Speaker impedance vs output power

[kubeek]

Recently someone told me that speaker impedance doubles when it approaches maximum power. Do you guys have any info about that, or maybe even a set of impedance graphs at various ouput power?
That would be the best I could imagne to get me some insight on this

[nigelwright7557]

Speaker impedance usually varies with frequency.

If xmax is passed then the impedance could change.

[kubeek]

There is no doubt about frequency, but i was wondering if the frequency-impedance plot changes significantly with power.

[Cal Weldon]

I wonder if they are referring to Power compression? Here's a little snippet I grabbed from the web:

Quote:
What is "power compression"? Speaker voice coils are made of copper or aluminum. As these voice coils increase in temperature during normal operation, the DC resistance of the voice coil increases. Greater voice coil resistance means less power transfer from the amplifier. As a result, the speaker will not play as loud when it's "warmed up" as it did when it was "cold". Some speakers may exhibit 3 to 6 dB of power compression. This means that power compression can have the same effect as taking away half of your PA!
End quote

[chris661]

Yeah, Cal's on to it.

If you lose 3dB, your impedance has doubled. Now imagine what that's done to the Qts that's helped you get that flat bass response...

[weltersys]

Originally Posted by chris661

Yeah, Cal's on to it.

If you lose 3dB, your impedance has doubled. Now imagine what that's done to the Qts that's helped you get that flat bass response...

And when the impedance changes, so does a passive crossover frequency...

[sofaspud]

-3dB and a doubling of impedance seems like an awful lot from a PTC. Is this a more extreme example of the concept? If so, does someone know a "what you can expect" figure for typical 6"-8" loudspeakers?

[weltersys]

Originally Posted by sofaspud

-3dB and a doubling of impedance seems like an awful lot from a PTC. Is this a more extreme example of the concept? If so, does someone know a "what you can expect" figure for typical 6"-8" loudspeakers?

You can expect 2-3 dB power compression when the speaker is run near the average rated power level .

Heavily compressed program (or a hard clipped amp) can hit that kind of average, but music with a wide dynamic range has only a fraction of the average power compared to peak.

A 28.3 volt sine wave into an eight ohm speaker is 100 watts, this could easily result in 2-3 dB power compression in a matter of seconds. You would be hard pressed to find any music with that much average power, with a more normal 10 dB dynamic range you would need 1000 watt peaks to average 100 watts.

If your "typical" speaker is driven from an un clipped 100 watt amp, the average power is only 10 watts from music with a 10 dB dynamic range, virtually no power compression would take place.

[Top Shelf]

[QUOTE=

If you lose 3dB, your impedance has doubled. Now imagine what that's done to the Qts that's helped you get that flat bass response...[/QUOTE]

I find that adding 10% to the port length helps to combat the upward shift in tuning/ Fs/ Qts due to Voice coil heating on my PA speakers. I also get the pole piece vent as close to the port as possible/ or pipe it there to help with VC cooling.

[sofaspud]

Thanks, weltersys. There was a recent thread concerning rated speaker power here and I don't believe power compression was brought up. I think the two threads provide a good summary of the situation.