Amplifier class
-
LIVE: Tech Talk Podcast with Scott Wilkinson, Episode 8 Click here for details.
- Status
1 - 19 of 19 Posts
Joined
·
1,782 Posts
I typed "amplifier class" into google and the 3rd link that came up was this: http://www.eatel.net/~amptech/elecdisc/ampclass.htm
It is for car audio, but amplifier classes don't change based on home vs car.
Long story short, the differences are efficiency & distortion. From least efficient and usually least distortion (A) to most (D). There are other more esoteric classes, but they are not quite as common as A, AB & D. Nearly all amplifiers you'll find in home theaters are AB amps. Class A amps tend to be hellishly expensive, not to mention space heaters. Class B amps have big distortion, so AB is the natural choice for most.
Sorny
It is for car audio, but amplifier classes don't change based on home vs car.
Long story short, the differences are efficiency & distortion. From least efficient and usually least distortion (A) to most (D). There are other more esoteric classes, but they are not quite as common as A, AB & D. Nearly all amplifiers you'll find in home theaters are AB amps. Class A amps tend to be hellishly expensive, not to mention space heaters. Class B amps have big distortion, so AB is the natural choice for most.
Sorny
Joined
·
225 Posts
This could be a lengthy description, but I'll try to keep it simple. It has to do with how the signal is driven by the output stage (transistors or tubes).
In a Class A configuration the output devices are always operating in their active region for the full (positive and negative) cycle of the signal. These are usually considered the best configuration for audio enthusiasts because they don't generate any crossover distortion. Because the output is always active, however, they are the least efficient (power-wise).
In Class B, the output transistor only operates in its active region, or outputs a signal, for one-half (either positive or negative) cycle. Pure class B amplifiers are not applicable to audio.
Most amplifiers operate in Class A-B, which in simplied terms is something like a pair of complementary class B stages in a push-pull configuration and slightly biased into conduction to minimize crossover distortion (the point at which the signal crosses zero, and the transistors switch on/off).
Class C amplifiers are active for less than one-half cycle and also are normally not applicable to audio.
Class D amplifiers introduce a digital aspect by driving the output using a PWM (pulse width modulator). These are highly complex and very efficient designs relative to the above, and can therefore produce more output power from a smaller physical package while generating less heat. But they are somewhat rare and can suffer from distortion if not designed well. The technology has only recently (last 5 years or so) become utilized in actual consumer audio products (both high and low end). Some cheap amplified computer speakers use dedicated off-the-shelf chips that contain small class D amplifiers.
In a Class A configuration the output devices are always operating in their active region for the full (positive and negative) cycle of the signal. These are usually considered the best configuration for audio enthusiasts because they don't generate any crossover distortion. Because the output is always active, however, they are the least efficient (power-wise).
In Class B, the output transistor only operates in its active region, or outputs a signal, for one-half (either positive or negative) cycle. Pure class B amplifiers are not applicable to audio.
Most amplifiers operate in Class A-B, which in simplied terms is something like a pair of complementary class B stages in a push-pull configuration and slightly biased into conduction to minimize crossover distortion (the point at which the signal crosses zero, and the transistors switch on/off).
Class C amplifiers are active for less than one-half cycle and also are normally not applicable to audio.
Class D amplifiers introduce a digital aspect by driving the output using a PWM (pulse width modulator). These are highly complex and very efficient designs relative to the above, and can therefore produce more output power from a smaller physical package while generating less heat. But they are somewhat rare and can suffer from distortion if not designed well. The technology has only recently (last 5 years or so) become utilized in actual consumer audio products (both high and low end). Some cheap amplified computer speakers use dedicated off-the-shelf chips that contain small class D amplifiers.
Class AB amplifiers will certainly have more distortion than class A amps, but this may or may not be important. A class AB amp will only have distortion at very high output levels, and if your speakers have much higher distortion at the same output levels, you won't notice the imperfection in the amp.
The heat generated by high-powered, full class A amps can be enormous.
TacT is making some really expensive class D amps these days.
Tim
The heat generated by high-powered, full class A amps can be enormous.
TacT is making some really expensive class D amps these days.
Tim
Keep in mind that what people are referring to as "distortion" here is actually "open-loop distortion" -- i.e., the distortion introduced by the final power stage in the absence of a negative feedback loop. In reality, today's audio amplifiers (even Class A power amps) all engage a generous amount of negative feedback, which pretty much completely eliminates any distortions -- hence those wonderful specs of less than 0.02% THD+N, etc., regardless of the power stage class.
So, although in theory a Class A amp requires a less amount of negative feedback than Class AB to achieve the same distortion spec and this fact has some benefits in itself, I think that in today's amps the difference is nearly undetectable by either measurements or listening tests. The markedly higher power consumption of Class A amps (even when the amp is idle) pose more problems in today's multi-channel settings.
By the way, a Class D amp (switching, or digital power stage) is often used in car audio power amps and subwoofer amps (where a compact and efficient yet high-power amp is desireble), along with some hybrid analog-digital designs such as the popular BASH amps.
So, although in theory a Class A amp requires a less amount of negative feedback than Class AB to achieve the same distortion spec and this fact has some benefits in itself, I think that in today's amps the difference is nearly undetectable by either measurements or listening tests. The markedly higher power consumption of Class A amps (even when the amp is idle) pose more problems in today's multi-channel settings.
By the way, a Class D amp (switching, or digital power stage) is often used in car audio power amps and subwoofer amps (where a compact and efficient yet high-power amp is desireble), along with some hybrid analog-digital designs such as the popular BASH amps.
"The markedly higher power consumption of Class A amps (even when the amp is idle) pose more problems in today's multi-channel settings."
Damn, sushi, does this mean I have to give up my Krell 6 million watt class A monoblocks? But those heatsinks look so
COOL!
Tim
Damn, sushi, does this mean I have to give up my Krell 6 million watt class A monoblocks? But those heatsinks look so
COOL!
Tim
nutt,
HAHAHA... They must be doubling as room heaters...? Oops! You are also in Texas. Have you checked the air conditioner for the summer? LOL
Phil,
Not necessarily. It all depends on the applications. In reality, 98% of today's receivers/power amps use a Class AB power stage. So, as long as you are shopping for an "ordinary" amp, you do not have much choice. Class A power stage is used in a small number of, well, high-end/exotic models (like Tim's Krell monoblocks! -- though I do not know if even his Krell is truly pure Class A... maybe it is Class A/AB). They are fairly expensive, especially considering the relatively modest power output. Audiophile grade Class D (digital) amps are also very expensive. Besides TacT, Sharp makes some $10K models.
HAHAHA... They must be doubling as room heaters...? Oops! You are also in Texas. Have you checked the air conditioner for the summer? LOL
Phil,
Not necessarily. It all depends on the applications. In reality, 98% of today's receivers/power amps use a Class AB power stage. So, as long as you are shopping for an "ordinary" amp, you do not have much choice. Class A power stage is used in a small number of, well, high-end/exotic models (like Tim's Krell monoblocks! -- though I do not know if even his Krell is truly pure Class A... maybe it is Class A/AB). They are fairly expensive, especially considering the relatively modest power output. Audiophile grade Class D (digital) amps are also very expensive. Besides TacT, Sharp makes some $10K models.
Joined
·
12,408 Posts
Contrary to what was stated above, an AB amplifier will likely have a higher THD at very low levels due to its wider margin for setting the output device's bias voltage level. Due to component cost reasons only a select few of the AVR brands actually have adjustable bias setting @ idle/low levels, instead they used fixed components and the output stage will likely have a wider variance for the bias level settings resulting in slightly higher THD..
Once the output devices are fully turned on/off and conducting, and producing higher output power the THD will decrease until the amplifier is maxed out and driven into clipping. For example, an amplifier may be rated at 0.08% THD 100W, @1 watt the THD may be 0.15% but then drops down to 0.05% as the power is increased up to 100W. When 100W is exceeded, then the THD will increase very rapidly..
Once the output devices are fully turned on/off and conducting, and producing higher output power the THD will decrease until the amplifier is maxed out and driven into clipping. For example, an amplifier may be rated at 0.08% THD 100W, @1 watt the THD may be 0.15% but then drops down to 0.05% as the power is increased up to 100W. When 100W is exceeded, then the THD will increase very rapidly..
Sinitro,
Huh? ALL amps, regardless of class, show higher THD at very low signal levels, and that is solely due to thermal and other noise (hence THD+N), not what you tried to explain (which does not make sense to me at all - sorry). Incidentally, virtually all modern solidstate power stage has one or another mechanism for auto-regulating the bias current/voltage, and so your "concern" is not warranted.
Huh? ALL amps, regardless of class, show higher THD at very low signal levels, and that is solely due to thermal and other noise (hence THD+N), not what you tried to explain (which does not make sense to me at all - sorry). Incidentally, virtually all modern solidstate power stage has one or another mechanism for auto-regulating the bias current/voltage, and so your "concern" is not warranted.
Joined
·
12,408 Posts
Quote:
NO.. not every class I said class AB.
The higher THD @ low power is due to the difference in gain within the output device pair as when one device may turn on before the other device is off. They are not totally matched, conventional AB amplifiers use fixed resistors for this. The fixed resistor value is fine if the output device gains are all matched to one another but this is the exception rather than the rule. By using variable trim resistors, the bias setting can be calibrated on the production line which will compensate for device gain differences. The end result is very audible on high resolution loudspeakers..
| Originally posted by sushi Sinitro, Huh? ALL amps, regardless of class, show higher THD at very low signal levels, and that is solely due to thermal and other noise (hence THD+N), not what you tried to explain (which does not make sense to me at all - sorry). Incidentally, virtually all modern solidstate power stage has one or another mechanism for auto-regulating the bias current/voltage, and so your "concern" is not warranted. |
The higher THD @ low power is due to the difference in gain within the output device pair as when one device may turn on before the other device is off. They are not totally matched, conventional AB amplifiers use fixed resistors for this. The fixed resistor value is fine if the output device gains are all matched to one another but this is the exception rather than the rule. By using variable trim resistors, the bias setting can be calibrated on the production line which will compensate for device gain differences. The end result is very audible on high resolution loudspeakers..
Well, I still disagree, Sinitro.
(1) Your argument will equally apply to Class A amps, if you think through. (2) The imbalance in bias current/voltage, if it is truly a problem, will result in distortions not only at low signal levels, but over the entire signal range. (3) Precisely in order to compensate for the power device mismatch that you mentioned, modern amplifiers engage one or another automatic bias regulator (bootstrap circuit for one). Perhaps the "conventional" amplifier design you mentioned was from '60s and early '70s? (4) The "hypothetical" distortion at low signal levels that you are talking about does not show at all in the actual measurements. The reality still remains that the higher THD+N values at low signal levels are all due to noise, regardless of the power stage class.
Perry, I am not ignoring your question. Just that I am not familiar with that amp at all. Could you link to some literature about it?
(1) Your argument will equally apply to Class A amps, if you think through. (2) The imbalance in bias current/voltage, if it is truly a problem, will result in distortions not only at low signal levels, but over the entire signal range. (3) Precisely in order to compensate for the power device mismatch that you mentioned, modern amplifiers engage one or another automatic bias regulator (bootstrap circuit for one). Perhaps the "conventional" amplifier design you mentioned was from '60s and early '70s? (4) The "hypothetical" distortion at low signal levels that you are talking about does not show at all in the actual measurements. The reality still remains that the higher THD+N values at low signal levels are all due to noise, regardless of the power stage class.
Perry, I am not ignoring your question. Just that I am not familiar with that amp at all. Could you link to some literature about it?
Sushi:
Many thanks for your kind response. You can look up some write-up on it at http://www.adireaudio.com/diy_audio/...ier/dt300.htm.
In fact, I was hoping there might be someone who can evaluate it based personal experience, if not what do its specs promise?
Best rgds
Perry
Many thanks for your kind response. You can look up some write-up on it at http://www.adireaudio.com/diy_audio/...ier/dt300.htm.
In fact, I was hoping there might be someone who can evaluate it based personal experience, if not what do its specs promise?
Best rgds
Perry
Joined
·
7,482 Posts
Keep in mind that manufacturers often consider their Class A/B amps "Class A" amps for marketing reasons ...
Quote:
Looks like alot of story that got passed on and on until it doesn't make sense anymore.
Class A or A/B is bias current on the output transistor. Class G is a power supply feature. They have nothing to do with each other.
For the amp to operate in class-A into 80 watts, each of those practically heatsink-less monoblock would be sucking over 600 watt from the wall and dissepating it as heat at idle.
| Originally posted by billfold I heard the Outlaw M200's were Class A up to 80 Watts and Class G above that...what does that mean?? Is that even right? |
Class A or A/B is bias current on the output transistor. Class G is a power supply feature. They have nothing to do with each other.
For the amp to operate in class-A into 80 watts, each of those practically heatsink-less monoblock would be sucking over 600 watt from the wall and dissepating it as heat at idle.
1 - 19 of 19 Posts
- Status
Join the discussion
