Lars Clausen said:
Check HIGH EFFICIENCY AUDIO POWER AMPLIFIERS from Ronan van der Zee. See http://www.ub.utwente.nl/webdocs/el/1/t000000d.pdf
In the first chapters he discusses amplifier efficiency and the characteristics of audio signals, w.r.t. average amplitude distribution in music. He also discusses some test signals.
SS A to VT A? Yes, but only once and it was with an early version of one of the Zen amps through some Tannoy 15's. Still no banana and I've done much better tube gear since then. I have a heap of FETs coming so I'm going to try again.noah katz said:
The LF on the system I'm building will be SS D, but that's only sub 100Hz where I care a lot less and the power is nice to be able to chew up with Linkwitz Transforms.
If I hear D or some other topology better what I have, I'll change, as I'm not hugely attached to one particular technology. Can still build a heap of guitar amps with what I have on hand and that'll keep me occupied until I die.
I just feel that for the energy usage difference for most people, justifying D over something else is fairly irrelevant in terms of the rest of the world. I'm also still not convinced re global warming - wasn't too long in my memory that we were headed for a new ice age. GW still has a Y2K smell about it, but I'm not getting into the argument. Matter of fact, I'm going to go downstairs and fire up my 6 litre V8 and do some donuts just for the hell of it.
Yeehahhhhhhh!!!!
I beg everyone. Please don't feed the trolls. 
I think this thread started with an important topic.
For me green is important. Every little bit helps.
Class-d is efficient, economical, and transparent.
In thinking of the first charts, The centers of measurement should probably be compared. I have a feeling that typical wattage is way down - below 1w for most listening. This isn't in the high efficiency curve for a high power class-d. Unless, the whole amp can act like a 1w amp when only 1w is needed. Maybe digital buffering could be done to allow the amp and power supply to only be a high power amp, at the times it needs to be - not all the time?
I think this thread started with an important topic.
For me green is important. Every little bit helps.
Class-d is efficient, economical, and transparent.
In thinking of the first charts, The centers of measurement should probably be compared. I have a feeling that typical wattage is way down - below 1w for most listening. This isn't in the high efficiency curve for a high power class-d. Unless, the whole amp can act like a 1w amp when only 1w is needed. Maybe digital buffering could be done to allow the amp and power supply to only be a high power amp, at the times it needs to be - not all the time?
Give one of these a listen
Hi, Noah:
My favorite tube amp was a triode-wired Quicksilver. I sold it in order to be able to purchase a Class A SET, and it sounded wonderful with a BLH. However, I eventually ditched it in disgust, as I was always having to repair it or have it repaired. I then tried a chip amp (a copy of Peter Daniel's Patek), and was surprised at how good it sounded. However, I kept hearing about something called a Sym-a-sym on the chip amp forum, and how a chip amp really couldn't hold a candle to it. Well, I built a pair of Sym-a-sym monoblocs (http://www.diyaudio.com/forums/showthread.php?postid=1312943#post1312943, with good pics beginning at http://www.diyaudio.com/forums/showthread.php?postid=1313813#post1313813), and paired them up with a pair of PAWOs (http://www.diyaudio.com/forums/showthread.php?postid=1315177#post1315177) and I'm never going back to tubes.
George
noah katz said:
Hi, Noah:
My favorite tube amp was a triode-wired Quicksilver. I sold it in order to be able to purchase a Class A SET, and it sounded wonderful with a BLH. However, I eventually ditched it in disgust, as I was always having to repair it or have it repaired. I then tried a chip amp (a copy of Peter Daniel's Patek), and was surprised at how good it sounded. However, I kept hearing about something called a Sym-a-sym on the chip amp forum, and how a chip amp really couldn't hold a candle to it. Well, I built a pair of Sym-a-sym monoblocs (http://www.diyaudio.com/forums/showthread.php?postid=1312943#post1312943, with good pics beginning at http://www.diyaudio.com/forums/showthread.php?postid=1313813#post1313813), and paired them up with a pair of PAWOs (http://www.diyaudio.com/forums/showthread.php?postid=1315177#post1315177) and I'm never going back to tubes.
George
Member
Joined 2003
George,
Thanks for the links.
DcibeL,
"Like combining Class-D and Class-H? I like the idea, but would a class-D amp suffer from the switching noise that has plagued the class-H topoligy? "
Yes, that's the idea.
I wasn't aware that class H had a noise problem. Isn't that what Carver and NAD designs are?
Is it audible, or just visible on a 'scope?
Thanks for the links.
DcibeL,
"Like combining Class-D and Class-H? I like the idea, but would a class-D amp suffer from the switching noise that has plagued the class-H topoligy? "
Yes, that's the idea.
I wasn't aware that class H had a noise problem. Isn't that what Carver and NAD designs are?
Is it audible, or just visible on a 'scope?
Sound quality aside:
Class D
After a little research of Class D modules(not including PS losses):
At idle:
Brand A: less than or equal to 14w/module
Brand C: 6.5watts/module to 10w/module
(Values for Brand A read from graph so are approximates)
Class A or AB draw considerably more power at idle than a Class D.
Again, power supply losses will increase this (35watts); switch-mode
will be more efficient than linear power supplies.
At 50% efficiency Brand A module:
14 watts total (7watts into 8ohms) [7watts used by amplifier]
26 watts total (13watts into 4ohms) [13watts used by amplifier]
At 80% efficiency Brand A module:
85 watts total (68watts into 8ohms)[17watts used by amplifier]
156.25 watts total (125watts into 4ohms)[31.25watts used by amplifier]
At 92% efficiency Brand A module:
163 watts total (150watts into 8ohms) [13watts used by amplifier]
298.9 watts total (275watts into 4ohms) [23.9watts used by amplifier]
Class A
At Elliott Sound Products this discussion covers Class A amplifiers
having an efficiencies between 13% to 40% at full power; efficiency
will fall as power output is less than maximum.
http://sound.westhost.com/efficiency.htm
A 35watt Class A amplifier running at full power:
At 13% efficiency:
269 watts total (35watts into 8ohms)[234watts used by amplifier]
At 40% efficiency:
87.5 watts total (35watts into 8ohms)[52.5watts used by amplifier]
Efficiency will be lower as power output is lowered.
Class AB
Class AB amplifiers will be in the range of 35% to 55%; theoretical
maximum of close to 78% isn't achievable due to losses in power
supply and circuit inefficiencies.
http://en.wikipedia.org/wiki/Amplifier
A 40 watts Class AB amplifier at full power:
At 35% efficiency:
114 watts total (40watts into 8ohms)[74watts used by amplifier]
At 55% efficiency:
73 watts total (40watts into 8ohms)[33watts used by amplifier]
Conclusions
How 'Green' is Class D compared to other amplifier classes?
Class D offers the most efficient power transfer ratio of all the
classes; at low wattages a Class D amplifier efficiency drops to
less than 10% (maybe even lower).
Class A and Class AB amplifiers also have poor efficiency ratings
at very low power output; these values can easily fall below
2% at 1 watt output.
Overall, Class D amplifiers consume less idle power than Class A or
Class AB amplifiers at idle; and most Class D amplifiers use less
power internally to output full power than Class A or Class AB
amplifiers use at their idle.
From a pure 'watts used' standpoint: Class D is 'greener' than Class A
or Class AB amplifiers at idle. With increased power output, all
amplifier classes increase efficiency as they get closer to their
maximum power output. I would rather pay the electric bill for
a Class D amplifier plugged in 24/7 than any other amplifier class.
Class D amplifiers will be most efficient if using switch-mode
power supplies; Class A and Class AB amplifiers could increase
efficiency by using switch-mode power supplies.
As far as electronic components used in the different classes of
amplifiers: they both use chemicals in their making that you would
not want in your house; they also use about the same amount of
raw materials -- except if using large transformers (more copper).
The 'greenest' way to go would be to use the smallest, most
efficient amplifier for the task that sounds good to you. A close
look at loudspeaker efficiency shows that they are even worse
than Class A amplifiers(watts to dB SPL); improvement could be made
in this category.
Class D
After a little research of Class D modules(not including PS losses):
At idle:
Brand A: less than or equal to 14w/module
Brand C: 6.5watts/module to 10w/module
(Values for Brand A read from graph so are approximates)
Class A or AB draw considerably more power at idle than a Class D.
Again, power supply losses will increase this (35watts); switch-mode
will be more efficient than linear power supplies.
At 50% efficiency Brand A module:
14 watts total (7watts into 8ohms) [7watts used by amplifier]
26 watts total (13watts into 4ohms) [13watts used by amplifier]
At 80% efficiency Brand A module:
85 watts total (68watts into 8ohms)[17watts used by amplifier]
156.25 watts total (125watts into 4ohms)[31.25watts used by amplifier]
At 92% efficiency Brand A module:
163 watts total (150watts into 8ohms) [13watts used by amplifier]
298.9 watts total (275watts into 4ohms) [23.9watts used by amplifier]
Class A
At Elliott Sound Products this discussion covers Class A amplifiers
having an efficiencies between 13% to 40% at full power; efficiency
will fall as power output is less than maximum.
http://sound.westhost.com/efficiency.htm
A 35watt Class A amplifier running at full power:
At 13% efficiency:
269 watts total (35watts into 8ohms)[234watts used by amplifier]
At 40% efficiency:
87.5 watts total (35watts into 8ohms)[52.5watts used by amplifier]
Efficiency will be lower as power output is lowered.
Class AB
Class AB amplifiers will be in the range of 35% to 55%; theoretical
maximum of close to 78% isn't achievable due to losses in power
supply and circuit inefficiencies.
http://en.wikipedia.org/wiki/Amplifier
A 40 watts Class AB amplifier at full power:
At 35% efficiency:
114 watts total (40watts into 8ohms)[74watts used by amplifier]
At 55% efficiency:
73 watts total (40watts into 8ohms)[33watts used by amplifier]
Conclusions
How 'Green' is Class D compared to other amplifier classes?
Class D offers the most efficient power transfer ratio of all the
classes; at low wattages a Class D amplifier efficiency drops to
less than 10% (maybe even lower).
Class A and Class AB amplifiers also have poor efficiency ratings
at very low power output; these values can easily fall below
2% at 1 watt output.
Overall, Class D amplifiers consume less idle power than Class A or
Class AB amplifiers at idle; and most Class D amplifiers use less
power internally to output full power than Class A or Class AB
amplifiers use at their idle.
From a pure 'watts used' standpoint: Class D is 'greener' than Class A
or Class AB amplifiers at idle. With increased power output, all
amplifier classes increase efficiency as they get closer to their
maximum power output. I would rather pay the electric bill for
a Class D amplifier plugged in 24/7 than any other amplifier class.
Class D amplifiers will be most efficient if using switch-mode
power supplies; Class A and Class AB amplifiers could increase
efficiency by using switch-mode power supplies.
As far as electronic components used in the different classes of
amplifiers: they both use chemicals in their making that you would
not want in your house; they also use about the same amount of
raw materials -- except if using large transformers (more copper).
The 'greenest' way to go would be to use the smallest, most
efficient amplifier for the task that sounds good to you. A close
look at loudspeaker efficiency shows that they are even worse
than Class A amplifiers(watts to dB SPL); improvement could be made
in this category.
From the Mackie M-2000, M-3000, M-4000 advertising. These are linear PA-amps.
"Adaptive Slewing circuitry allows M-Series amplifiers to pass the fastest transient attacks—like kick and snare drum, or any bass player trying to imitate Marcus Miller—while minimizing Class-H switching distortion. As a result, the M-Series have the lowest distortion of any Class-H amplifier on the market. And no matter how hard and fast you thrash, the M-Series amps are gonna be a couple of steps ahead, making sure the output is clean."
When using a digital storage system for musical reproduction such as CDP, DVD, hard disc it seems possible to store the information in a buffer there analyzing the amplitudes coming and calculate the need for power; given the efficiency of the loudspeaker and average listening level. If the music leaves the system with one second of delay, I don´t care.
If multiple amps are used (bi- or triamping ) the amps have to receive their signal in a synchronized manner, or even at slightly different times to compensate for differences in the drivers phase response. Goldmund of Switzerland have done some promising research about compensating (in the time domain) both loudspeakers with conventional and active crossovers.
There may have been some intention of promotion when starting this thread, but let´s forget about this if the discussion has clearified a few other things...(or made us more confused).
"Adaptive Slewing circuitry allows M-Series amplifiers to pass the fastest transient attacks—like kick and snare drum, or any bass player trying to imitate Marcus Miller—while minimizing Class-H switching distortion. As a result, the M-Series have the lowest distortion of any Class-H amplifier on the market. And no matter how hard and fast you thrash, the M-Series amps are gonna be a couple of steps ahead, making sure the output is clean."
When using a digital storage system for musical reproduction such as CDP, DVD, hard disc it seems possible to store the information in a buffer there analyzing the amplitudes coming and calculate the need for power; given the efficiency of the loudspeaker and average listening level. If the music leaves the system with one second of delay, I don´t care.
If multiple amps are used (bi- or triamping ) the amps have to receive their signal in a synchronized manner, or even at slightly different times to compensate for differences in the drivers phase response. Goldmund of Switzerland have done some promising research about compensating (in the time domain) both loudspeakers with conventional and active crossovers.
There may have been some intention of promotion when starting this thread, but let´s forget about this if the discussion has clearified a few other things...(or made us more confused).
tosspot?
I'm sorry, I came into the thread late and after further reading
I see that the original poster does seem affiliated with a company.
But, that shouldn't make the original question any less valid.
I think any discussion of quality of sound from the different classes
of amplifiers should be in their own thread. Some good thread
topics I would like to see would be:
Class A tubes sound better than Class D amplifiers.
Class AB amplifiers use less complex circuits than Class D amplifiers.
Class AB amplifiers can be made more efficient.
Class D amplifiers are used in at least one stage of the music
we hear everyday.
I'm sorry, I came into the thread late and after further reading
I see that the original poster does seem affiliated with a company.
But, that shouldn't make the original question any less valid.
I think any discussion of quality of sound from the different classes
of amplifiers should be in their own thread. Some good thread
topics I would like to see would be:
Class A tubes sound better than Class D amplifiers.
Class AB amplifiers use less complex circuits than Class D amplifiers.
Class AB amplifiers can be made more efficient.
Class D amplifiers are used in at least one stage of the music
we hear everyday.
The golden mean said:From the Mackie M-2000, M-3000, M-4000 advertising. These are linear PA-amps.
When using a digital storage system for musical reproduction such as CDP, DVD, hard disc it seems possible to store the information in a buffer there analyzing the amplitudes coming and calculate the need for power; given the efficiency of the loudspeaker and average listening level. If the music leaves the system with one second of delay, I don´t care.
That is a real benefit of being all digital. . .the amp can get ready
for the upcoming signal. A big plus to increasing efficiency. I wish
I had though of it. There might not even be a 1 second delay. . .it might
be less that 50mS given how fast processing is today.
And with a linear power supply. . . voltage rail switching. . . slick
technology. . .would certainly make the linear power supply more
efficient.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.