Class A power amps and electricity crisis..

[rando]

Through training and by repeatedly being taught how to rationalize accepting what we see every day.

One can quickly lose all ability to discern the equation Edison developed from.
From the one his accountant later soured him with.

 

[jep123]

My impetus to examine class D in more depth was simply driven by hearing some that really didn't sound bad at all. I came to realize that this rising star of the last 15 years was reaching ascendancy and if I didn't take it seriously I might get left behind. That was 6 1/2 years ago. It took us a while but we got the amp going. Now as some of you know my company makes all triode class A2 zero feedback fully differential balanced OTLs. They've gotten plenty of nice reviews over the decades as well as many awards.

I switched over from the tube amps at home to the class D about a year ago. Its nice to be able to say it is just as smooth in the mids and highs as the tubes were. Its a bit more transparent. I don't miss the tubes at all.

My point here is raw materials, cost, overhead and all that had nothing to do with it. The sound and the sound alone did. Its a happy coincidence that the amp draws so much less power, makes almost no heat, is smaller and that goes with that. IMO tube power amps are on borrowed time and class A of any type is too.

I couldn`t agree more. Class D done right sounds very musical and without a sound that many associate with typical "class D sound". Such as listening-fatigue, dryness, etc. I experience none of this, my system sounds more like a tube-based system. IMO!

JP

 

[Atmasphere]

But all assumptions aside does anyone here have any background in life cycle assessment or any specific data on embedded energy (gross process requirements especially) versus operational energy for any of the general types of amplifiers being discussed here let alone specific models of amplifiers.

It takes far less 'embedded energy' to build a class D amp than it does to build a tube amplifier! Tubes take a lot of energy to build- more than most semiconductors. Plus you have to mine the materials and take the energy to build the output transformers and provide for a filament supply. Tubes of course have fairly short life cycles.

in a class A or AB amplifier the energy to build the amp is also higher than class D on account of the materials needed to make the larger heatsinks, larger chassis and so on. Since there is more heat the filter capacitors might have a shorter life. The output devices have a lifespan related to their heating cycles. I'm sure its better now than when I was in school; back then you needed hundreds of thousands of heat cycles to short the output devices so I think that's a wash these days.

The semiconductor industry seems like they want to make switching devices rather than linear devices. If that is so the energy used to make output transistors for class D applications should be lower.

The actual energy to run a class D for a given output power is considerably lower as you know. So any way you look at it class D amps take less energy regardless of how you classify the energy.

 

[the sound of Tao]

It takes far less 'embedded energy' to build a class D amp than it does to build a tube amplifier! Tubes take a lot of energy to build- more than most semiconductors. Plus you have to mine the materials and take the energy to build the output transformers and provide for a filament supply. Tubes of course have fairly short life cycles.

in a class A or AB amplifier the energy to build the amp is also higher than class D on account of the materials needed to make the larger heatsinks, larger chassis and so on. Since there is more heat the filter capacitors might have a shorter life. The output devices have a lifespan related to their heating cycles. I'm sure its better now than when I was in school; back then you needed hundreds of thousands of heat cycles to short the output devices so I think that's a wash these days.

The semiconductor industry seems like they want to make switching devices rather than linear devices. If that is so the energy used to make output transistors for class D applications should be lower.

The actual energy to run a class D for a given output power is considerably lower as you know. So any way you look at it class D amps take less energy regardless of how you classify the energy.

Ralph, can you give actual measured data in terms of embedded energy of any class d amplifier and the embedded energy of any tube amplifier or are you just making assumptions in this?

 

[the sound of Tao]

Yes, thanks

Can you outline what criteria are typically involved in a full sustainability life cycle assessment of a product or service?…. I’m sure if you could that you’d know exactly how much research would be required to create any real finding that one product is more “environmentally friendly” than another.

Even just an essential carbon accounting of a product or service requires extraordinary levels of measurement and research. I’m genuinely waiting to see any proper peer reviewed science happen in this discussion at all.

 

[Atmasphere]

Ralph, can you give actual measured data in terms of embedded energy of any class d amplifier and the embedded energy of any tube amplifier or are you just making assumptions in this?

I suggest you seek video of what is required to make a vacuum tube and produce the elements of which its composed.

Then seek the same for an integrated circuit. Please keep in mind that what can be done with one IC might take a number of tubes. There's enough metal in the base of any tube that would be enough for a few chips, especially since the advent of surface mount technology.

Honestly I don't see how anyone could arrive at a different conclusion.

 

[christoph]

Can you outline what criteria are typically involved in a full sustainability life cycle assessment of a product or service?…. I’m sure if you could that you’d know exactly how much research would be required to create any real finding that one product is more “environmentally friendly” than another.

Even just an essential carbon accounting of a product or service requires extraordinary levels of measurement and research. I’m genuinely waiting to see any proper peer reviewed science happen in this discussion at all.

Toa, why don't you listen to some great music instead of wasting your precious time here in this thread?
Life is so much better if you don't jump over every stick someone holds in front of you

 

[the sound of Tao]

Toa, why don't you listen to some great music instead of wasting your precious time here in this thread?
Life is so much better if you don't jump over every stick someone holds in front of you

Wise words Christoph. Lol and yes, but I do find it strange that some talk about the value of measurements and I usually balance that with the importance of subjective evaluation… and then at last when I’m finally agreeing that in an area measurements are the only important thing it gets all unscientific and any truth is being lost behind the repeated clouds of guesswork.

 

[christoph]

Wise words Christoph. Lol and yes, but I do find it strange that some talk about the value of measurements and I usually balance that with the importance of subjective evaluation… and then at last when I’m finally agreeing that in an area measurements are the only important thing it gets all strangely unscientific and any truth is being lost behind the clouds of guesswork.

If the positions are completely diametral, there will never be a consensus and then it's better to agree to disagree.
At least that's how I tread. Saves nerves and energy

 

[Hear Here]

Ralph, can you give actual measured data in terms of embedded energy of any class d amplifier and the embedded energy of any tube amplifier or are you just making assumptions in this?

Frankly if you apply a modicum of imagination and common sense, you'll realise it's obvious. Why is something so obvious so difficult for one or two here to understand? It's a bit like asking someone to prove that trees are bigger than shrubs!

 

[microstrip]

Wise words Christoph. Lol and yes, but I do find it strange that some talk about the value of measurements and I usually balance that with the importance of subjective evaluation… and then at last when I’m finally agreeing that in an area measurements are the only important thing it gets all unscientific and any truth is being lost behind the repeated clouds of guesswork.

Well, if you can't separate informal coffee talk from audio debates it is your problem. This discussion is pure WBF speculation in our usual style, driven by intuition. Trying to connect it with audio measurements and subjective evaluation - I praise both a lot - is absurd.

 

[the sound of Tao]

Well, if you can't separate informal coffee talk from audio debates it is your problem. This discussion is pure WBF speculation in our usual style, driven by intuition. Trying to connect it with audio measurements and subjective evaluation - I praise both a lot - is absurd.

sigh…

 

[retired farmer]

I always wonder about the green idea people. From a personal standpoint I owned and used a tractor with record breaking horsepower per gallon hour rating and my combine was one small step down in efficiency. In my house I have spent 60000 dollars on new high efficiency windows and upgraded furnaces and water heaters. I have upgraded the furnaces in over twenty rental units to high efficiency ones. Insulation has been blown in to r60 in every rental. The house flips I have done have all had the same Insulation upgrade. I drive a tiny fuel efficient car and a small truck not a big car and full sized pickup. And I have a set of 33h mark levinson amps that are run in stand by when not in use so what? Many times I think there is a hint of jealousy within this thinking in the cases where that is the case those people need to work harder and buy what the lust after instead of getting upset at the people who have done so.

 

[Atmasphere]

@retired farmer
What does 'green' have to do with it?

The reason class A is slowly going away has nothing to do with 'green', whatever that means. It has everything to do with how you can apply feedback to the amp and its overall cost vs performance. Its a simple fact that you can now build class D amps that are the sonic equal of class A amps. That wasn't true 20 years ago but it is now.

Take this from someone who has been manufacturing class A amps for 50 years. The biggest complaint we get is heat and our amps aren't the hottest ones out there (amazingly enough). People don't want to have to run the air conditioning to play the stereo. So when class D came along the benefits on the manufacturing side were glaringly obvious to designers!!

I can go into the technical side if you like. In a nutshell its the market, not 'green' that is fueling the change to class D.

 

[InfigoAudio]

There is another option, the one I have been focusing on for many years.

The objective of class A operation is to keep both halves (in case of a push pull design) of the amplifier active at all times.
The easiest way to achieve this is to run more current through the output stage than will ever be needed to go to the speaker at any time.
This 'brute force' method creates the energy consumption problem as discussed in this thread.
Another way to do this is by running a guaranteed amount of current through both halves at all times, which can be smaller than the current required by the speaker. Having all devices conducting at all times means it never enters class AB.
This requires a different approach in driving the output stage, but the result is the same, if not better than brute force class A operation.
Of course it is never as efficient as a class D operating amplifier, but it is possible to achieve an idle power usage of only 30+ W per channel this way.

One can argue that this is not class A operation but I beg to differ.
In both cases only one half of the amplifier contributes to the speaker current during each phase, while the other half is dissipating current.
The objective is the same and the results are the same but a lot more efficient when this current is not maxed out as in brute force class A.

Hans.

 

[Atmasphere]

The objective of class A operation is to keep both halves (in case of a push pull design) of the amplifier active at all times.

This isn't quite correct! The objective is: by having both sides conduct (not go into cutoff) and by placing both halves in their most linear operating region, distortion is minimized. It is not to simply keep both halves 'active'.

This technique is only required to reduce distortion if the design cannot employ sufficient feedback or if the design uses none at all.

 

[InfigoAudio]

This isn't quite correct! The objective is: by having both sides conduct (not go into cutoff) and by placing both halves in their most linear operating region, distortion is minimized. It is not to simply keep both halves 'active'.

This technique is only required to reduce distortion if the design cannot employ sufficient feedback or if the design uses none at all.

Agreed, I swapped method and objective, but essentially meant to say the same.
The idea here is to limit the amount of overall feedback needed to get sufficiently low distortion.

 

[Atmasphere]

Agreed, I swapped method and objective, but essentially meant to say the same.
The idea here is to limit the amount of overall feedback needed to get sufficiently low distortion.

Why do you need to limit feedback?

 

[InfigoAudio]

Why do you need to limit feedback?

The discussion on feedback is always an interesting one, the school of thought to have as much feedback as possible and the one to have limited or no feedback. Most of the reasoning and results are pretty subjective, which is what makes this such an interesting field.
Applying as much feedback as possible certainly has its merits in lowering output impedance and distortion measurements, but it doesn't play nice with our way to keep both halves conducting at all times.
It then comes down to a simple choice: upping the feedback and make the amp go in class AB again (but with even better measurement results) or lowering the feedback to suit our design to stay conducting.
I chose the latter because it simply sounds better to me and all people that helped listening during testing/development...
It's great to have choices...

 

[Atmasphere]

The discussion on feedback is always an interesting one, the school of thought to have as much feedback as possible and the one to have limited or no feedback. Most of the reasoning and results are pretty subjective, which is what makes this such an interesting field.
Applying as much feedback as possible certainly has its merits in lowering output impedance and distortion measurements, but it doesn't play nice with our way to keep both halves conducting at all times.

Here's food for thought then

In most amplifiers the feedback is applied to a non-linear point which might be the cathode of an input tube, the base of a transistor or gate of an FET.

In these cases the feedback signal is distorted before it can even do its job, so higher ordered harmonics are generated. You also get intermodulations. The way to get around this is to use the most linear feedback node that you can! This will minimize the distortion otherwise caused by its application.

If you look at an opamp, the feedback is a simple resistive divider network: the feedback resistor and the input resistor meet at the input of the opamp rather than internally. I think amplifier designers need to take a strong look at this! Using one of the transistors in a differential pair at the input of the amp is for the birds. In a tube amp, the amps that use highly linear tubes where the feedback is applied are not surprisingly some of the better sounding tube amps that use feedback.

If you don't have problems with phase margin, if you have enough Gain Bandwidth Product to support the feedback at all audio frequencies, this approach will allow you to run much higher levels of feedback without the usual caveats that have plagued its application in the last 70 years...