[cporro]

so a speaker with a sensitivity of 93dB @ 1w / 1 meter means 1 watt of
power will drive an input signal at 1khz (sine wave?) to a spl of 93db
at 1 meter?

but isn't a 1khz test tone unrealistic? wouldn't pink noise be a far
better test? and wouldn't pink noise require a lot more power for the
same volume?

and couldn't i figure out how much power i need in an amp much better
if pink noise was used?

the other number is the power range or rating. that is supposed to say
how much power the speaker can handle before it erupts? and then on
the low side the practical minimum required to drive it?

[Don Pearce[_3_]]

On Fri, 23 Dec 2011 13:50:53 -0800 (PST), cporro
wrote:

so a speaker with a sensitivity of 93dB @ 1w / 1 meter means 1 watt of
power will drive an input signal at 1khz (sine wave?) to a spl of 93db
at 1 meter?

but isn't a 1khz test tone unrealistic? wouldn't pink noise be a far
better test? and wouldn't pink noise require a lot more power for the
same volume?

and couldn't i figure out how much power i need in an amp much better
if pink noise was used?

the other number is the power range or rating. that is supposed to say
how much power the speaker can handle before it erupts? and then on
the low side the practical minimum required to drive it?


So close to Christmas. So large a can of worms.

d

[cporro]

yeah i know. just trying to find some meaning in the number. hoho.

[Arny Krueger[_4_]]

"cporro" wrote in message
...

so a speaker with a sensitivity of 93dB @ 1w / 1 meter means 1 watt of
power will drive an input signal at 1khz (sine wave?) to a spl of 93db
at 1 meter?

In an anechoic room...

but isn't a 1 KHz test tone unrealistic?

Test was probably done with a band of noise centered at 1 KHz.

wouldn't pink noise be a far better test?

Or, a warble tone.

But not pink noise because pink noise is wideband noise, and this spec is
about a narrow band of frequencies around 1 KHz.


and wouldn't pink noise require a lot more power for the
same volume?

No because we would measure the tone going into the speaker and the sound
coming out of it the same way.

and couldn't i figure out how much power i need in an amp much better
if pink noise was used?

Not really.

the other number is the power range or rating. that is supposed to say
how much power the speaker can handle before it erupts? and then on
the low side the practical minimum required to drive it?

The minimum power needed is very much dependent on the application, which
includes room, location of listeners, type of program material, actual
bandwidth of speakers, etc.


The way I look at speaker specs is that if one speaker is rated at 93 dB/W
and another is rated at say 96 dB/W and they are very similar otherwise,
then I know I could possibly get away with a smaller amp with the more
efficient speaker.

[William Sommerwerck]

"cporro" wrote in message
...

so a speaker with a sensitivity of 93dB @ 1w / 1 meter means
1 watt of power will drive an input signal at 1khz (sine wave?)
to an SPL of 93db at 1 meter?

First of all, that is not a sensitivity rating (which would be based on a
voltage input), but an efficiency rating. Regardless, such ratings give a
fair idea of how loud the speaker will play for a given input. High
sensitivity or efficiency is not, in and of itself, a mark of quality.

I'm old enough to remember when 60W/ch was high power, and AR speakers were
considered uncomfortably inefficient. Today you can buy $300 receivers that
are spec'd at 100W/ch -- for 8 channels.


but isn't a 1khz test tone unrealistic? wouldn't pink noise be a far
better test? and wouldn't pink noise require a lot more power for the
same volume?

That's too complicated an issue to give a complete and and honest answer to.
Since the output spec is based on a (sort-or) well-defined standard, you can
make a reasonably valid comparison among speakers.


and couldn't i figure out how much power i need in an amp much better
if pink noise was used?

The only way you can know for sure is to "PLAY REAL LOUD!".


the other number is the power range or rating. that is supposed to say
how much power the speaker can handle before it erupts? and then on
the low side the practical minimum required to drive it?

Power-handling ratings are largely meaningless. Most buyers misinterpret
them as indicating how much /continuous/ power the speaker can handle. If
you drove the speaker at such a level, it would be badly damaged.

What the spec is /supposed/ to mean is that, if you drove the speaker with
an amp of the stated value, the speaker could handle the amp's instantaneous
/peak/ output without being damaged or audibly distorting. But the speaker's
power-handling capacity varies with the music. The peak level of highly
compressed music is only slightly above the average level, whereas with jazz
or classical, it's about 16dB above. So, for a given peak power level,
rotten'roll is far more likely to blow out the speaker.

I don't know how common the practice is, but rock listeners are notorious
for cranking up the volume past the point of audible distortion. William
Michael Watson Dayton-Wright told me that every time he improved the
power-handling capacity of his ICBM speaker, rock listeners kept damaging
it.

[Trevor]

"William Sommerwerck" wrote in message
...
Today you can buy $300 receivers that
are spec'd at 100W/ch -- for 8 channels.

Spec'd perhaps. Not all channels at once though, not for very long either,
and usually not an "RMS" rating.
(yes I know that term is technically incorrect but widely used anyway)

Trevor.

[Mike Rivers]

On 12/23/2011 5:22 PM, cporro wrote:
yeah i know. just trying to find some meaning in the number. hoho.

Don't try too hard. It's just a number that manufacturers
have to publish because there are people like you, so they do.

I've always used pink noise when calibrating monitor level
settings, for what that's worth. I've never had a problem
with underpowering speakers because I usually have at least
25 watts behind them and 93 dB SPL is plenty loud enough for me.


--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[William Sommerwerck]

"Trevor" wrote in message
...
"William Sommerwerck" wrote in message
...

Today you can buy $300 receivers that
are spec'd at 100W/ch -- for 8 channels.

Spec'd perhaps. Not all channels at once though, not for very long either,
and usually not an "RMS" rating.
(Yes, I know that term is technically incorrect, but widely used anyway.)

The correct term is "continuous average".

If we start using the correct term, we perhaps can change the world!

[Don Pearce[_3_]]

On Sat, 24 Dec 2011 04:01:01 -0800, "William Sommerwerck"
wrote:

"Trevor" wrote in message
...
"William Sommerwerck" wrote in message
...

Today you can buy $300 receivers that
are spec'd at 100W/ch -- for 8 channels.

Spec'd perhaps. Not all channels at once though, not for very long either,
and usually not an "RMS" rating.
(Yes, I know that term is technically incorrect, but widely used anyway.)

The correct term is "continuous average".

If we start using the correct term, we perhaps can change the world!


Average will do - no need for continuous. Of course there may be a
difference when it comes to short term and long term - temperature
profiles will be a factor in this.

d

[William Sommerwerck]

Average will do -- no need for continuous.

Average, by itself, is ambiguous.

[Don Pearce[_3_]]

On Sat, 24 Dec 2011 04:34:51 -0800, "William Sommerwerck"
wrote:

Average will do -- no need for continuous.

Average, by itself, is ambiguous.


Continuous, when added to average, is meaningless. What, for example,
happens if I switch the system off for a while? The continuous average
is going to drop.

Average power is easy. Just measure at least one full cycle of signal,
take the average and you have the number. Making the measurement
continuous adds nothing but confirmation.

d

[cporro]

thx all. so it sounds like the test signal at 1khz actually is
noise...just centered around 1khz. that makes more sense. and should
be more useful. no?

and the power rating is based on a peak maximum power handling. not an
rms value. ok.

its confusing because for these speakers there is this huge range. a
recommended power rating of 10-200 watts. and a sensitivity (er,
efficiency) of 93db at 1watt. the active speakers have a 60 watts
RMS@ .005% THD (1KHz) amp.

if i monitor at 63db which i do sometimes then a 1 watt amp could
power these with 30db headroom. right?

and a 200 watt amp would get me...110 db at 1 meter? ouch.

so what is a practical way to look at these numbers? btw, i already
have the amp, bryston 2b. overkill? for sure but might come in handy
up the road wither other speakers.

[Don Pearce[_3_]]

On Sat, 24 Dec 2011 07:13:20 -0800 (PST), cporro
wrote:

thx all. so it sounds like the test signal at 1khz actually is
noise...just centered around 1khz. that makes more sense. and should
be more useful. no?

and the power rating is based on a peak maximum power handling. not an
rms value. ok.

its confusing because for these speakers there is this huge range. a
recommended power rating of 10-200 watts. and a sensitivity (er,
efficiency) of 93db at 1watt. the active speakers have a 60 watts
RMS@ .005% THD (1KHz) amp.

if i monitor at 63db which i do sometimes then a 1 watt amp could
power these with 30db headroom. right?

and a 200 watt amp would get me...110 db at 1 meter? ouch.

so what is a practical way to look at these numbers? btw, i already
have the amp, bryston 2b. overkill? for sure but might come in handy
up the road wither other speakers.

The practical way is to take your average listening level - 63dB (or
probably more like 73dB if you think about it). That is just an
average. Add 20dB for when you feel like winding it up a bit - and you
will - making 93dB. That's a watt.Now you need to add some more to
accommodate the peaks above average, accounting for perhaps another
20dB.

So a 100 watt amp would give you just about an ideal amount of power
with some in reserve. 100 watt amps are plentiful and cheap for a very
good reason.

d

[Scott Dorsey]

cporro wrote:
thx all. so it sounds like the test signal at 1khz actually is
noise...just centered around 1khz. that makes more sense. and should
be more useful. no?

Could be. Might not be. There's no FTC standard for this like there is
power amplifier ratings.

and the power rating is based on a peak maximum power handling. not an
rms value. ok.

No, the power rating is made up by someone in the marketing department
and has no bearing on any actual speaker measurements.

its confusing because for these speakers there is this huge range. a
recommended power rating of 10-200 watts. and a sensitivity (er,
efficiency) of 93db at 1watt. the active speakers have a 60 watts
RMS@ .005% THD (1KHz) amp.

if i monitor at 63db which i do sometimes then a 1 watt amp could
power these with 30db headroom. right?

and a 200 watt amp would get me...110 db at 1 meter? ouch.

Right. If you were at one meter in an anechoic chamber. Since you're in
a real room, it will probably be louder than that.

But, you'll find that since the speaker impedance varies with frequency,
the speaker will tend to pull more power at lower frequencies near the
box resonance, where the efficiency will be poorer.

so what is a practical way to look at these numbers? btw, i already
have the amp, bryston 2b. overkill? for sure but might come in handy
up the road wither other speakers.

Ignore them completely.

Right now I am using a pair of Magnepan MG1.4s, which have a sticker on
the back warning you of dire consequences if you run them with an amplifier
of less than 250W. I've been running them off an old 60-watt Citation II
for the past 20 years and they've been fine.

Try the amp. If it sounds good, it's fine. If it clips, turn it down.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Peter Larsen[_3_]]

William Sommerwerck wrote:

... The peak level of highly compressed music is only slightly
above the average level, whereas with jazz or classical, it's about
16dB above.

Not correct, mostly you should add 10 dB to that, at least for resonably
natural appearing music. Generally speaking a peak to average below 25 dB
needs explaining, but 18 or so dB is not unlikely for church organ and bass
heavy rock. There may be more than one way to define "average", in this
context I define it as "average" as per Cool Edits statistics with a 300 ms
time window. Quite many previous posts on this are googlable.

Kind regards

Peter Larsen

[Peter Larsen[_3_]]

Trevor wrote:

"William Sommerwerck" wrote in message
...
Today you can buy $300 receivers that
are spec'd at 100W/ch -- for 8 channels.

Spec'd perhaps. Not all channels at once though, not for very long
either, and usually not an "RMS" rating.
(yes I know that term is technically incorrect but widely used anyway)

Generally it is "any two channels driven" somewhere in the small print.

Trevor.

Kind regards

Peter Larsen

[cporro]

hhhh.

sorry i asked. who needs numbers anyway?

[William Sommerwerck]

"Peter Larsen" wrote in message
k...
William Sommerwerck wrote:

... The peak level of highly compressed music is only slightly
above the average level, whereas with jazz or classical, it's about
16dB above.

Not correct, mostly you should add 10 dB to that, at least for resonably
natural appearing music. Generally speaking a peak to average below 25 dB
needs explaining, but 18 or so dB is not unlikely for church organ and
bass
heavy rock. There may be more than one way to define "average", in this
context I define it as "average" as per Cool Edits statistics with a 300
ms
time window. Quite many previous posts on this are googlable.

Actually, I should retract my statement. I was thinking in terms of crest
factor, which really applies only to solo instruments..

[Don Pearce[_3_]]

On Sat, 24 Dec 2011 08:45:33 -0800, "William Sommerwerck"
wrote:

"Peter Larsen" wrote in message
. dk...
William Sommerwerck wrote:

... The peak level of highly compressed music is only slightly
above the average level, whereas with jazz or classical, it's about
16dB above.

Not correct, mostly you should add 10 dB to that, at least for resonably
natural appearing music. Generally speaking a peak to average below 25 dB
needs explaining, but 18 or so dB is not unlikely for church organ and
bass
heavy rock. There may be more than one way to define "average", in this
context I define it as "average" as per Cool Edits statistics with a 300
ms
time window. Quite many previous posts on this are googlable.

Actually, I should retract my statement. I was thinking in terms of crest
factor, which really applies only to solo instruments..


No, crest factor is a function of a waveform - any waveform. It is the
ratio of the peak value to the RMS value. The waveform can be a sine
wave, square wave, a voice, one instrument or many instruments. It
still has a crest factor, which is still calculated just the same way.

d

[William Sommerwerck]

"Don Pearce" wrote in message
...
On Sat, 24 Dec 2011 08:45:33 -0800, "William Sommerwerck"
wrote:

"Peter Larsen" wrote in message
. dk...
William Sommerwerck wrote:

... The peak level of highly compressed music is only slightly
above the average level, whereas with jazz or classical, it's about
16dB above.

Not correct, mostly you should add 10 dB to that, at least for
resonably
natural appearing music. Generally speaking a peak to average below 25
dB
needs explaining, but 18 or so dB is not unlikely for church organ and
bass
heavy rock. There may be more than one way to define "average", in this
context I define it as "average" as per Cool Edits statistics with a
300
ms
time window. Quite many previous posts on this are googlable.

Actually, I should retract my statement. I was thinking in terms of crest
factor, which really applies only to solo instruments..

No, crest factor is a function of a waveform - any waveform. It is the
ratio of the peak value to the RMS value. The waveform can be a sine
wave, square wave, a voice, one instrument or many instruments. It
still has a crest factor, which is still calculated just the same way.

Correct, but I was thinking in terms of uncompressed waveforms. Regardless,
my example was not appropriate.

[Sean Conolly]

"Arny Krueger" wrote in message
...

The minimum power needed is very much dependent on the application, which
includes room, location of listeners, type of program material, actual
bandwidth of speakers, etc.

But in general, people tend to drastically overestimate how much power they
need, or just how loud a single watt of power can be. Many years ago I put
together a stupid cheap system for my father, using a little Radio Shack amp
with a whopping 1.5W per side, and he loved it. The speakers were very
efficient and were rated at about 15W max, but they sounded better than a
lot of stuff available at the time. It wasn't a disco, but it was still
pretty convincing with the 1812 overture.


The way I look at speaker specs is that if one speaker is rated at 93 dB/W
and another is rated at say 96 dB/W and they are very similar otherwise,
then I know I could possibly get away with a smaller amp with the more
efficient speaker.

Yes a higher efficiency speaker requires less powerful amp, but for
nearfields where you're sitting about a meter away from the speakers (or
less) you really need an amp which sounds very good at single digit watts.
At 200 watts I'd be much more concerned about the impact on ME than the
speakers!

Sean

[William Sommerwerck]

"Sean Conolly" wrote in message
...
"Arny Krueger" wrote in message
...

The minimum power needed is very much dependent on the application, which
includes room, location of listeners, type of program material, actual
bandwidth of speakers, etc.

But in general, people tend to drastically overestimate how much power
they
need, or just how loud a single watt of power can be. Many years ago I put
together a stupid cheap system for my father, using a little Radio Shack
amp
with a whopping 1.5W per side, and he loved it. The speakers were very
efficient and were rated at about 15W max, but they sounded better than a
lot of stuff available at the time. It wasn't a disco, but it was still
pretty convincing with the 1812 overture.

I own large planar speakers, biamped, and even at high volume levels,
there's only a couple of volts across them.

[Trevor]

"William Sommerwerck" wrote in message
...
Spec'd perhaps. Not all channels at once though, not for very long
either,
and usually not an "RMS" rating.
(Yes, I know that term is technically incorrect, but widely used anyway.)

The correct term is "continuous average".

If we start using the correct term, we perhaps can change the world!

I gave that up long ago. If you say "maximum continuous average power" most
people don't understand, and I simply don't have the desire or patience to
teach them any more.
The majority of amplifier buyers don't even know what RMS stands for, so
it's a pontless exercise in any case.

Trevor.

[Trevor]

"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.


Of course there may be a
difference when it comes to short term and long term - temperature
profiles will be a factor in this.

Right, as well as power supply regulation etc.

Trevor.

[Trevor]

"Don Pearce" wrote in message
...
Continuous, when added to average, is meaningless. What, for example,
happens if I switch the system off for a while? The continuous average
is going to drop.

Well duh, obviously you measure while the power is applied, and for a known
period of more than ONE lousy cycle at 1kHz!


Average power is easy. Just measure at least one full cycle of signal,
take the average and you have the number. Making the measurement
continuous adds nothing but confirmation.

Exactly the problem, power output for one cycle of a 1kHz sine wave is
*very* different from the continuous average.

Trevor.

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 19:12:14 +1100, "Trevor" wrote:

Average power is easy. Just measure at least one full cycle of signal,
take the average and you have the number. Making the measurement
continuous adds nothing but confirmation.

Exactly the problem, power output for one cycle of a 1kHz sine wave is
*very* different from the continuous average.

No it isn't, it is precisely the same. And if you think the frequency
is important, you have the whole thing seriously misconceived.

d

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 19:07:46 +1100, "Trevor" wrote:


"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.

No you are not. That is an entirely different subject.


Of course there may be a
difference when it comes to short term and long term - temperature
profiles will be a factor in this.

Right, as well as power supply regulation etc.

And that was it.

d

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 19:04:52 +1100, "Trevor" wrote:


"William Sommerwerck" wrote in message
...
Spec'd perhaps. Not all channels at once though, not for very long
either,
and usually not an "RMS" rating.
(Yes, I know that term is technically incorrect, but widely used anyway.)

The correct term is "continuous average".

If we start using the correct term, we perhaps can change the world!

I gave that up long ago. If you say "maximum continuous average power" most
people don't understand, and I simply don't have the desire or patience to
teach them any more.
The majority of amplifier buyers don't even know what RMS stands for, so
it's a pontless exercise in any case.

Trevor.


The product of RMS voltage and RMS current is average power. Provided
nobody tries to talk about RMS power, all is reasonably well.

d

[Trevor]

"Don Pearce" wrote in message
...
Exactly the problem, power output for one cycle of a 1kHz sine wave is
*very* different from the continuous average.

No it isn't, it is precisely the same.

You probably have the only amplifier in the world for which that is the
case. It MUST contain a regulated power supply for a start, which is usually
considered an expensive negative by most amplifier manufacturers.


And if you think the frequency
is important, you have the whole thing seriously misconceived.

Having measured MANY amps, I know only a few few can maintain their full
rated output at 20kHz or above, or 20Hz and below either. Obviously you
don't.
I guess the PMPO concept was invented for people just like you! :-)

Trevor.

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 19:44:23 +1100, "Trevor" wrote:


"Don Pearce" wrote in message
...
Exactly the problem, power output for one cycle of a 1kHz sine wave is
*very* different from the continuous average.

No it isn't, it is precisely the same.

You probably have the only amplifier in the world for which that is the
case. It MUST contain a regulated power supply for a start, which is usually
considered an expensive negative by most amplifier manufacturers.


OK, we are talking about different things. I am talking about the
power in a sine wave, you are talking about - well - something else to
do with unstable amplifiers. I have to tell you that I don't have any
amplifiers that change their gain from moment to moment in such a way
that measuring one cycle is different from measuring several.


And if you think the frequency
is important, you have the whole thing seriously misconceived.

Having measured MANY amps, I know only a few few can maintain their full
rated output at 20kHz or above, or 20Hz and below either. Obviously you
don't.
I guess the PMPO concept was invented for people just like you! :-)

Trevor.

Again this is an utterly different subject from the power in a sine
wave. Take a sine wave of a given RMS voltage and current, and the
power is independent of frequency.

d

[Trevor]

"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.

No you are not. That is an entirely different subject.


HOW is that a different subject, just because it makes your whole argument
wrong!
It is because of power supply regulation and transistor heating etc that
long term power measurement of amplifiers are significantly different from
the single cylce measurement you advocated when claiming 1mS (single cycle)
average was the same as long term average when clearly it is NOT.

Trevor.

[Trevor]

"Don Pearce" wrote in message
...
The product of RMS voltage and RMS current is average power. Provided
nobody tries to talk about RMS power, all is reasonably well.

Right, provided the person you are talking to is another engineer, all is
well. Most of those I deal with aren't, but have heard the term RMS power
many times.
As I said, life is too short for me to tech them electronic engineering for
free.

Trevor.

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 19:51:52 +1100, "Trevor" wrote:


"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.

No you are not. That is an entirely different subject.


HOW is that a different subject, just because it makes your whole argument
wrong!
It is because of power supply regulation and transistor heating etc that
long term power measurement of amplifiers are significantly different from
the single cylce measurement you advocated when claiming 1mS (single cycle)
average was the same as long term average when clearly it is NOT.

Trevor.


I am talking about measurement. You need to MEASURE only one complete
cycle to know the average power. If you can't hold the power constant
for the duration of the measurement, then you are going to be no
better off with a continuous measurement anyway.

d

[Trevor]

"Don Pearce" wrote in message
...
OK, we are talking about different things. I am talking about the
power in a sine wave, you are talking about - well - something else to
do with unstable amplifiers.

Well IF you want to rely to a statement on amplifier output power rating
with one that is obviously purely theoretical, it's up to YOU to say so
rather than waste our time on a completely unspecified argument.

I have to tell you that I don't have any
amplifiers that change their gain from moment to moment in such a way
that measuring one cycle is different from measuring several.

Are you now changing the argument from the original "long term maximum
average power output" to "gain" instead. How disingenious, since you could
scrap the mention of average or power as well as long term! :-)


Again this is an utterly different subject from the power in a sine
wave. Take a sine wave of a given RMS voltage and current, and the
power is independent of frequency.

Again this is an utterly different thing to what everyone else is talking
about, other than you it seems. The discussion was about REAL amplifier
power output ratings. And more specifically mention was made of surround
sound multi-channel reciever ratings. What you are on about and why is
anybody's guess.

Trevor.

[Trevor]

"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.

No you are not. That is an entirely different subject.


HOW is that a different subject, just because it makes your whole argument
wrong!
It is because of power supply regulation and transistor heating etc that
long term power measurement of amplifiers are significantly different from
the single cylce measurement you advocated when claiming 1mS (single
cycle)
average was the same as long term average when clearly it is NOT.


I am talking about measurement. You need to MEASURE only one complete
cycle to know the average power. If you can't hold the power constant
for the duration of the measurement, then you are going to be no
better off with a continuous measurement anyway.


That's the trouble with measurements made by people who have NO idea what
they are actually measuring. In this case we were talking about REAL
surround sound amplifier maximum power ratings, NOT whatever you seem to
think applies. I'm glad the IEC has different ideas to yours :-)

Trevor.

[Don Pearce[_3_]]

On Sun, 25 Dec 2011 20:13:29 +1100, "Trevor" wrote:


"Don Pearce" wrote in message
...
Average will do - no need for continuous.

Of course there is, or you are right back in the mS duration PMPO game.

No you are not. That is an entirely different subject.


HOW is that a different subject, just because it makes your whole argument
wrong!
It is because of power supply regulation and transistor heating etc that
long term power measurement of amplifiers are significantly different from
the single cylce measurement you advocated when claiming 1mS (single
cycle)
average was the same as long term average when clearly it is NOT.


I am talking about measurement. You need to MEASURE only one complete
cycle to know the average power. If you can't hold the power constant
for the duration of the measurement, then you are going to be no
better off with a continuous measurement anyway.


That's the trouble with measurements made by people who have NO idea what
they are actually measuring. In this case we were talking about REAL
surround sound amplifier maximum power ratings, NOT whatever you seem to
think applies. I'm glad the IEC has different ideas to yours :-)


No, I'm not having this. Power is power and there is no magic
involved. I stand by my statement that to measure power, you only need
to measure a single cycle. I'm not allowing any argument with that. If
you are measuring the maximum output power of an amplifier, it will
vary for any number of reasons - thermal effects, power supply sag
etc. We know that. That doesn't change what power is.

So when you are measuring this notional soggy amplifiers, you run a
sine wave through it and measure the power coming out. It is slowly
dropping as you measure. So what exactly is it you are measuring? Do
you keep measuring for, say, ten minutes and take the average power
over that time? Or should you only measure the reduced power coming
out at the end of that time?

If you choose to measure the compromised power at the end of that
time, you still only need to measure it over one cycle of the waveform
to get a true measurement. That doesn't change because the amplifier
has been running at full power for a while.

d

[Mark]

If you choose to measure the compromised power at the end of that
time, you still only need to measure it over one cycle of the waveform
to get a true measurement. That doesn't change because the amplifier
has been running at full power for a while.

d

Don,

the point is that for amps with wimpy power supply regulation, if you
apply a continuous tone, the power averaged over the FIRST cycle will
have the benefit of the charged up filter caps, after a second or so
the power supply rails will drop and the power averaged over the
1000th cycle will be less.

Since audio has a crest factor, the higher power at the start of the
burst , even over a short term duration is a benefit.

The fact that the power droops over time might even help protect your
speakers, you get the benefit of the higher power for the short term,
but the lower power over the long term might save your speakers.

Nothing is ever easy... :-)

Happy Holidays

Mark

[William Sommerwerck]

"Don Pearce" wrote in message
...

I am talking about measurement. You need to MEASURE only one
complete cycle to know the average power. If you can't hold the power
constant for the duration of the measurement, then you are going to
be no better off with a continuous measurement anyway.

This is true mathematically, but it has practical negative ramifications.
ahem

How many people here remember "music power" ratings? The claim was than an
amplifier could produce more power for a fraction of a second than it could
continuously.

Assuming this is true, it applies only to amplifiers with poor power-supply
regulation. The "better" amps, with large, stable supplies, gain little or
nothing in such a measurement.

[Scott Dorsey]

In article , Trevor wrote:
"Don Pearce" wrote in message
...
The product of RMS voltage and RMS current is average power. Provided
nobody tries to talk about RMS power, all is reasonably well.

Right, provided the person you are talking to is another engineer, all is
well. Most of those I deal with aren't, but have heard the term RMS power
many times.

So, why don't we just say "power measured by FTC method" which encompasses
a whole range of things from warmup time to average measurements to power
bandwidth?
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Les Cargill[_4_]]

Trevor wrote:
"Don wrote in message
...
Continuous, when added to average, is meaningless. What, for example,
happens if I switch the system off for a while? The continuous average
is going to drop.

Well duh, obviously you measure while the power is applied, and for a known
period of more than ONE lousy cycle at 1kHz!


Average power is easy. Just measure at least one full cycle of signal,
take the average and you have the number. Making the measurement
continuous adds nothing but confirmation.

Exactly the problem, power output for one cycle of a 1kHz sine wave is
*very* different from the continuous average.

Trevor.



I am not sure I agree with that...

---
Les Cargill