On 2/7/2017 8:43 PM, Phil Allison wrote:
Mike Rivers wrote:
No. Only someone ignorant of how loudspeakers are rated would interpret
it as the maximum power rating of the amplifier that's driving it.
** But that is often how speaker makers come up with their inflated numbers.
Its the amp power ( usually rms sine wave) that they figure is OK with the speaker under normal programme conditions with little or no clipping. The average power being delivered is then at most 1/4 ( -6dB )of the amp's rated power.
At lot hen depends on what the speaker's intended use, ie home hi-fi, live sound, disco or guitar.
poorly built and stupidly rated speaker (both cases are not uncommon)
driven by a 50 watt amplifier that's clipping badly is more likely to
suffer damage than when driven by a 500 watt amplifier that's
undistorted when running 100 watts.
** Well, a 50watt sine wave rated amp will deliver at most 100W when heavily clipped - so matching the output of the 500watt model unclipped.
The idea of using a LARGER amp to solve speaker burn out problems is a complete nonsense.
Section 3.2 of my article on speaker failures makes the case pretty clear.
http://sound.whsites.net/articles/speaker-failure.html
From your link: "Nominal Impedance = R plus 15% - where R is the
DC resistance in ohms
The "nominal impedance" of a woofer or instrument speaker is the
LOWEST value of the REAL impedance that driver exhibits in the audio
range and at room temperature. The actual minimum typically occurs in
the band between 200 Hz and 500 Hz and the usual test frequencies are
250 Hz or 400 Hz.
Being an impedance minimum means that it is a pure resistance too, with
current and voltage in phase. The extra 15% comes from energy losses in
the suspension, eddy currents the iron magnet structure and radiated sound."
So this estimated nominal impedance ignores the inductive
reactance of the voice coil, because it assumes lower testing
frequencies, where XL=2pifL will be low, and can be ignored?
Also from your link: "A "nominal watt" is based purely on a
simple, but absurd, calculation that assumes the speaker maintains it
nominal impedance at all frequencies and under all operating conditions.
The usual power handling test done on a high powered woofer is to
install it in a large cabinet or perhaps in free air, and feed it with
modified pink noise filtered to the 50 Hz to 500 Hz band or possibly the
50 Hz to 5,000 Hz band. (See note below.)
The output level from the amp is then adjusted upwards until the voice
coil is dangerously hot and left like that for a couple of hours. The
RMS voltage being delivered by the amp is measured, the value squared
and divided by the nominal impedance to give "max watts". See AES-2 1984
"Speaker Testing" link 2.
As a result of this patent absurdity - the actual watts dissipated by
the speaker during such testing may well be only 20 to 25% of the
published max watts figure."
I assume trying to calculate Power from Voltage x Current by
putting a current sensing resistor of 0.1 Ohm in series with the
speaker under test, would affect the measurements too much?
How hot is a "dangerously hot" voice coil?
Seems to me the best thing to do would be to measure
the RMS voltages, while incrementally increasing the pink noise output
level from the amplifier, until the speakers blows. Then you
could use the last measured RMS voltage (when the speaker still
survived) in the calculation.