; including 0W power dissipation when the speakers are disconnected.

Of course no device is 100% efficient, speaker impedance is all over the place and there is a huge difference between the the peak power and average power of an audio signal. The general observations by most people are; modern class D amps tend compact and feel tepid to touch, class AB amps are bigger and feel warm to the touch - hot when working hard. Class A amps - especially valve amps - can be useful for central heating.

Your CA amp is rated at 120W in to a 4ohm load; the Version 2 model. That's 240W Pk, 580W both channels driven. Specifying a 600 ~ 700W maximum power is not unusual. According to the specs standby power is < 10W and idle power is 35W; the amp is not drawing 600W continuously.

There are lots of cheap power meters available. They should be okay monitoring average power. I'm not sure how good they are at detecting the peaks. Peak power could be easily be 20 times the average.

 

[jst]

The ideal class A amp can only be 25% efficient; a 100W amp will dissipate at at least 400W overall, even when the speakers are unplugged.

Wait...25% efficient means it will draw 4x ? And that's an ideal, while in reality it will draw more than 4x ?

Your CA amp is rated at 120W in to a 4ohm load; the Version 2 model. That's 240W Pk, 580W both channels driven. Specifying a 600 ~ 700W maximum power is not unusual. According to the specs standby power is < 10W and idle power is 35W; the amp is not drawing 600W continuously.

The spec is 75w for 8 ohm and 100watt for 4 ohm, per channel, for v2.
( https://www.manualslib.com/manual/23922/Cambridge-Audio-Azur-640a-V2.html )

Do you mean the power load, not the speaker watt ? I'm a bit confused here. But even if it's 120w per ch (4ohm), isn't that x2 = 240watt both channel, why it goes to 580 ? Is that assuming the efficiency is 50% then it's doubled ?

That's good to know if it draws only 35 at idle lol, coz I often left it on from the moment I woke up till I slept, but now I hardly do that anymore, I use Aiyima A07, it uses much less power at 160watt max and more or less the same loudness. But I just bought new speakers and I'm curious about feeding it more power.

​

 

[pjug]

I can't find dynamic power measurements for the 640A but the 160W rated 840A can do 485W into 2 ohms.

https://www.fransvaneeckhout.be/cambridge-audio/images/Hi-FiNewsJune06840Ahi-res.pdf

Suppose the 640A can do 200W dynamic power into 2 ohms. So 200W x 2-chan, and it is about 65% efficient. That gets you to drawing a little more than 600W.

 

[rdenney]

The other thing is that the rating on the label of a power supply is related to safe operation (meaning: It won't overheat), not to completing waveforms. A power supply rated at 32 volts and 5 amps might actually be able to produce much more for brief periods, in terms of providing enough power to complete waveforms. Amps take a signal and increase its magnitude, which means increasing the voltage (the vertical dimension of the waveform). But if the amp runs our of current-supply capability, the waveform will become truncated. Given that current and voltage are not always in phase, there is an additional power factor to take into consideration as well, which isn't the same thing as inefficiency.

But supplying enough current to fill that waveform only has to last long enough for the peak of the waveform, and not for the valley that immediately follows it in the frequency domain. So power supplies and power amp figures are given for the average of the waveform, which is averages using root-mean-square, or RMS.

But peak power is different from RMS. It's the ability to supply a given RMS for a short period. How short is short? That's usually not specified and one of the reasons we do tests. Usually, it's thermally constrained, not power-supply constrained. Even compressed pop music is not uniformly loud, like a test signal turned to maximum. It's just louder more consistently than uncompressed music.

My B&K amps have a label on the back that claims 600 watts of power draw at 120 VAC, 60 Hz. The amps are rated at 125 watts into 8 ohms or 185 watts into four ohms, with a 1.6 dB headroom for each channel. It's nearly impossible to really know what the power output actually is into a speaker load unless you can measure the impedance and voltage at the frequencies being played. And that's the average of the waveform. B&K didn't just run the calculation. They created a design worst case and sized the power supply for that, and then determined what the power supply components ratings would need to be if it drew that much power. The label on the back is related to that. It's an AB amp, of course, so that's going to be 50% more than the amplifier actually needs, just to account for the heat it is also dissipating.

TL;DR; It's not easy at all to relate the power supply rating and the amplifier power output--they are measured differently and for different purposes.

Rick "look at measurements to see at what power the amp starts to clip--that's what counts" Denney

 

[kchap]

Wait...25% efficient means it will draw 4x ? And that's an ideal, while in reality it will draw more than 4x ?


The spec is 75w for 8 ohm and 100watt for 4 ohm, per channel, for v2.
( https://www.manualslib.com/manual/23922/Cambridge-Audio-Azur-640a-V2.html )

Do you mean the power load, not the speaker watt ? I'm a bit confused here. But even if it's 120w per ch (4ohm), isn't that x2 = 240watt both channel, why it goes to 580 ? Is that assuming the efficiency is 50% then it's doubled ?

That's good to know if it draws only 35 at idle lol, coz I often left it on from the moment I woke up till I slept, but now I hardly do that anymore, I use Aiyima A07, it uses much less power at 160watt max and more or less the same loudness. But I just bought new speakers and I'm curious about feeding it more power.

​

25% is about right. My theory is a bit rough but I think transformer coupled push amps a re more efficient; 50% ? In real life a valve push pull class A would be lucky to get anywhere near 40%. In class A the output transistors or valves have to be conducting at all times even in the absence of a load.

My figures where based on the manual, see image. Maybe I have the wrong manual. 100W or 120W we are still in the same ball park. Let us say 100W into 4ohm, because the maths is easier. That represents RMS voltage of 20V. For a sine wave the peak voltage would 28V. So the instantaneous power at the peak is 28^2 x 4, 200W. We have 2 channels so instantaneous power is now 400W. Allow for various additional power losses and a margin of error and you are back around the 600W figure. For the purposes of calculating the load on upstream transformers etc. that peak power is important. Also the amp can probably deliver higher powers for short periods. The speaker impedance might be as low as 2ohms at some frequencies.

Your speakers should be ok. The average powers are far lower. If you can find a suitable PSU the Aiyima A07 can run up to 48V; according to the Aiyima web site. Simple rule, if the speakers sound distressed, turn the volume down.

 

[DonH56]

Ideally single-ended class A is at best ~27% efficient IIRC. Push-pull class A is ideally 50% efficient. Class B is ~67% efficient. Class AB is someplace between class A and class B depending upon how it is biased. Class D is ~100% efficient (assuming no losses in the (superconducting?) output switches and zero power in the stages driving them). That is with no additional losses in the output stage, no power consumed in driving and other low-level circuits, no power for any other circuits like processing and power lights, etc. That is also at full power output; efficiency falls at lower power.

 

[AnalogSteph]

The most drastic differences between AB and D amplifier efficiency are found at partial power output. AB amps famously tend to heat up the most at about 1/3 of maximum power output. This is because a conventional amplifier like that essentially is one big linear regulator that has to dissipate (Vrail -Vout)*Iout. By contrast, highest power dissipation in Class D amps occurs at maximum power, and even then efficiency still commonly hits >90%. Idle power dissipation also favors Class D although it may be little more than a factor of 2 for a whole integrated amp.

 

[kchap]

My job is done, I've attracted the experts. Off to bed. Cheers everyone.

 

[GeorgeBynum]

The other thing is that the rating on the label of a power supply is related to safe operation (meaning: It won't overheat), not to completing waveforms. A power supply rated at 32 volts and 5 amps might actually be able to produce much more for brief periods, in terms of providing enough power to complete waveforms.

Rick "look at measurements to see at what power the amp starts to clip--that's what counts" Denney

Very, VERY, true for linear power supplies where filter capacitors help things along. Usually less so with switching supplies which will often "fold back" as a form of short circuit protection. I've seen switchers shut off (by design, not unintended) at less than 10% "overload". In industrial valve (hydraulic) applications this has bitten some users; a valve may be advertised as needing 2A @ 24V but with some controls require 3.5A for up to 60 ms. Put it on a 2.4A supply and it will result in "fold back", then see the "short" is resolved only to repeat the cycle every 2 or 3 seconds.

Your clip comment is analogous to my losing valve capability.

 

[DVDdoug]

Wait...25% efficient means it will draw 4x ? And that's an ideal, while in reality it will draw more than 4x ?

That's "best case". Note that he said, "even with the speakers unplugged". If you are putting 1W into the speakers it's less than 0.25% efficient and you'll never get close to 25% unless you are running constant full-power test-tones.

IMO - Class A is a "dumb design" for a power amp. It's OK for a preamp but if I was building a preamp I'd use an op-amp and the odds are the op-amp would be class AB internally.

 

[jst]

Thanks for all the input, so..is it safe to say :
Class A 25% efficiency
Class B 67% efficiency
Class AB 50% efficiency
Class D 100% efficiency

to produce 100watt output to per channel speaker then power draw would be like :
Class A 100 * 4 = 400 watt power draw per ch (more or less)
Class B 100 * 1.49 = 149 watt power draw per ch (more or less)
Class AB 100 * 2 = 200 watt power draw per ch (more or less)
Class D 100 * 1 = 100 watt power draw per ch (more or less)

Am I slightly right or not even close ? trying to make sense of it based from what I read here.

Maybe that's why most class A amps are tube amp with low wattage and need high efficiency speaker lol, to make up with the inefficiency of power with it.

 

[ZolaIII]

Class B (AB) is up to 70%, class D is as efficient as its power supply is (75~86 lower end and 85~96 for higher end).

 

[DonH56]

Not unreasonable. Most class-A audio amplifiers are push-pull so 50% ideally. Some of the smaller ones, like SETs and some low-power transistor amps and kits, are single-ended and drop to the 25% level.

 

[Raindog123]

In reality, a Class D will be about 90% efficient, Class AB just under 50%. This is when closer to their [steady state] nominal max power — at lower power levels the efficiency will be lower (for the sake of discussion, by about 10%).

Also keep in mind, when they quote a power-supply wattage (or current), it’s usually its “max“ [steady state, sustained] value. Most of the time, for your listening volume level your power supply would operate at a fraction of this power/current.

 

[jst]

How about this amp ?

Spoiler: pa350

Brand/model : Ashley PA-350
Spec :
- 2 x 350 watt @8ohm
- 2 x 525 watt @4ohm
- 1050 watt 8 ohm bridge
- Circuit type : class AB

Here's the transformer 60V 10A =600 watt

Spoiler: transformer ct60v ac

The question is, how come a class AB with 600 watt power consumtion and about 50% efficiency, can produce 350 watt per ch = 700 watt total ? Is the spec a lie ? that's outputting even more power to the speaker than the power it consumes from the wall.

 

[alex-z]

The question is, how come a class AB with 600 watt power consumtion and about 50% efficiency, can produce 350 watt per ch = 700 watt total ? Is the spec a lie ? that's outputting even more power to the speaker than the power it consumes from the wall.

You answered your own question. The efficiency range of class AB is between 39.3 and 78.5%, depending on the bias point. If a class AB amplifier pulls 600 watts from the wall, it can never output more than 471 watts continuous. Peak power can be higher, if enough capacitance is available.

 

[Willem]

So in this day and age of higher energy prices and global warming class A is really not a good idea anymore, and class D is the way of the future. I would not bin it if I owned a good class AB design, but the next one would surely be a class D.
For us, this is precisely what we are doing each time we buy a new replacement appliance of one kind or another. Our new vacuum cleaner consumes only half of what the old one from a well respected brand did, is quieter, and works better. Current refrigerators consume about a quarter of those from 15 years ago, and my new desktop computer is ultra efficient as well (and efficient enough not to need a fan). Of course, in absolute terms audio equipment does not consume that much, but it is the sum of all these little things.

 

[PierreV]

The idea that one had to distill a lot of "dirty watts" from the power lines to produce a few "clean watts" had a lot of appeal in the audiophile community of the 80ies. It still has I guess. Jean Hiraga (again) was partly to blame.

 

[kchap]

How about this amp ?

Spoiler: pa350

View attachment 159858

Brand/model : Ashley PA-350
Spec :
- 2 x 350 watt @8ohm
- 2 x 525 watt @4ohm
- 1050 watt 8 ohm bridge
- Circuit type : class AB

Here's the transformer 60V 10A =600 watt

Spoiler: transformer ct60v ac

View attachment 159857

The question is, how come a class AB with 600 watt power consumtion and about 50% efficiency, can produce 350 watt per ch = 700 watt total ? Is the spec a lie ? that's outputting even more power to the speaker than the power it consumes from the wall.

The transformer rating is probably does as much about avoiding overheating. For short periods of time I'm sure it could push out more power than the rated 600W.