Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the impedance to be any good?

It has been my contention that power amps that double their power as
impedance halves are the product of manipulation of specifications for
marketing, not how real-world audio technology actually works. Here's a
real world data point for the controversy about power amps needing to be
able to put out twice the power for half the impedance:

I ran some bench tests on what may be one of the most overbuilt power amps
in the history of audio - the Threshold SA/4e from the early 1990s.

No visible effort was spared to build a robust 100 wpc power amp, as this
article shows:

http://www.thresholdlovers.com/artic...?lng=en&pg=394

The Threshold SA/4e weighs about 110 pounds, according to my calibrated
biceps. You really want two people to tote it around, and it facilitates
this by having two comfortable handles on each end.

In contrast, in similar tests a Behringer A500 which weighs about 15 pounds
puts out about the same amount of power at clipping at 8 ohms, but puts out
about 144 watts at both 4 and 2 ohms. This is what I would expect from a
typical contemporary mainstream amplifier.

Among other features, the Threshold SA/4e amplifier has extremely soft
clipping. While most power amps, even tubed power amps have a fairly
well-defined clipping point on the bench, not so with the SA/4e. In the end
I decided to measure its output power at about 0.1% THD @ 1KHz, as observed
with a RTA.

The actual measured power is as listed he

http://cgi.audiogon.com/cgi-bin/auc....82027&auc&3&4&

Namely 112 Watts per channel @ 8 ohms with both channels driven @ 1000 Hz,
196 Watts @ 4 ohms, 264 Watts @ 2 ohms.

This is still very far from doubling power as load impedance halves,
particularly at 2 ohms.


  #2   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the

"ScottW2" wrote in message

On Aug 18, 5:51 am, "Arny Krueger"
wrote:
It has been my contention that power amps that double
their power as impedance halves are the product of
manipulation of specifications for marketing, not how
real-world audio technology actually works. Here's a
real world data point for the controversy about power
amps needing to be able to put out twice the power for
half the impedance:

I ran some bench tests on what may be one of the most
overbuilt power amps in the history of audio - the
Threshold SA/4e from the early 1990s.

No visible effort was spared to build a robust 100 wpc
power amp, as this article shows:

http://www.thresholdlovers.com/artic...?lng=en&pg=394

The Threshold SA/4e weighs about 110 pounds, according
to my calibrated biceps. You really want two people to
tote it around, and it facilitates this by having two
comfortable handles on each end.

In contrast, in similar tests a Behringer A500 which
weighs about 15 pounds puts out about the same amount of
power at clipping at 8 ohms, but puts out about 144
watts at both 4 and 2 ohms. This is what I would expect
from a typical contemporary mainstream amplifier.

Among other features, the Threshold SA/4e amplifier has
extremely soft clipping. While most power amps, even
tubed power amps have a fairly well-defined clipping
point on the bench, not so with the SA/4e. In the end I
decided to measure its output power at about 0.1% THD @
1KHz, as observed with a RTA.

The actual measured power is as listed he

http://cgi.audiogon.com/cgi-bin/auc....82027&auc&3&4&

Namely 112 Watts per channel @ 8 ohms with both channels
driven @ 1000 Hz, 196 Watts @ 4 ohms, 264 Watts @ 2 ohms.

This is still very far from doubling power as load
impedance halves, particularly at 2 ohms.


This is going to largely depend upon the speakers the amp
is matched with and if the amp is driven near clipping
into it's load at whatever max spl the listener wants
from their system.


I think that the one of the major goals of the SA/4e was to produce an amp
that was as independent as possible of the characteristics of its speaker
load.

Ideally, if you want an amp that can
drive any speaker load you can come up with, an amp that
doubles power from 8 to 4 and doubles again to 2 will
probably be more versatile than one that can't.


There is no such thing as a simple amplfiier that doubles power when you
halve impedance. Vastly overbuilt whales like the SA/4e prove this out.

I'd suggest most folks should select their speakers and then
decide what amp requirements are appropriate.


They should avoid speakers that put extreme demands on amplifiers, as bad
design in the area of impedance curve suggests bad design elsewhere.

  #3   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"Arny Krueger" wrote:

"ScottW2" wrote in message

On Aug 18, 5:51 am, "Arny Krueger"
wrote:
It has been my contention that power amps that double
their power as impedance halves are the product of
manipulation of specifications for marketing, not how
real-world audio technology actually works. Here's a
real world data point for the controversy about power
amps needing to be able to put out twice the power for
half the impedance:

I ran some bench tests on what may be one of the most
overbuilt power amps in the history of audio - the
Threshold SA/4e from the early 1990s.

No visible effort was spared to build a robust 100 wpc
power amp, as this article shows:

http://www.thresholdlovers.com/artic...?lng=en&pg=394

The Threshold SA/4e weighs about 110 pounds, according
to my calibrated biceps. You really want two people to
tote it around, and it facilitates this by having two
comfortable handles on each end.

In contrast, in similar tests a Behringer A500 which
weighs about 15 pounds puts out about the same amount of
power at clipping at 8 ohms, but puts out about 144
watts at both 4 and 2 ohms. This is what I would expect
from a typical contemporary mainstream amplifier.

Among other features, the Threshold SA/4e amplifier has
extremely soft clipping. While most power amps, even
tubed power amps have a fairly well-defined clipping
point on the bench, not so with the SA/4e. In the end I
decided to measure its output power at about 0.1% THD @
1KHz, as observed with a RTA.

The actual measured power is as listed he

http://cgi.audiogon.com/cgi-bin/auc....82027&auc&3&4&

Namely 112 Watts per channel @ 8 ohms with both channels
driven @ 1000 Hz, 196 Watts @ 4 ohms, 264 Watts @ 2 ohms.

This is still very far from doubling power as load
impedance halves, particularly at 2 ohms.


This is going to largely depend upon the speakers the amp
is matched with and if the amp is driven near clipping
into it's load at whatever max spl the listener wants
from their system.


I think that the one of the major goals of the SA/4e was to produce an amp
that was as independent as possible of the characteristics of its speaker
load.

Ideally, if you want an amp that can
drive any speaker load you can come up with, an amp that
doubles power from 8 to 4 and doubles again to 2 will
probably be more versatile than one that can't.


There is no such thing as a simple amplfiier that doubles power when you
halve impedance. Vastly overbuilt whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do precisely that
(and being nearly a perfect voltage source is pretty automatic for a
decent solid-state amp). It is important to understand, though, that you
*must* limit the lower limit of impedance for which the amp is allowed
to be a voltage source, or things will melt...

Isaac

  #4   Report Post  
Posted to rec.audio.high-end
mcdonaldREMOVE TO ACTUALLY REACH [email protected] mcdonaldREMOVE TO ACTUALLY REACH ME@scs.uiuc.edu is offline
external usenet poster
 
Posts: 42
Default Do power amps need to deliver twice the power into half the

isw wrote:

Ideally, if you want an amp that can
drive any speaker load you can come up with, an amp that
doubles power from 8 to 4 and doubles again to 2 will
probably be more versatile than one that can't.

There is no such thing as a simple amplfiier that doubles power when you
halve impedance. Vastly overbuilt whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do precisely that
(and being nearly a perfect voltage source is pretty automatic for a
decent solid-state amp).


Actually thats not true. Solid state amps are indeed near perfect
voltage sources until the voltage drop across a near saturated
bipolar transistor becomes non-negligible, or a FET nears its low
limit of resistance. Then the effective series resistance will
cut power at high current (low load impedance).

Now of course this can be fix by installing a high quality
fixed voltage limit of the output sage. In other words, very
hard fix-voltage clipping. Then down to some low load
resistance value, the power output really will be inversely proportional
to load resistance. This is the opposite of "soft clipping"
and is not exactly in fact to be desired.

Doug McDonald

  #5   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the impedance to be any good?

In article ,
"Arny Krueger" wrote:

It has been my contention that power amps that double their power as
impedance halves are the product of manipulation of specifications for
marketing, not how real-world audio technology actually works.


A pure voltage source (not a bad thing for an audio amp to be -- up to a
point) will indeed double its power output every time the load impedance
is halved. This is OK for 16, 8, 4, maybe 2, and possibly 1 ohm loads.
But every time the load halves, the current doubles (because the voltage
is constant), so it's probably a good idea to have some mechanism (other
than the flying leads inside the output devices) that prevents the thing
from trying to deliver a megamp or two into a micro-ohm load.

Isaac


  #6   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the

"isw" wrote in message


There is no such thing as a simple amplfiier that
doubles power when you halve impedance. Vastly overbuilt
whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do
precisely that (and being nearly a perfect voltage source
is pretty automatic for a decent solid-state amp).


Simply not so. While modern power amps behave like constant-voltage sources,
they only do that within bounds.

The bounds are set by gross losses in the output stage and power supply that
won't go away and can't be corrected by negative feedback.

It is
important to understand, though, that you *must* limit
the lower limit of impedance for which the amp is allowed
to be a voltage source, or things will melt...


Every power amp has limited output power, even at 8 ohms.

  #7   Report Post  
Posted to rec.audio.high-end
Sonnova Sonnova is offline
external usenet poster
 
Posts: 1,337
Default Do power amps need to deliver twice the power into half the

On Thu, 20 Aug 2009 06:39:20 -0700, isw wrote
(in article ):

In article ,
"Arny Krueger" wrote:


[quoted text deleted -- deb]

There is no such thing as a simple amplfiier that doubles power when you
halve impedance. Vastly overbuilt whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do precisely that
(and being nearly a perfect voltage source is pretty automatic for a
decent solid-state amp). It is important to understand, though, that you
*must* limit the lower limit of impedance for which the amp is allowed
to be a voltage source, or things will melt...

Isaac


Since Power = V squared / R, I don't see how any power amp can be "divorced"
from it's load. Anyone can see that power increases as R gets smaller and
that a dead short (R=1 Ohm. It won't be zero because no solid-state output
device has zero "on" resistance) results in power being, essentially, the
same as V squared. Since it's doubtful that any amp can supply that much
voltage, the power supply becomes the limiting factor. For instance:

assuming a V swing of 100 Volts * 100 volts squared is 10,000/8 = 1250 watts
and 10,000/4 = 2500 Watts and 10,000/2 = 5,000 watts etc.

Since power is also a product of current (voltage, current and resistance
are, as we know, inexorably linked together) as in P = VI and I squared times
R. The power supply has to be able to deliver enough current across the load
to develop sufficient voltage to produce the power associated with the
example given. I.E. in real applications, the current limits of the amp's
power supply will insure that less and less voltage is developed across the
speaker as the load impedance drops, but at some point, usually before the
power supply taps out, the collector current In a bi-polar output device
(drain current in a FET) is going to exceed the device's rating (I = E/R)
and the transistor is going to fail.

So the truth is that in order for any amplifier to keep doubling it's power
as the load impedance drops, it would have to have a power supply capable of
maintaining the amp's rail voltages regardless of load by supplying unlimited
current (and only Morbius' ancient Krell in "Forbidden Planet" could manage
that) and it would require unlimited I sub c on the output transistors. Even
wildly over-built amplifiers like the previous poster's SA/4e cannot meet
those requirements. Of course you can parallel output devices to increase
current handling, but there's a practical limit to that as well.

  #8   Report Post  
Posted to rec.audio.high-end
Sonnova Sonnova is offline
external usenet poster
 
Posts: 1,337
Default Do power amps need to deliver twice the power into half the

On Thu, 20 Aug 2009 16:55:38 -0700, Arny Krueger wrote
(in article ):

"isw" wrote in message


There is no such thing as a simple amplfiier that
doubles power when you halve impedance. Vastly overbuilt
whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do
precisely that (and being nearly a perfect voltage source
is pretty automatic for a decent solid-state amp).


Simply not so. While modern power amps behave like constant-voltage sources,
they only do that within bounds.

The bounds are set by gross losses in the output stage and power supply that
won't go away and can't be corrected by negative feedback.

It is
important to understand, though, that you *must* limit
the lower limit of impedance for which the amp is allowed
to be a voltage source, or things will melt...


Every power amp has limited output power, even at 8 ohms.


Because P =VI and where V (rail Voltage) is a constant, I, current, must
double every time R (load impedance) is halved (P= I squared times R) to
maintain V. All amps have a current limit, but up to that point, power does,
indeed, double every time the load impedance is halved.
  #9   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"Arny Krueger" wrote:

"isw" wrote in message


There is no such thing as a simple amplfiier that
doubles power when you halve impedance. Vastly overbuilt
whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do
precisely that (and being nearly a perfect voltage source
is pretty automatic for a decent solid-state amp).


Simply not so. While modern power amps behave like constant-voltage sources,
they only do that within bounds.


Well, that's what I *thought* I was saying...

Every power amp has limited output power, even at 8 ohms.


Of course; limited by the available supply voltage, if nothing else.

Let's try this:

Acting like a voltage source is a very good thing for an audio amplifier
to do. IF there are circumstances where the amp fails to act as a
voltage source, then it's probably not the right amp for the job.

Put another way: If your amplifier fails to act as a pure voltage source
IN ANY WAY while you're trying to play music through it, you need a
different amplifier. Period. An amp failing to act as a voltage source
is guaranteed to cause distortion. (This applies to domestic
reproduction; not professional sound reinforcement situations).

Isaac

  #10   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the

"isw" wrote in message


Let's try this:

Acting like a voltage source is a very good thing for an
audio amplifier to do.


True within limits. Those limits are often known as "Power ratings". They
are generally strongly dependent on load impedance.

IF there are circumstances where
the amp fails to act as a voltage source, then it's
probably not the right amp for the job.


Every real world amp has circumstances where it fails to act like a voltage
source. Run it outside its ratings. Run it into too-low of an impedance.
Don't give it enough power line power and/or voltage.

Put another way: If your amplifier fails to act as a pure
voltage source IN ANY WAY while you're trying to play
music through it, you need a different amplifier.


You've got at least one other option which is to obtain more effcient
speakers.

Period.


An unqualified statement, and therefore likely to be false on the face of
it. In fact, its false.

An amp failing to act as a voltage source is guaranteed
to cause distortion. (This applies to domestic
reproduction; not professional sound reinforcement
situations).


Professional sound reinforcment systems follow the same rules of electrical
engieneering basic as home hi fi.

I can state unequivocally that An amp failing to act as a voltage source is
guaranteed to cause distortion, whether its home audio or pro audio.




  #11   Report Post  
Posted to rec.audio.high-end
mcdonaldREMOVE TO ACTUALLY REACH [email protected] mcdonaldREMOVE TO ACTUALLY REACH ME@scs.uiuc.edu is offline
external usenet poster
 
Posts: 42
Default Do power amps need to deliver twice the power into half the

Arny Krueger wrote:


I can state unequivocally that An amp failing to act as a voltage source is
guaranteed to cause distortion, whether its home audio or pro audio.



Well, yes, if you include frequency response changes as distortion.

But consider a perfect amp with a one ohm resistor hidden inside it
between the feedback loop and the output terminal.

Fed into a pure resistance it will have no distortion whether you measure
voltage at output terminal or current.

Fed into a reactive load it will have frequency and phase
response distortion. Fed into a nonlinear response load
(e.g. a diode) it will generate harmonic and intermodulation distortion,
even measured as voltage.

Doug
  #12   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the

"mcdonaldREMOVE TO ACTUALLY REACH wrote
in message
Arny Krueger wrote:


I can state unequivocally that An amp failing to act as
a voltage source is guaranteed to cause distortion,
whether its home audio or pro audio.


Well, yes, if you include frequency response changes as
distortion.


Calling frequency response changes "Linear Distortion" is accepted
practice.

But consider a perfect amp with a one ohm resistor hidden
inside it between the feedback loop and the output
terminal.


Fed into a pure resistance it will have no distortion
whether you measure voltage at output terminal or current.


Actually, there will be a broadband frequency response error - a level
shift.

Fed into a reactive load it will have frequency and phase
response distortion. Fed into a nonlinear response load
(e.g. a diode) it will generate harmonic and
intermodulation distortion, even measured as voltage.


True.



  #13   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"Arny Krueger" wrote:

"isw" wrote in message


Let's try this:

Acting like a voltage source is a very good thing for an
audio amplifier to do.


True within limits. Those limits are often known as "Power ratings". They
are generally strongly dependent on load impedance.

IF there are circumstances where
the amp fails to act as a voltage source, then it's
probably not the right amp for the job.


Every real world amp has circumstances where it fails to act like a voltage
source. Run it outside its ratings. Run it into too-low of an impedance.
Don't give it enough power line power and/or voltage.

Put another way: If your amplifier fails to act as a pure
voltage source IN ANY WAY while you're trying to play
music through it, you need a different amplifier.


You've got at least one other option which is to obtain more effcient
speakers.


Which, if I'm not mistaken would (should) let the amplifier work without
overloading (i.e. act like a voltage source again).


Period.


An unqualified statement, and therefore likely to be false on the face of
it. In fact, its false.


OK. Please explain how an amplifier can behave NOT as a voltage source
and not generate distortion in the signal.


An amp failing to act as a voltage source is guaranteed
to cause distortion. (This applies to domestic
reproduction; not professional sound reinforcement
situations).


Professional sound reinforcment systems follow the same rules of electrical
engieneering basic as home hi fi.

I can state unequivocally that An amp failing to act as a voltage source is
guaranteed to cause distortion, whether its home audio or pro audio.


I stand corrected. I tried to make that comment too succinct and failed.
What I was intending to say is that while there are possibly good
reasons to allow amplifying gear to operate NOT like a voltage source in
professional sound reinforcement applications, there's really no good
reason to allow non-linear operation when you're trying to do a good job
of reproducing music in a home.

Isaac
  #14   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"mcdonaldREMOVE TO ACTUALLY REACH wrote:

Arny Krueger wrote:


I can state unequivocally that An amp failing to act as a voltage source
is
guaranteed to cause distortion, whether its home audio or pro audio.



Well, yes, if you include frequency response changes as distortion.


But of course they are; they represent a failure of the amp to be a
voltage source, no?

But consider a perfect amp with a one ohm resistor hidden inside it
between the feedback loop and the output terminal.


Well, that makes it a voltage source with a one ohm series resistance in
the output. Not too different from a real-world situation where speaker
leads, crossover inductors, and voice coils have quite measurable
resistances.

Fed into a pure resistance it will have no distortion whether you measure
voltage at output terminal or current.


Well, if you vary your pure resistance, you'll see that (because of that
one-ohm resistor) the gain of the amplifier appears to be a function of
load; the lower the load resistance, the lower the output voltage. Would
you define that as a sort of distortion? I'm not sure, but am inclined
to; certainly it violates the "straight wire with gain" notion.

Fed into a reactive load it will have frequency and phase
response distortion.


True in theory, but we're talking about *audio* amplifiers here. Do you
know of any actual loudspeakers that have enough reactance to cause a
significant problem (let's ignore electrostats here, because they
usually require special treatment in other ways)?

Fed into a nonlinear response load
(e.g. a diode) it will generate harmonic and intermodulation distortion,
even measured as voltage.


Again true in theory, but how many loudspeakers have diodes inside
(excepting back-to-back zeners used for protection; they should never
conduct except under pathological conditions)?

Isaac
  #15   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"mcdonaldREMOVE TO ACTUALLY REACH wrote:

isw wrote:

Ideally, if you want an amp that can
drive any speaker load you can come up with, an amp that
doubles power from 8 to 4 and doubles again to 2 will
probably be more versatile than one that can't.
There is no such thing as a simple amplfiier that doubles power when you
halve impedance. Vastly overbuilt whales like the SA/4e prove this out.


That's not so; any amp that is a voltage source will do precisely that
(and being nearly a perfect voltage source is pretty automatic for a
decent solid-state amp).


Actually thats not true. Solid state amps are indeed near perfect
voltage sources until the voltage drop across a near saturated
bipolar transistor becomes non-negligible, or a FET nears its low
limit of resistance. Then the effective series resistance will
cut power at high current (low load impedance).

Now of course this can be fix by installing a high quality
fixed voltage limit of the output sage. In other words, very
hard fix-voltage clipping. Then down to some low load
resistance value, the power output really will be inversely proportional
to load resistance. This is the opposite of "soft clipping"
and is not exactly in fact to be desired.


IMO if your amplifier clips (i.e. if it *ever* fails to be a voltage
source while you're using it to listen to music), you need a bigger one
(or, as Arnie said, more efficient speakers).

Isaac



  #16   Report Post  
Posted to rec.audio.high-end
mcdonaldREMOVE TO ACTUALLY REACH [email protected] mcdonaldREMOVE TO ACTUALLY REACH ME@scs.uiuc.edu is offline
external usenet poster
 
Posts: 42
Default Do power amps need to deliver twice the power into half the

isw wrote:
In article ,
"mcdonaldREMOVE TO ACTUALLY REACH wrote:


But consider a perfect amp with a one ohm resistor hidden inside it
between the feedback loop and the output terminal.


Fed into a reactive load it will have frequency and phase
response distortion.


True in theory, but we're talking about *audio* amplifiers here. Do you
know of any actual loudspeakers that have enough reactance to cause a
significant problem


Oh yes ... absolutely. Loudspeakers are highly, exceedingly reactive. Normal
ones expect to be driven from a voltage source. I should add that speakers
are reactive even if you consider a single driver with no external
electronics. The mechanical part of the electro-mechanical system
has resonances etc. and these back up into the electrical system
and look like reactance.


Doug

  #17   Report Post  
Posted to rec.audio.high-end
isw isw is offline
external usenet poster
 
Posts: 182
Default Do power amps need to deliver twice the power into half the

In article ,
"mcdonaldREMOVE TO ACTUALLY REACH wrote:

isw wrote:
In article ,
"mcdonaldREMOVE TO ACTUALLY REACH wrote:


But consider a perfect amp with a one ohm resistor hidden inside it
between the feedback loop and the output terminal.


Fed into a reactive load it will have frequency and phase
response distortion.


True in theory, but we're talking about *audio* amplifiers here. Do you
know of any actual loudspeakers that have enough reactance to cause a
significant problem


Oh yes ... absolutely. Loudspeakers are highly, exceedingly reactive. Normal
ones expect to be driven from a voltage source. I should add that speakers
are reactive even if you consider a single driver with no external
electronics. The mechanical part of the electro-mechanical system
has resonances etc. and these back up into the electrical system
and look like reactance.


I don't doubt that there are reactances associated with speakers --
reactances are *everywhere*.

But, do you know of any speakers with sufficient reactance *at audio
frequencies* to cause problems such as frequency response errors or
phase errors? (Sure, I know there might be crossovers inside;
presumably they are doing the "right thing" -- whatever the designer
intends).

Isaac
  #18   Report Post  
Posted to rec.audio.high-end
Dick Pierce[_2_] Dick Pierce[_2_] is offline
external usenet poster
 
Posts: 151
Default Do power amps need to deliver twice the power into half the

"mcdonaldREMOVE TO ACTUALLY REACH wrote:
True in theory, but we're talking about *audio* amplifiers here. Do you
know of any actual loudspeakers that have enough reactance to cause a
significant problem


Oh yes ... absolutely. Loudspeakers are highly,
exceedingly reactive.


No, they are not "exceedingly" reactive, not by any
reasonably objective definition of the term. If, for
example, be "exceedingly reactive," you mean that
over their specified bandwidth (or over the full
audio bandwidth of 20 Hz to 20 kHz), the magnitude of
the reactive portion of the impedance exceeds the
resistive portion the majority of the time, then no
loudspeaker system meets that definition of "exceedingly
reactive."

Even if you use a definition that is far more generous,
like the reactive portion of the impedance dominates over
at least, oh, 20% of the speaker's bandwidth, then maybe
1 in 100 models and probably 1 in 1,000 speakers sold may
barely meet that qualification.

For the VAST majority of speakers hooked to the VAST
majority of amplifiers, the impedance almost never is
dominantly reactive. The phase angle of the impedance
almost alwyas approaches, but exceedingly ever wanders
outside the +-45 degree limits that signal the objective
border between "domanantly resistive" and "dominantly
reactive."

Normal ones expect to be driven from a voltage source.


This is true.

I should add that speakers are reactive even if you
consider a single driver with no external electronics.


As is this.

The mechanical part of the electro-mechanical system
has resonances etc. and these back up into the
electrical system and look like reactance.


Generally true. One thing many commentators and, regrettably,
self-appointed luminaries and designers in the audio industry
fail utterly to comprehend is that, to the amplifier, there is
no difference between a speaker with its mechanical components
forming a collection of resonant systems and an equivalent
collection of resistors, capacitors and inductors. Magical,
special and myetrious attributes are assigned to masses attached
to voices coils moving in magnetic fields and "back EMF" that
are, pgysically, NO different electrically than a parallel
resonant tank circuit consisting of an inductor, capacitor
and resistor. Electrically, they behave EXACTLY the same way,
and elctrically is the ONLY way the amplifier knows about the
speaker.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+
  #19   Report Post  
Posted to rec.audio.high-end
Arny Krueger Arny Krueger is offline
external usenet poster
 
Posts: 17,262
Default Do power amps need to deliver twice the power into half the

"mcdonaldREMOVE TO ACTUALLY REACH wrote
in message
isw wrote:
In article ,
"mcdonaldREMOVE TO ACTUALLY REACH
wrote:


But consider a perfect amp with a one ohm resistor
hidden inside it between the feedback loop and the
output terminal.


Fed into a reactive load it will have frequency and
phase
response distortion.


True in theory, but we're talking about *audio*
amplifiers here. Do you know of any actual loudspeakers
that have enough reactance to cause a significant problem


Oh yes ... absolutely. Loudspeakers are highly,
exceedingly reactive.


In the cosmic scheme of things, not exceedingly. For example, consider a
standard Fluorescent balast.

Normal ones expect to be driven
from a voltage source.


Agreed.

I should add that speakers are
reactive even if you consider a single driver with no
external electronics. The mechanical part of the
electro-mechanical system has resonances etc. and these back up into the
electrical
system and look like reactance.


Right.


  #20   Report Post  
Posted to rec.audio.high-end
Dick Pierce[_2_] Dick Pierce[_2_] is offline
external usenet poster
 
Posts: 151
Default Do power amps need to deliver twice the power into half the

Dick Pierce wrote:
For the VAST majority of speakers hooked to the VAST
majority of amplifiers, the impedance almost never is
dominantly reactive. The phase angle of the impedance
almost alwyas approaches, but exceedingly ever wanders
outside the +-45 degree limits


Sorry, I meant to write "only exceedingly rarely" instead
of the gobbldeygook I did.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+



  #21   Report Post  
Posted to rec.audio.high-end
Robert Peirce[_2_] Robert Peirce[_2_] is offline
external usenet poster
 
Posts: 15
Default Do power amps need to deliver twice the power into half the

In article ,
Dick Pierce wrote:

Dick Pierce wrote:
For the VAST majority of speakers hooked to the VAST
majority of amplifiers, the impedance almost never is
dominantly reactive. The phase angle of the impedance
almost alwyas approaches, but exceedingly ever wanders
outside the +-45 degree limits


Sorry, I meant to write "only exceedingly rarely" instead
of the gobbldeygook I did.


But some are and some go to very low impedances at some point of their
frequency response. I'm not sure if double is required, but it can't
hurt.

I have a pair of Apogee Divas hooked up to Classe power amps. I don't
recall the specifics, but I remember they get to some ridiculous low
impedance and the Classe amps will double output when impedance drops in
half. They work with no audible strain.

I can't say something less wouldn't work as well, but I know anything
way less, like a Dynaco Stereo 70, won't because I fried one playing
with it.

--
Robert B. Peirce, Venetia, PA 724-941-6883
bob AT peirce-family.com [Mac]
rbp AT cooksonpeirce.com [Office]
  #22   Report Post  
Posted to rec.audio.high-end
Dick Pierce[_2_] Dick Pierce[_2_] is offline
external usenet poster
 
Posts: 151
Default Do power amps need to deliver twice the power into half the

Robert Peirce wrote:
In article ,
Dick Pierce wrote:


Dick Pierce wrote:

For the VAST majority of speakers hooked to the VAST
majority of amplifiers, the impedance almost never is
dominantly reactive. The phase angle of the impedance
almost alwyas approaches, but exceedingly ever wanders
outside the +-45 degree limits


Sorry, I meant to write "only exceedingly rarely" instead
of the gobbldeygook I did.



But some are and some go to very low impedances at some point of their
frequency response. I'm not sure if double is required, but it can't
hurt.

I have a pair of Apogee Divas hooked up to Classe power amps. I don't
recall the specifics, but I remember they get to some ridiculous low
impedance and the Classe amps will double output when impedance drops in
half. They work with no audible strain.


You missed the point in several dimensions.

First, they may well get to a few femtoOhms impedance (they
don't). That does not necessarily make them "exceedingly
reactive." That makes them low impedance. Low impedance and
"exceedingly" reactive are NOT equivalent. Even normal impedance
at some frequencies and low at others does not constitute
"exceedingly reactive." Over the substantial portions bandwidth
of the system, does the phase angle of the impedance exceeding
+-45 degrees (it does not)? If not, they are not "exceedingly
reactive."

In fact, the impedance curves of the Divas is probably as close
to proving my point as anything you could have picked: the present
about as close to a resistive load over the entire bandwidth as
any speaker could imaginably do without restoring to complex
compensation networks. Their impedance varies from about 3.5
to about 4,5 ohms, and the phase angle seldom exceeds +-20
degrees.

Secondly, Apogee Divas are but one of several hundreds of
high end speaker models. And, given the numbers of them out
there, the comprise a vanishing, insignificant minority of
the speakers available. So even if the impedance phase angle
spent most of its time between +- 45 degress and +- 90 degrees
(which is most assuredly does not), that does not mean that,
as the poster claimed, "speakers are highly, exceedingly
reactive." That generalization is simply at variance with
the known physical facts.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

  #23   Report Post  
Posted to rec.audio.high-end
mcdonaldREMOVE TO ACTUALLY REACH [email protected] mcdonaldREMOVE TO ACTUALLY REACH ME@scs.uiuc.edu is offline
external usenet poster
 
Posts: 42
Default Do power amps need to deliver twice the power into half the

Dick Pierce wrote:

In fact, the impedance curves of the Divas is probably as close
to proving my point as anything you could have picked: the present
about as close to a resistive load over the entire bandwidth as
any speaker could imaginably do without restoring to complex
compensation networks. Their impedance varies from about 3.5
to about 4,5 ohms, and the phase angle seldom exceeds +-20
degrees.

Secondly, Apogee Divas are but one of several hundreds of
high end speaker models. And, given the numbers of them out
there, the comprise a vanishing, insignificant minority of
the speakers available. So even if the impedance phase angle
spent most of its time between +- 45 degress and +- 90 degrees
(which is most assuredly does not), that does not mean that,
as the poster claimed, "speakers are highly, exceedingly
reactive." That generalization is simply at variance with
the known physical facts.


We simply took different views on "exceedingly reactive". I suppose
you are right, though you have to admit 90 degrees is "exceedingly
reactive" and I'd say 80 degrees was too. I was considering 60 degrees,
which some speakers reach (and not at ultrasonic frequencies),
"exceedingly reactive". I do agree that most speakers stop at roughly 45 degrees.

In any case, unless you put in an 8 ohm series resistance, it
does not matter.

Doug McDonald
  #24   Report Post  
Posted to rec.audio.high-end
Dick Pierce[_2_] Dick Pierce[_2_] is offline
external usenet poster
 
Posts: 151
Default Do power amps need to deliver twice the power into half the

"mcdonaldREMOVE TO ACTUALLY REACH wrote:
Dick Pierce wrote:

In fact, the impedance curves of the Divas is probably as close
to proving my point as anything you could have picked: the present
about as close to a resistive load over the entire bandwidth as
any speaker could imaginably do without restoring to complex
compensation networks. Their impedance varies from about 3.5
to about 4,5 ohms, and the phase angle seldom exceeds +-20
degrees.

Secondly, Apogee Divas are but one of several hundreds of
high end speaker models. And, given the numbers of them out
there, the comprise a vanishing, insignificant minority of
the speakers available. So even if the impedance phase angle
spent most of its time between +- 45 degress and +- 90 degrees
(which is most assuredly does not), that does not mean that,
as the poster claimed, "speakers are highly, exceedingly
reactive." That generalization is simply at variance with
the known physical facts.


We simply took different views on "exceedingly reactive".


Luckily, there are objective definitions for reactivity
on which we can reliably depend.

I suppose you are right, though you have to admit 90
degrees is "exceedingly reactive"


And, there does not exist a single speaker anywhere that
comes close to 90 degrees total phase shift. None. That
is something that can be said with certainty.

and I'd say 80 degrees was too.


And I have, in the thousands of speaker systems I have
measured, to the best of my memory, never encountered
a speaker that reached 80 degrees.

I was considering 60 degrees, which some speakers reach


Very few speakers exceed 45 degrees, and then only over
very narrow ranges of frequency.

(and not at ultrasonic frequencies), "exceedingly reactive".
I do agree that most speakers stop at roughly 45 degrees.


No, they may approach it, but few if any ever make it there
in any substantive fashion.

Simply stated, if the phase shift of the impedance is
between -45 degrees and +45 degrees, the impedance is
dominated by the resistive portion. Only when it exceeds
+-45 degrees does the reactive portion of the impedance
dominate.

And it is NEVER +-90 degrees and it NEVER exceeds +-90
degrees. The reasons why are left as an excercise to
the student. :-)

In any case, unless you put in an 8 ohm series resistance, it
does not matter.


Or, you hook it to some amplifier designed by some incompent
high-end loonie, of which there are a disturbing number of
examples.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

  #25   Report Post  
Posted to rec.audio.high-end
Trevor Wilson Trevor Wilson is offline
external usenet poster
 
Posts: 776
Default Do power amps need to deliver twice the power into half the impedance to be any good?

"Arny Krueger" wrote in message
...
It has been my contention that power amps that double their power as
impedance halves are the product of manipulation of specifications for
marketing, not how real-world audio technology actually works.


**Of course. It is impossible for an amplifier to manage such a feat in the
real world. OTOH, many manufacturers deliberately understate the 8 Ohm (and
4 Ohm) power levels, such that it APPEARS that they are doubling their
maximum power into successively lower impedances.

Here's a
real world data point for the controversy about power amps needing to be
able to put out twice the power for half the impedance:

I ran some bench tests on what may be one of the most overbuilt power amps
in the history of audio - the Threshold SA/4e from the early 1990s.

No visible effort was spared to build a robust 100 wpc power amp, as this
article shows:

http://www.thresholdlovers.com/artic...?lng=en&pg=394

The Threshold SA/4e weighs about 110 pounds, according to my calibrated
biceps. You really want two people to tote it around, and it facilitates
this by having two comfortable handles on each end.

In contrast, in similar tests a Behringer A500 which weighs about 15
pounds
puts out about the same amount of power at clipping at 8 ohms, but puts
out
about 144 watts at both 4 and 2 ohms. This is what I would expect from a
typical contemporary mainstream amplifier.

Among other features, the Threshold SA/4e amplifier has extremely soft
clipping. While most power amps, even tubed power amps have a fairly
well-defined clipping point on the bench, not so with the SA/4e. In the
end
I decided to measure its output power at about 0.1% THD @ 1KHz, as
observed
with a RTA.

The actual measured power is as listed he

http://cgi.audiogon.com/cgi-bin/auc....82027&auc&3&4&

Namely 112 Watts per channel @ 8 ohms with both channels driven @ 1000 Hz,
196 Watts @ 4 ohms, 264 Watts @ 2 ohms.

This is still very far from doubling power as load impedance halves,
particularly at 2 ohms.


**Of course. That would be an impossibility. Saturation effects in output
devices precludes such a thing.

BTW: Did Mr Pass ever claim such a thing about his amplifiers?

--
Trevor Wilson
www.rageaudio.com.au


  #26   Report Post  
Posted to rec.audio.high-end
Robert Peirce[_2_] Robert Peirce[_2_] is offline
external usenet poster
 
Posts: 15
Default Do power amps need to deliver twice the power into half the

In article ,
Dick Pierce wrote:

You missed the point in several dimensions.

First, they may well get to a few femtoOhms impedance (they
don't). That does not necessarily make them "exceedingly
reactive."


You are correct.

--
Robert B. Peirce, Venetia, PA 724-941-6883
bob AT peirce-family.com [Mac]
rbp AT cooksonpeirce.com [Office]
  #27   Report Post  
Posted to rec.audio.high-end
mpresley mpresley is offline
external usenet poster
 
Posts: 2
Default Do power amps need to deliver twice the power into half the

The best scientific explanation of real world power in amplification:

http://www.youtube.com/watch?v=EbVKWCpNFhY

[ For people who don't want to click through without some
context -- it's the "Goes to 11" scene from "This is Spinal
Tap" -- dsr ]

Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
WTB: power transformers (2) for UREI half rack moules Tim Luce Pro Audio 1 April 15th 06 02:27 AM
45Z3 - Diode Half wave power rectifier Engineer Vacuum Tubes 0 July 8th 04 02:00 AM
How Do I Determine A Power-Amp's Input Impedance? Clytemnestra Tech 180 January 18th 04 04:20 PM
How Do I Determine A Power-Amp's Input Impedance? Clytemnestra High End Audio 1 January 14th 04 12:32 AM


All times are GMT +1. The time now is 06:41 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"