Quality Power - Page 20 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #571 of 637 Old 03-22-2014, 07:22 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Frank Derks View Post


For preamplifiers or dedicated headphone amplifiers class A is still the best output topology. The energy lost is negligible in these cases.

The vast majority of all headphone and earphone amplifiers are in portable equipment that runs off of batteries so the excess energy use by Class A output stages remains a very important consideration.
arnyk is offline  
Sponsored Links
Advertisement
 
post #572 of 637 Old 03-22-2014, 08:08 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
I did design and build my amplifiers but never had the pleasure of a distortion measuring test set. All I could afford was an oscilloscope and a dvm. My AC voltmer at the time was a Phillips GM6012 rolleyes.gif
What I did was measure the voltage difference between the output and the input at the driver stage on a 2 channel scope in differential mode.
With a 10 Hz 2Vpp sine wave across an 1 Ohm load I adjusted the bias to the point where the diff signal had the lowest variation.

Fast forward 15 years later and an affordable distortion measurement can be had in the form of a good pc soundcard and software. Not AP standards but useful.

Agree that for power amplifiers optimal biased class B is more than good enough with modern power BJT's.
Frank Derks is offline  
post #573 of 637 Old 03-22-2014, 09:32 AM
AVS Special Member
 
kiwi2's Avatar
 
Join Date: Nov 2010
Location: New Zealand
Posts: 1,629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 144
Quote:
Originally Posted by Frank Derks View Post

Despite the inefficiency of class A it is still the most linear output topology for an output stage in an amplifier.

If you look at the input then output waveform class D is the most linear...

http://en.wikipedia.org/wiki/Electronic_amplifier#Class_A
kiwi2 is offline  
post #574 of 637 Old 03-22-2014, 09:43 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by kiwi2 View Post

Quote:
Originally Posted by Frank Derks View Post

Despite the inefficiency of class A it is still the most linear output topology for an output stage in an amplifier.

If you look at the input then output waveform class D is the most linear...

http://en.wikipedia.org/wiki/Electronic_amplifier#Class_A

Really? Your reference says:

"Class-D amplifiers have been widely used to control motors, but they are now also used as audio power amplifiers, with some extra circuitry to allow analogue to be converted to a much higher frequency pulse width modulated signal.

High quality class-D audio power amplifiers have now appeared on the market. These designs have been said to rival traditional AB amplifiers in terms of quality. An early use of class-D amplifiers was high-power subwoofer amplifiers in cars. Because subwoofers are generally limited to a bandwidth of no higher than 150 Hz, the switching speed for the amplifier does not have to be as high as for a full range amplifier, allowing simpler designs. Class-D amplifiers for driving subwoofers are relatively inexpensive in comparison to class-AB amplifiers."

I read that as saying that the idea that Class D amplifiers rival traditional AB amplifiers in terms of quality is a controversial point. I know that Class D amplifiers have broken new ground in distortion mechanisms.

Here is an article by a proponent of Class D amplifiers:

http://www.irf.com/technical-info/appnotes/an-1071.pdf

For example:

arnyk is offline  
post #575 of 637 Old 03-22-2014, 09:58 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
Quote:
Originally Posted by kiwi2 View Post

If you look at the input then output waveform class D is the most linear...

http://en.wikipedia.org/wiki/Electronic_amplifier#Class_A

That may be true but there is a price to pay. The feedback loop in a conventional amplifier is just two resistors. The feedback to the differential input pair is instantaneous.
In class D the output must be filtered This introduces a phase shift in the feedback path rendering feedback less effective.
Frank Derks is offline  
post #576 of 637 Old 03-22-2014, 11:57 AM
Member
 
dbrown3611's Avatar
 
Join Date: Sep 2013
Location: The Great Plains of Nebraska, USA
Posts: 42
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 24
Quote:
Originally Posted by arnyk View Post

Really? Your reference says:

"Class-D amplifiers have been widely used to control motors, but they are now also used as audio power amplifiers, with some extra circuitry to allow analogue to be converted to a much higher frequency pulse width modulated signal.

High quality class-D audio power amplifiers have now appeared on the market. These designs have been said to rival traditional AB amplifiers in terms of quality. An early use of class-D amplifiers was high-power subwoofer amplifiers in cars. Because subwoofers are generally limited to a bandwidth of no higher than 150 Hz, the switching speed for the amplifier does not have to be as high as for a full range amplifier, allowing simpler designs. Class-D amplifiers for driving subwoofers are relatively inexpensive in comparison to class-AB amplifiers."

I read that as saying that the idea that Class D amplifiers rival traditional AB amplifiers in terms of quality is a controversial point. I know that Class D amplifiers have broken new ground in distortion mechanisms.

Here is an article by a proponent of Class D amplifiers:

http://www.irf.com/technical-info/appnotes/an-1071.pdf

For example:


That quoted Wiki entry section on Class D amps has no cited references, so it is possibly anecdotal and/or outdated.

Pioneer uses the International Rectifier silicon. It would seem the boys and girls in the design department at Pioneer are very competent in utilizing the IR parts. Witness their D3 amps:
http://www.digitaltrends.com/home-theater/how-pioneer-made-digital-amplification-more-musical/

"Direct Signal Path Design – Pioneer designed the amplifier section to have the shortest audio signal path possible, requiring no EMI filters, feedback loops, impedance switching or current limiters, which results in the elimination of induced coloration, ringing or noise, especially in the high frequency audio band. In addition, the amp uses surface mounted MOSFETs and discrete 7- and 9-channel inputs."

Above was taken from here: http://www.pioneerelectronics.com/PUSA/Press-Room/Home-Entertainment/Pioneer+Expands+Elite+SC+Series+Receivers+Models
dbrown3611 is offline  
post #577 of 637 Old 03-22-2014, 12:14 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
Quote:
Originally Posted by dbrown3611 View Post

That quoted Wiki entry section on Class D amps has no cited references, so it is possibly anecdotal and/or outdated.

Pioneer uses the International Rectifier silicon. It would seem the boys and girls in the design department at Pioneer are very competent in utilizing the IR parts. Witness their D3 amps:
http://www.digitaltrends.com/home-theater/how-pioneer-made-digital-amplification-more-musical/

"Direct Signal Path Design – Pioneer designed the amplifier section to have the shortest audio signal path possible, requiring no EMI filters, feedback loops, impedance switching or current limiters, which results in the elimination of induced coloration, ringing or noise, especially in the high frequency audio band. In addition, the amp uses surface mounted MOSFETs and discrete 7- and 9-channel inputs."

Above was taken from here: http://www.pioneerelectronics.com/PUSA/Press-Room/Home-Entertainment/Pioneer+Expands+Elite+SC+Series+Receivers+Models

Warning, marketing blurb.

"Direct Signal Path Design – Pioneer designed the amplifier section to have the shortest audio signal path possible, requiring no EMI filters, feedback loops, impedance switching or current limiters, which results in the elimination of induced coloration, ringing or noise, especially in the high frequency audio band. In addition, the amp uses surface mounted MOSFETs and discrete 7- and 9-channel inputs."

And we get the picture:

No EMI filters?

Frank Derks is offline  
post #578 of 637 Old 03-22-2014, 12:24 PM
Member
 
dbrown3611's Avatar
 
Join Date: Sep 2013
Location: The Great Plains of Nebraska, USA
Posts: 42
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 24
Quote:
Originally Posted by Frank Derks View Post

Warning, marketing blurb.

"Direct Signal Path Design – Pioneer designed the amplifier section to have the shortest audio signal path possible, requiring no EMI filters, feedback loops, impedance switching or current limiters, which results in the elimination of induced coloration, ringing or noise, especially in the high frequency audio band. In addition, the amp uses surface mounted MOSFETs and discrete 7- and 9-channel inputs."

And we get the picture:

No EMI filters?

You grabbed a pic of the old style board, the new D3 board is the top picture.
http://www.digitaltrends.com/home-theater/how-pioneer-made-digital-amplification-more-musical/
dbrown3611 is offline  
post #579 of 637 Old 03-22-2014, 12:35 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
Quote:
Originally Posted by dbrown3611 View Post

You grabbed a pic of the old style board, the new D3 board is the top picture.
http://www.digitaltrends.com/home-theater/how-pioneer-made-digital-amplification-more-musical/


Oops,

Here is the correct pic. Stil rfi filtering in there, but harder to spot.

Frank Derks is offline  
post #580 of 637 Old 03-22-2014, 12:59 PM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Frank Derks View Post

Quote:
Originally Posted by dbrown3611 View Post

You grabbed a pic of the old style board, the new D3 board is the top picture.
http://www.digitaltrends.com/home-theater/how-pioneer-made-digital-amplification-more-musical/


Oops,

Here is the correct pic. Stil rfi filtering in there, but harder to spot.


Agreed. They went from old-style unencapsulated open winding ferrite-core toroids to whatever they have squirreled away in the 9 little silver and black boxes. They may have kicked the switching frequency up a little, so the coils and caps might be a tad smaller.
arnyk is offline  
post #581 of 637 Old 03-22-2014, 01:24 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
If you follow the link and zoom in on the picture you can see a hint of a coil between the gray and black area of the 'mystery box'.

The marketing speak used is way to suspicious. The word 'musical' is used too often and all the usual audiophile 'horrors' where mentioned.
Frank Derks is offline  
post #582 of 637 Old 03-22-2014, 01:28 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
Frank Derks is offline  
post #583 of 637 Old 03-22-2014, 01:43 PM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Frank Derks View Post

Also the 0.003% distortion mentioned is also rather optimistic...

http://www.hometheaterhifi.com/receivers/receivers-reviews/pioneer-sc68-receiver/page-4-on-the-bench.html

Yes, this picture tells a far less optimistic story, and at only about 3 watts into an 8 ohm load:



High order harmonics through the ninth are clearly visible. This one is dirty enough that it might even be fun to ABX, but I wouldn't hope for too much.
arnyk is offline  
post #584 of 637 Old 03-22-2014, 01:51 PM
 
Jonathan Parker's Avatar
 
Join Date: Mar 2014
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by arnyk View Post


What I mean is that according to the USA Federal Trade Commission (FTC) audio amplifiers for consumer use must be advertised and have power ratings based on measurements that model music as a pure, steady sine wave being delivered to a resistive load. This has a strong influence on the world market. Music has a high crest factor while steady pure sine waves have a very low crest factor. Speakers typically have far higher average impedance than their rated impedance.

These asymmetries between law and reality make audio gear bigger, heavier, and more expensive than it needs to be.

While naive audiophiles lust over power amps that weigh a lot, amps that sound best and produce the most clean power possible at any price point would be smaller and lighter.


The way I see it, if amplifiers were overbuilt to perform well on the test bench as you suggest then they can only perform better with real music and loudspeakers, which should put customers at ease as far as performance goes. Rather go big or go home, as some people say. I don't see the harm in that, in fact I think it makes for a better, more robust design.

Jonathan Parker is offline  
post #585 of 637 Old 03-22-2014, 02:02 PM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Jonathan Parker View Post

Quote:
Originally Posted by arnyk View Post

What I mean is that according to the USA Federal Trade Commission (FTC) audio amplifiers for consumer use must be advertised and have power ratings based on measurements that model music as a pure, steady sine wave being delivered to a resistive load. This has a strong influence on the world market. Music has a high crest factor while steady pure sine waves have a very low crest factor. Speakers typically have far higher average impedance than their rated impedance.


These asymmetries between law and reality make audio gear bigger, heavier, and more expensive than it needs to be.


While naive audiophiles lust over power amps that weigh a lot, amps that sound best and produce the most clean power possible at any price point would be smaller and lighter.

The way I see it, if amplifiers were overbuilt to perform well on the test bench as you suggest then they can only perform better with real music and loudspeakers, which should put customers at ease as far as performance goes. Rather go big or go home, as some people say. I don't see the harm in that.

The way I see it, if Mr. Engineer blows his budget on parts that are just cooling their heels all of the time, the guy down the street might be able to figure out how to spend the same money on parts that work a little harder and make an audible improvement. The market may even figure out that his gear is a little cleaner a little louder. Too bad for you!

I think a power amp with a time/power delivery curve that better duplicates music might be able to crank out about 6 dB more power for about the same parts cost, size and weight. It won't be twice as loud, but it will be enough louder to hear the difference in an ABX test pretty clearly.
arnyk is offline  
post #586 of 637 Old 03-22-2014, 02:16 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
A low bass note sustained for a couple of seconds at full power is what the power supply needs to be able to cope with.
Amplifiers don't discriminate so testing with a 1kHz or 10kHz sine wave at the same power levels makes no difference for an amplifier.
Frank Derks is offline  
post #587 of 637 Old 03-23-2014, 03:16 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Frank Derks View Post

A low bass note sustained for a couple of seconds at full power is what the power supply needs to be able to cope with.
Amplifiers don't discriminate so testing with a 1kHz or 10kHz sine wave at the same power levels makes no difference for an amplifier.

Sustained bass notes aren't anything like pure sine waves. I actually spent a good part of yesterday afternoon looking at sustained pure bass notes just to double-check my previous findings. What I have found over the decades is though bass notes dominate the recording, considerable energy is in other notes that are played at the same time. In addition, the bass notes themselves aren't pure tones at the note's fundamental frequency. It is not unusual for the majority of the energy in the bass note to be in the harmonics of the fundamental. All of these influences work together to considerably increase the crest factor of the music from the 3 dB we see for pure since waves to 10 dB or more. This vastly reduces the average drain on power supplies and the heating if the power transistors and heat sinks serving the output stages. It is also why people can plug power amps capable of putting out thousands of watts of sine waves into a single ordinary house outlet and never blow the circuit breaker as long as they play recordings of regular music or drama.
arnyk is offline  
post #588 of 637 Old 03-23-2014, 07:55 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,768
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 493 Post(s)
Liked: 323
Quote:
Originally Posted by arnyk View Post

Sustained bass notes aren't anything like pure sine waves. I actually spent a good part of yesterday afternoon looking at sustained pure bass notes just to double-check my previous findings. What I have found over the decades is though bass notes dominate the recording, considerable energy is in other notes that are played at the same time. In addition, the bass notes themselves aren't pure tones at the note's fundamental frequency. It is not unusual for the majority of the energy in the bass note to be in the harmonics of the fundamental. All of these influences work together to considerably increase the crest factor of the music from the 3 dB we see for pure since waves to 10 dB or more.
Can you please post the waveform and the method you used to compute the crest factor?

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is online now  
post #589 of 637 Old 03-23-2014, 08:49 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by amirm View Post

Quote:
Originally Posted by arnyk View Post

Sustained bass notes aren't anything like pure sine waves. I actually spent a good part of yesterday afternoon looking at sustained pure bass notes just to double-check my previous findings. What I have found over the decades is though bass notes dominate the recording, considerable energy is in other notes that are played at the same time. In addition, the bass notes themselves aren't pure tones at the note's fundamental frequency. It is not unusual for the majority of the energy in the bass note to be in the harmonics of the fundamental. All of these influences work together to considerably increase the crest factor of the music from the 3 dB we see for pure since waves to 10 dB or more.
Can you please post the waveform and the method you used to compute the crest factor?

Sure:




However the above in my opinion represents a fool's mission because it can't be heard and its spectral content was not asked for.

In this case crest factor was calculated from Average RMS power and Peak Amplitude. The display and analysis tool was Adobe Audition. Note that the actual crest factor is more than 10 dB, which is what I meant by "10 dB or more".

I expect major hypercritical nit-picking to follow! ;-)
arnyk is offline  
post #590 of 637 Old 03-23-2014, 08:58 AM
 
Jonathan Parker's Avatar
 
Join Date: Mar 2014
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by arnyk View Post


Sustained bass notes aren't anything like pure sine waves. I actually spent a good part of yesterday afternoon looking at sustained pure bass notes just to double-check my previous findings. What I have found over the decades is though bass notes dominate the recording, considerable energy is in other notes that are played at the same time. In addition, the bass notes themselves aren't pure tones at the note's fundamental frequency. It is not unusual for the majority of the energy in the bass note to be in the harmonics of the fundamental. All of these influences work together to considerably increase the crest factor of the music from the 3 dB we see for pure since waves to 10 dB or more. This vastly reduces the average drain on power supplies and the heating if the power transistors and heat sinks serving the output stages. It is also why people can plug power amps capable of putting out thousands of watts of sine waves into a single ordinary house outlet and never blow the circuit breaker as long as they play recordings of regular music or drama.


So you are saying that a high crest factor allows the power supplies to cool down so they can handle higher instantaneous peaks?

Jonathan Parker is offline  
post #591 of 637 Old 03-23-2014, 09:19 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Jonathan Parker View Post

Quote:
Originally Posted by arnyk View Post

Sustained bass notes aren't anything like pure sine waves. I actually spent a good part of yesterday afternoon looking at sustained pure bass notes just to double-check my previous findings. What I have found over the decades is though bass notes dominate the recording, considerable energy is in other notes that are played at the same time. In addition, the bass notes themselves aren't pure tones at the note's fundamental frequency. It is not unusual for the majority of the energy in the bass note to be in the harmonics of the fundamental. All of these influences work together to considerably increase the crest factor of the music from the 3 dB we see for pure since waves to 10 dB or more. This vastly reduces the average drain on power supplies and the heating if the power transistors and heat sinks serving the output stages. It is also why people can plug power amps capable of putting out thousands of watts of sine waves into a single ordinary house outlet and never blow the circuit breaker as long as they play recordings of regular music or drama.


So you are saying that a high crest factor allows the power supplies to cool down so they can handle higher instantaneous peaks?

The short answer is the real world is better than that - the power supplies and amp heat sinks never heat up because the peaks are so short. The peaks in music often range between 2 thousandths of a second and 20 thousandths of a second. I've verified this by inspecting hundreds of recordings, some I made myself to ensure that there was no peak limiting or compression in the recording process.

However, heat isn't the issue - amplifier power is largely dictated by power supply voltage and that is the issue. Typically, amp power supplies have large electrolytic capacitors that hold enough charge to significantly flywheel the amp through the short peaks described above.

Almost nobody tests amps with speaker loads because it is risky and and awkward process. Only a few people (including yours truly) have tested power amps for power and distortion with music and a loudspeaker load. When one switches from sine waves and resistor loads to music and a loudspeaker like load the amp works like it is basically is on break. Its working, but it doesn't look that way! ;-)
arnyk is offline  
post #592 of 637 Old 03-23-2014, 09:33 AM
 
Jonathan Parker's Avatar
 
Join Date: Mar 2014
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by arnyk View Post


The short answer is the real world is better than that - the power supplies and amp heat sinks never heat up because the peaks are so short. The peaks in music often range between 2 thousandths of a second and 20 thousandths of a second. I've verified this by inspecting hundreds of recordings, some I made myself to ensure that there was no peak limiting or compression in the recording process.

However, heat isn't the issue - amplifier power is largely dictated by power supply voltage and that is the issue. Typically, amp power supplies have large electrolytic capacitors that hold enough charge to significantly flywheel the amp through the short peaks described above.

Almost nobody tests amps with speaker loads because it is risky and and awkward process. Only a few people (including yours truly) have tested power amps for power and distortion with music and a loudspeaker load. When one switches from sine waves and resistor loads to music and a loudspeaker like load the amp works like it is basically is on break. Its working, but it doesn't look that way! ;-)



I assumed heat was a contributing factor that prevented the amp power supplies from delivering on short term peaks. A high crest factor would allow the components to cool down in less dynamic sequences, allowing the capacitors to hold more charge more consistently. I don't know, perhaps I'm way off base here.

Jonathan Parker is offline  
post #593 of 637 Old 03-23-2014, 09:52 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Jonathan Parker View Post

Quote:
Originally Posted by arnyk View Post

The short answer is the real world is better than that - the power supplies and amp heat sinks never heat up because the peaks are so short. The peaks in music often range between 2 thousandths of a second and 20 thousandths of a second. I've verified this by inspecting hundreds of recordings, some I made myself to ensure that there was no peak limiting or compression in the recording process.


However, heat isn't the issue - amplifier power is largely dictated by power supply voltage and that is the issue. Typically, amp power supplies have large electrolytic capacitors that hold enough charge to significantly flywheel the amp through the short peaks described above.


Almost nobody tests amps with speaker loads because it is risky and and awkward process. Only a few people (including yours truly) have tested power amps for power and distortion with music and a loudspeaker load. When one switches from sine waves and resistor loads to music and a loudspeaker like load the amp works like it is basically is on break. Its working, but it doesn't look that way! ;-)



I assumed heat was a contributing factor that prevented the amp power supplies from delivering on short term peaks.

Simply not so. Let me emphasize that the whole amp stays relatively cool when amplifying music into a speaker like load as opposed to sine waves into resistive loads. The muscle is for the test bench, not the music.
Quote:
A high crest factor would allow the components to cool down in less dynamic sequences, allowing the capacitors to hold more charge more consistently.

That would be correct if the components got very hot in the first place. The typical crest factor of music is about 10 dB or more, and that means that the amp dissipates many times less heat then it does in tests used to establish power ratings.

I've been bench testing power amps since 1959. My first bench test was done in the HS physics lab using the school's 'scope. In that test I found that the output transformer I was using which came from a car radio was generating high distortion in a narrow band around 7 KHz. I upgraded it to a purpose built part. A few years later I found a similar transformer being used in a stereo receiver that Kenwood built under the Lafayette name. A few months later one of the big audio magazine's lab tests confirmed that finding.

The first amps I tested got pretty hot regardless, because they had tubes!

I have two Audio Precision test sets on hand and use them occasionally, but I prefer using high quality sound cards as monitoring/analysis devices because I can't afford the newer AP equipment that can keep up with them.
arnyk is offline  
post #594 of 637 Old 03-23-2014, 10:21 AM
 
Jonathan Parker's Avatar
 
Join Date: Mar 2014
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by arnyk View Post


Simply not so. Let me emphasize that the whole amp stays relatively cool when amplifying music into a speaker like load as opposed to sine waves into resistive loads. The muscle is for the test bench, not the music.
That would be correct if the components got very hot in the first place. The typical crest factor of music is about 10 dB or more, and that means that the amp dissipates many times less heat then it does in tests used to establish power ratings.

I've been bench testing power amps since 1959. My first bench test was done in the HS physics lab using the school's 'scope. In that test I found that the output transformer I was using which came from a car radio was generating high distortion in a narrow band around 7 KHz. I upgraded it to a purpose built part. A few years later I found a similar transformer being used in a stereo receiver that Kenwood built under the Lafayette name. A few months later one of the big audio magazine's lab tests confirmed that finding.

The first amps I tested got pretty hot regardless, because they had tubes!

I have two Audio Precision test sets on hand and use them occasionally, but I prefer using high quality sound cards as monitoring/analysis devices because I can't afford the newer AP equipment that can keep up with them.


Okay, so you are saying that amps just don't get all that hot in actual usage using music and drama? And hence have more headroom available as opposed to sine wave testing? Sorry, sometimes your posts are hard to follow.

Jonathan Parker is offline  
post #595 of 637 Old 03-23-2014, 12:34 PM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 402 Post(s)
Liked: 1046
Quote:
Originally Posted by Jonathan Parker View Post

Quote:
Originally Posted by arnyk View Post

Simply not so. Let me emphasize that the whole amp stays relatively cool when amplifying music into a speaker like load as opposed to sine waves into resistive loads. The muscle is for the test bench, not the music.

That would be correct if the components got very hot in the first place. The typical crest factor of music is about 10 dB or more, and that means that the amp dissipates many times less heat then it does in tests used to establish power ratings.


I've been bench testing power amps since 1959. My first bench test was done in the HS physics lab using the school's 'scope. In that test I found that the output transformer I was using which came from a car radio was generating high distortion in a narrow band around 7 KHz. I upgraded it to a purpose built part. A few years later I found a similar transformer being used in a stereo receiver that Kenwood built under the Lafayette name. A few months later one of the big audio magazine's lab tests confirmed that finding.


The first amps I tested got pretty hot regardless, because they had tubes!


I have two Audio Precision test sets on hand and use them occasionally, but I prefer using high quality sound cards as monitoring/analysis devices because I can't afford the newer AP equipment that can keep up with them.


Okay, so you are saying that amps just don't get all that hot in actual usage using music and drama?

Kinda. Its all relative but what is sure is that they don't get as hot as they would while being tested on the bench with pure sine waves being delivered to load resistors.

I've seen AVRs get too hot while playing music but the ventilation was bad.
Quote:
And hence have more headroom available as opposed to sine wave testing?

That too. I've never seen an amp whose peak output playing signals that varied wasn't more than what it was with steady sine waves.
arnyk is offline  
post #596 of 637 Old 03-23-2014, 06:25 PM
AVS Special Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 5,912
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 230
Many amplifiers are limited by the power supply, which includes filter/decoupling capacitors. Voltage sag causes power loss faster and more often than thermal (heat) issues. Thermal issues tend to be long-term rather than limiting short=term peaks.

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is offline  
post #597 of 637 Old 03-23-2014, 06:58 PM
AVS Special Member
 
JHAz's Avatar
 
Join Date: Mar 2009
Posts: 3,848
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 32 Post(s)
Liked: 146
Quote:
Originally Posted by Jonathan Parker View Post


Okay, so you are saying that amps just don't get all that hot in actual usage using music and drama? And hence have more headroom available as opposed to sine wave testing? Sorry, sometimes your posts are hard to follow.
I think that is correct generally speaking. A sine wave is the same cycle after cycle. But look at an uncompressed sound from say an electric bass and the first half cycle is ten to twelve dB loaders thane the beginning of the sustain, and the sustain loses way more than three dB per second. A piano will be roughly the same. Bowed instruments could if played by robots with no idea how to be musical be far more consistent but in the hands of humans attempting to be expressive will have lots more dynamic range than a sine wave. Not to mention the effects of the harmonics on the power envelope.
JHAz is offline  
post #598 of 637 Old 03-25-2014, 12:46 PM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,768
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 493 Post(s)
Liked: 323
Quote:
Originally Posted by arnyk View Post

Sure:




However the above in my opinion represents a fool's mission because it can't be heard and its spectral content was not asked for.

In this case crest factor was calculated from Average RMS power and Peak Amplitude. The display and analysis tool was Adobe Audition. Note that the actual crest factor is more than 10 dB, which is what I meant by "10 dB or more".

I expect major hypercritical nit-picking to follow! ;-)
I apologize for causing you angst Arny. I was genuine in my question. I have seen the argument you are making that music has a lower crest factor (difference between peak and average) than sine wave used for amplifier power testing but I have yet to find one that says how they computed that "10 db" number. A few say that they used the DR Meter which is inappropriate for this type of analysis.

Anyway, looking at your sample data, I see that the background is white. That tells me that you have selected the entire song. I thought the discussion was focused around short-term peak activity and you said something about analyzing individual notes, not the whole song. The amplifier only gets instantaneous levels to play. It is not able to analyze a whole song like Adobe Audition does for you above. As such, the amplifier cannot meter and optimize its power reserves across the whole song. Such analysis then is inappropriate with regards to this topic. You need to do what was mentioned which is analyze short segments.

To test your hypothesis and not look like I am cherry picking, I selected the French Horn track from suite of MPEG standardized audio test clips. I randomly picked a dense part of the waveform and analyzed it as you did except that mine only goes after the marked section (my copy of Audition is much newer than yours but has the same field):

i-Pwxnd5V-XL.png

Focusing on the right channel, we have an RMS (average) value of -11.37 db and peak of -8.46 db (I have highlighted them in red). The difference between them is just 2.91 db. As you know, sine wave has a crest factor of 3 db so this clip actually stresses the amp more than sine wave, not less. And not remotely close to the 10 db value you mention.

No doubt different clips and tracks will result in different ratios. And therein lies the problem with this argument. You can't know what is "music." Music is whatever it wants to be smile.gif. The famous demo track of No Sanctuary from Chris Jones starts with what sounds like a pure tone. Unusual but there it is. We don't get to rule that out. There is no test when music is produced to say, "if crest factor is less than 3, redo the mix." The talent produces the music he/she wants and if it happens to have a pure french horn in it like the above sample I analyzed, no one is going to object. If the amp distorts while reproducing this clip, then we have too little power. It matters not if another or hundred other tracks don't distort.

I hope this didn't come across as "major hypercritical nit-picking." smile.gif The analysis behind this argument is key to its validity.

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is online now  
post #599 of 637 Old 03-25-2014, 01:55 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,885
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 89 Post(s)
Liked: 86
Sinewave testing is kind of a worst case scenario. It can be done with short burst too if cooling is not engineered for continuous sine wave loads.
For eaxemples 7 amps sharing a PSY and 'intentional' limited heatsink capability in av amps is designed with a reasonable crest faktor chosen by the designer.

Point is that with a sinewave test and a 8Ohm load were dealing with two knowns as opposed to dragging an undefined speaker load and crest factor into the mix.
From this simple test we get a distortion level and the actual amount of power at a given distortion level.

Simple and repeatable.
Frank Derks is offline  
post #600 of 637 Old 03-25-2014, 03:35 PM
AVS Special Member
 
mcnarus's Avatar
 
Join Date: Nov 2007
Posts: 6,008
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 262
Quote:
Point is that with a sinewave test and a 8Ohm load were dealing with two knowns as opposed to dragging an undefined speaker load and crest factor into the mix.
From this simple test we get a distortion level and the actual amount of power at a given distortion level.

Simple and repeatable.
Exactly. A spec is just a spec, and a measurement is just a measurement. Neither is a universal truth. They're just indications of an amp's capabilities. And as an indicator, they work well enough.

If you can't explain how it works, you can't say it doesn't.—The High-End Creed

mcnarus is online now  
Reply Audio theory, Setup and Chat

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off