[Mat Nieuwenhoven]

On Fri, 26 Apr 2019 18:52:00 -0700 (PDT), wrote:

Don Pearce wrote:


Nope. Didn't understand a word of that.



** Really ??

So " I squared R " has no meaning in your world ?

The resistive losses that increase a speaker's mid band, resistive impedance beyond the DC ohms value are also mysterious to you ?

A multiplying power meter, using analogue multiplier ICs, is also a mystery ?

Where have you been hiding Don.

Under a rock?



.... Phil

I squared R has plenty of meaning. Unfortunately a speaker is not an
R. It's an X. And I squared X gives you VA, not Watts. Your
measurement method, applied to a pure capacitor, would apparently
yield a power level. You can't put any power into a capacitor.

d


Oops - let me correct that. A speaker is a whole slew of different R +
jX, with the values being different at every frequency over the
measurement band. No scalar instrument can tell you about power
transfer into that. You can do it with a vector instrument at a single
frequency - and I regularly do that.


But a noise signal with a scalar voltmeter and a scalar ammeter? Nope.
Not on this earth.



** You are making a very basic mistake.

Consider the following:

If you apply a noise voltage to a resistor and measure the current flow with a wide band, true rms meter the formula "I squared R " applies exactly.

If you add an ideal reactance in series with the resistor the formula "I squared R " still applies since the resistor is the only power dissipating item.

A speaker can be accurately modelled as several resistances in series with a lossless reactive component, also in series. snip

Can it? I thought the "Wright emperical model" is the standard model
since 1990.
For interested readers, see
http://www.wavecor.com/Transducer_eq...nt_circuit.pdf for a 1 page
overview
or
https://www.researchgate.net/publica...pirical_Model_
for_Loudspeaker_Motor_Impedance

Wavecor specifically states that the traditional model does not do a
good job at non-low frequencies.

Mat Nieuwenhoven