or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Quality Power
New Posts  All Forums:Forum Nav:

Quality Power - Page 11

post #301 of 637
Quote:
Originally Posted by mtn-tech View Post

Of course one of the claims of Class A amps is that they have less crossover distortion, but good Class AB amplifiers have distortion + noise less than 0.005% which is so far below the threshold of audibility that it can be ignored. I thought that the other claim of Class A amps was a simpler circuit topology - that Krell circuitry certainly doesn't sound simple.

One question that I have is about crossover distortion is "fully balanced output" amplifiers or "bridged" amplifiers - where one amplifier drives the signal positive to the positive output terminal and another amplifier drives the signal negative to the negative output terminal - both amplifiers idle at zero and each drives half the voltage in opposite directions when amplifying a signal - this type of amplifier wouldn't have crossover distortion either - would it?

Unless I'm mistaken, your second paragraph still describes A/B output. The amp will presumably employ means to squash any crossover distortion to well under audible limits as you describe in your first paragraph.

The only types that completely avoid crossover distortion that I'm aware of would be real single ended, or something unconventional like Peter Walker's current dumping amps.
Edited by Wayne Highwood - 11/1/13 at 8:25am
post #302 of 637
Quote:
Originally Posted by mtn-tech View Post

Quote:
Originally Posted by whoaru99 View Post

The Owner's Manual describes the function in non techincal terms.

Which are:

"
BIAS LEVEL METER

The KSA amplifiers utilize Sustained Plateau Biasing, a proprietary KRELL
design. The Bias display meter on the front panel of the amplifier displays
which level of bias the amplifier is utilizing. Unlike sliding bias designs,
Sustained Plateau Bias only changes level when the demand situation
relative to input and source material changes. The signal passes through a
section called an Anticipator circuit. Once the level of bias is determined by
the Anticipator circuit, it automatically biases the output ~f the amplifier
before the signal is passed through the output stage. The Anticipator circuit
is several times faster than the output stage itself. This technique insures
all signal is passed through the amplifier in pure Class A operation. If the
amplifier only requires a low level of Class A bias, the amplifier will bias itself
to use only the amount of Class A power necessary to the demand
parameters. Should the demand parameters exceed the bias level the
amplifier is presently working under, the amplifier will increase its bias to
the next level to allow all signal to come through in Class A operation. If the
demand parameters are less then the bias level indicates for more than 15-
20 seconds, the amplifier will drop to the next lowest bias level.

The bias levels are indicated by appropriate LEDs in the faceplate meter
display as they are activated.

Should the demand parameters reach a point at which the temperature of
the external heatsink reaches approximately 80 degrees Celsius, the last two
levels of bias 3 and 4 will be disabled until the heatsink temperature drops
to a safe operating level. The overall output power of the amplifier is not
changed when this occurs, only the amount of class A power. You will notice
the LEDs corresponding to the top two levels of bias will not illuminate. The
LEDs for levels 3 and 4.will come back on when the heatsink temperature
is at safe operating levels. This temperature fluctuation will not harm the
amplifier. It is provided for consumer safety in a home listening environment.
"
Quote:
Apparently the amp uses what Krell dubs an "Anticipator" circuit to monitor the input signal and predicts bias needed to remain in Class A, then reacts accordingly.

Of course one of the claims of Class A amps is that they have less crossover distortion, but good Class AB amplifiers have distortion + noise less than 0.005% which is so far below the threshold of audibility that it can be ignored. I thought that the other claim of Class A amps was a simpler circuit topology - that Krell circuitry certainly doesn't sound simple.

This appears to yet another situation where common wisdom doesn't tell exactly the whole story. I know for a fact that Class A operation does not give the lowest possible nonlinear distortion based on my own experiences with SS power amplifiers on the test bench, and two of the most highly regarded authors of recent books on this topic appear to agree with me. My experiences and that of these two learned gentlemen being Douglas Self and Robert Cordell is that there is a optimum and fairly minimum amount of bias that minimizes distortion and this amount of bias is the same for any reasonable output power. Here is some evidence:



This figure from Self''s "Audio Power Amplifier Design Handbook" shows three conditions: Underbias, optimal bias, and overbias. Obviously Class A is an extreme case of overbias. Both underbias and overbias cause more distortion than optimal bias.



Cordell goes on to say in "Designing Audio Amplifiers": "Bias stability is important to class AB output stages because crossover distortion is sensitive to the quiescent bias current Iq." and clearly states that bias stability, that is maintaining the same bias level regardless of other changes in the operation of the amplifier including temperature and power output level is highly desirable. This agrees with my own experiments.

Therefore a scheme for varying bias based on expected amplifier output is a solution looking for a problem.
Quote:
One question that I have is about crossover distortion is "fully balanced output" amplifiers or "bridged" amplifiers - where one amplifier drives the signal positive to the positive output terminal and another amplifier drives the signal negative to the negative output terminal - both amplifiers idle at zero and each drives half the voltage in opposite directions when amplifying a signal - this type of amplifier wouldn't have crossover distortion either - would it?

If power amplifiers could be truly identical and symmetrical then the distortion in the signal that bridged nominally identical amplifiers delivered to their load would cancel out to zero. Real world power amplifiers are never perfectly identical and/or symmetrical so the load for bridged amplifiers always seems to see a distorted signal. Since bridging significantly increases the current delivered by each output stage at the higher output levels that bridging enables, and generally drawing more current from a power amplifier usually increases its distortion, bridged amplifier have more distortion than the same amplifier run as a stereo amplifier.
post #303 of 637
And bridging roughly doubles the output impedance, making them typically less stable into lower-impedance loads. Most bridge designs do not handle as low a load as their "regular" brethren. Check out the specs on the majority of bridgeable amps and you'll find e.g. a 2-ohm amp is rated for 4 ohms when bridged.
post #304 of 637
^^^
What's your point?
post #305 of 637
^^attack of the spambots again.... classic just signed up 2 posts per usual, and always ends with.thanks

EDIT... wow you mods are on top of it! spammer gone! thanks
Edited by ljmart - 11/2/13 at 11:03pm
post #306 of 637
Does avsforum.com match up IPs? I've been on other forums that filter "new" members this way or is that obsolete now? I'm not up on my spambots in any case....seems like someone who actually read some of the background in any case....
post #307 of 637
Quote:
Originally Posted by whoaru99 View Post

^^^
What's your point?

Uh, pardon me? The question was about bridging amps, I was hoping to provide a little more info (?)
post #308 of 637
Quote:
Originally Posted by DonH50 View Post

Quote:
Originally Posted by whoaru99 View Post

^^^
What's your point?

Uh, pardon me? The question was about bridging amps, I was hoping to provide a little more info (?)

fyi, wasn't directed at u. last night spambot got inbetween your posts, which is now removed!
post #309 of 637
Yes, that's correct, wasn't directed at you, DonH50.

I didn't immediately recognize the (now deleted) spambot post. It was much more subtle than I'm accustom to seeing.
post #310 of 637
Oops, sorry guys, thanks. - Don
post #311 of 637
Please correct me if I'm wrong, but a simple THD vd Power curve is only done at 1 frequency. Is this good enough to compare distortion across the frequency spectrum?
post #312 of 637
Quote:
Originally Posted by Heinrich S View Post

a simple THD vd Power curve is only done at 1 frequency. Is this good enough to compare distortion across the frequency spectrum?

You are correct, that is not sufficient. If I didn't already link my recent video explaining audio specs and measurements, here it is:

AES Damn Lies

--Ethan
post #313 of 637
Quote:
Originally Posted by Heinrich S View Post

Please correct me if I'm wrong, but a simple THD vd Power curve is only done at 1 frequency. Is this good enough to compare distortion across the frequency spectrum?


Or said another way: A THD curve taken at 1 KHz does not tell you about the THD at 20 Hz, right?

This is one of those questions that makes me shake my head. It's like asking "If you have no information about the weather in Hawaii, do you know what the weather is like in Hawaii?". Well, you probably can't go too wrong saying that the weather is Hawaii is nice, but that's really just a guess..

What about this one? A THD curve taken at 1 KHz does not tell you about the THD at 2 KHz, right? Well, yes and no. If you actually do the experiment, or look at someone else's data, you find that the distortion at 2 KHz is rarely much different than the distortion at 1 KHz

Furthermore, if you understand where distoriton in power amps comes from, higher distortion at 20 Hz is usually caused by the power supply being less capable for that kind of signal.

Higher distortion at 20 KHz comes from two sources: .

(1) The amplifier usually has less gain at 20 KHz and so there is less surplus gain to support the corrective benefits of negative feedback.

(2) Many ampliifers have a small inductor in series with their output that improves stability at the cost of about 5% loss in maximum output. In some cases this has even caused clipping at 20 KHz when the specified power was too close to the amp's real world limits.
Edited by arnyk - 1/20/14 at 1:50pm
post #314 of 637
So then the noise and distortion as shown in S&V doesn't really show anything useful?
post #315 of 637
Quote:
Originally Posted by Heinrich S View Post

So then the noise and distortion as shown in S&V doesn't really show anything useful?

It provides you with an estimate of power output at middle frequencies. Power output at 20 and 20 KHz is usually from 0.5 to 1 dB less, all other things being equal.
post #316 of 637
Quote:
Please correct me if I'm wrong, but a simple THD vd Power curve is only done at 1 frequency. Is this good enough to compare distortion across the frequency spectrum?
But this is true of all measurements. They can never tell you everything. They are only indicators of a component's capabilities.

But certain indicators have been adopted as standard precisely because they are indicative of broader performance capabilities. An amp that can deliver 100w continuously into a static 8 ohm load is certainly powerful enough to drive >99.9% of all nominal 8-ohm speakers. An amp with 0.01% THD at 1 kHz is not going to have 3% THD at some other frequency unless it is very, very broken.
post #317 of 637
Quote:
Originally Posted by Heinrich S View Post

Please correct me if I'm wrong, but a simple THD vd Power curve is only done at 1 frequency. Is this good enough to compare distortion across the frequency spectrum?
As noted, the standard measurement is done at one frequency. Hometheater HiFi shows measurements at one or two power levels and full frequency range. Here are some examples:

rotel-rb-1092-amplifier-thd-vs-frequency-5-volts.gif

Looks like it rises from 0.05 to 0.15 or 3X increase.

emotiva-power-amplifier-thd%2Bn-vs.frequency.gif

Looks like it rises from 0.005 to 0.4 or an increase of 80X. Note that this is specifying the input voltage but not output so not sure if it is the same as others.

mark-levinson-436-amplifier-thd+n-vs-frequency-5-volts.gif

Looks like it is as neutral as one can get.

lamm-m22-amplifier-thd-vs-frequency-5-volts.gif

Another neutral one to 20 Khz.

Just picked these at random. Feel free to search for others yourself.

I can also measure some of the AVRs I have if there is interest.
post #318 of 637
Quote:
Originally Posted by amirm View Post

Hometheater HiFi shows measurements at one or two power levels and full frequency range. Here are some examples:

Excellent Amir.

--Ethan
post #319 of 637
Quote:
Originally Posted by amirm View Post

As noted, the standard measurement is done at one frequency. Hometheater HiFi shows measurements at one or two power levels and full frequency range. Here are some examples:


emotiva-power-amplifier-thd%2Bn-vs.frequency.gif

Looks like it rises from 0.005 to 0.4 or an increase of 80X. Note that this is specifying the input voltage but not output so not sure if it is the same as others.

I'd think not the same. Emo amps are typ. 29dB gain; it would seem to be at considerably higher output voltage than the others.
post #320 of 637
I have been reading widely across the forums in regard to my interest in adding a power amp to my system.

I knew that some underlying theory would be critical and so I just completed reading this entire thread.

My current system:

B&W 683

Outlaw RR 2150 (100 watts per channel)

Rega P3

NAD CD

There are many, many posts that suggest the 683s are "power hungry" although they are rated at the top end for 200 watts and suggesting amps in the 250 watt + category.

Yet, after reading this thread, and noting that I am not having issues with clipping, I am doubtful that more power will produce a fuller and smoother dynamic experience. As noted, I am not of the view that the sound is lacking pure volume nor is it distorting. Instead, I am dissatisfied by the overall presence of the music (mostly late 50s jazz and some indy/alt rock) and the consistency of the sound. I am not looking for louder...

What I am taking away though, is that with systems at my level, there is little or no need to buy large power amps and that money is better spent elsewhere. That my $600 Outlaw (which I actually bought, new, at a silent auction for $150), will sound about the same as a $5,000 amp/pre-amp combo. It is only with more demanding speakers do such investments make sense.

Am I missing something here or have I read correctly?
post #321 of 637
Quote:
Originally Posted by thr61 View Post

What I am taking away though, is that with systems at my level, there is little or no need to buy large power amps and that money is better spent elsewhere.

Yes, exactly. If your system plays loud enough while remaining clean, then buying a more powerful amplifier will make no difference. Since you're still unsatisfied, I suggest you consider room acoustics, which have vastly more influence on sound quality than what model amplifier you use. This short article is mainly about home recording, but all the same principles apply to hi-fi and home theater acoustics too:

Acoustic Basics

--Ethan
post #322 of 637
Quote:
Originally Posted by Ethan Winer View Post

Yes, exactly. If your system plays loud enough while remaining clean, then buying a more powerful amplifier will make no difference. Since you're still unsatisfied, I suggest you consider room acoustics, which have vastly more influence on sound quality than what model amplifier you use. This short article is mainly about home recording, but all the same principles apply to hi-fi and home theater acoustics too:

Acoustic Basics

--Ethan

Thank you for validating and the link. I will spend some time on acoustics and see what happens. In my reading and research, it seems that my components are fairly evenly matched. I remain a bit confused about the many claims I have read and been told that the 683s will "sound better" with more power, since this thread clearly repudiates thinking along these lines.

To go further, is it accurate to say that the role of more powerful and sophisticated power systems are only in relation to the speakers they are driving? Meaning that for me to really "need" more power, I would need to change speakers to something even more advanced?

Again, thanks for the viewpoint and saving me $1,000+ dollars!
post #323 of 637
When I see the phrase "hungry for power" I immediately move on.
post #324 of 637
Quote:
Originally Posted by Marantz guy View Post

When I see the phrase "hungry for power" I immediately move on.

From the opinion or the product?
post #325 of 637
Quote:
Originally Posted by thr61 View Post

Thank you for validating and the link. I will spend some time on acoustics and see what happens. In my reading and research, it seems that my components are fairly evenly matched. I remain a bit confused about the many claims I have read and been told that the 683s will "sound better" with more power, since this thread clearly repudiates thinking along these lines.

To go further, is it accurate to say that the role of more powerful and sophisticated power systems are only in relation to the speakers they are driving? Meaning that for me to really "need" more power, I would need to change speakers to something even more advanced?

Again, thanks for the viewpoint and saving me $1,000+ dollars!
The dependency is on what music you listen to (i.e. its dynamic range), the size of your room, where you sit in the room, the speaker and its directivity and finally your hearing sensitivity. The best test I know is to play something with very high dynamic range and keep turning up the volume while listening to the bass. Does it get thin at some point? And is that point above your comfort level? If it is, then you don't need more power. But if it is, you do smile.gif. No one else can answer this question for you because they can't assess the variables I just mentioned.
post #326 of 637
Quote:
From the opinion or the product?
From the people expressing the opinion.
post #327 of 637
Quote:
Originally Posted by thr61 View Post

From the opinion or the product?

The opinion.
post #328 of 637
Quote:
Originally Posted by Marantz guy View Post

When I see the phrase "hungry for power" I immediately move on.

It is clearly one of those phrases that seems to mean whatever the author wants it to mean.

To me it means something like a speaker with 83 dB/W sensitivity and an impedance curve that is below 4 ohms and stays there without coming up like most of them do.



(no make or model or link because its a fiction that I made up)

What I find it often being attached to is speakers with 90+ dB/W sensitivity and an impedance curve that generally stays above 4 ohm and often goes well beyond that.



As the discussion evolves I often find someone who has money burning a hole in his pocket, who thinks that more expensive amplifiers are the solution to every audio problem, and who has not much of a clue about the tremendous importance of room acoustics.
post #329 of 637
You are far more charitable than I.
post #330 of 637
Way easier to swap in an amp than perform room analysis and put up room treatments... Back in the dark ages I lost a sale for the store and really PO'd my boss when I told the client that changing from ML to Krell amps would do nothing compared to fixing the acoustical disaster that was his listening room.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Audio theory, Setup and Chat
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Quality Power