volume damage to KEF LS50 Meta

[underthetable]

Hello. I was measuring the KEF LS50 Meta, and accidentally played the REW -12dbFs pink noise signal without any extra volume control, and it was loud! I think that corresponds to around 19 Vrms delivered to the speakers, from these numbers: 0.25 (-12 dBFs) * 4 Vrms (DAC out max) * 19 (25.5 dB amp gain). That should end up at about 45W to 90W of power depending on the impedance, or about 100dB of SPL by looking at the speaker sensitivity. All this seems to be just within spec of the KEFs, I still wonder if the speakers are damaged. If anyone else any thoughts or suggestions on whether the speakers are fine, I'd be interested in hearing them.

 

[Sokel]

Short or long sweep?
90W of power is unlikely to damage a speaker unless is a tiny one.

Kef are well made, just measure them at one meter or so and see their distortion (this time at normal levels)

 

[pma]

Hello. I was measuring the KEF LS50 Meta, and accidentally played the REW -12dbFs pink noise signal without any extra volume control, and it was loud!

It should be OK if it was a pink noise. Try to measure distortion with a sine sweep at much lower amplitude, like 2 - 3 Vrms. You might like to check the amp output (without a speaker) with a DMM before measuring.

 

[underthetable]

So, this was

Short or long sweep?
90W of power is unlikely to damage a speaker unless is a tiny one.

Kef are well made, just measure them at one meter or so and see their distortion (this time at normal levels)

Thankfully, I didn't kick off a sweep at -12 dBFs, it was just 3 seconds of pink noise from REW's "Check levels" button in the "Measure" panel.

Good idea with regards to the distortion measurements, would that just be the distortion plots you get when you perform a REW measurement? I have those plots (attached) from measurements (at much lower level) I did after the loud pink noise. However, the measurements were done with the speaker far away from mic (>3m), and sitting high on top of a bookshelf. The second plot shows the measurement with some EQ applied via REW, and distortion is increased.

 

[underthetable]

It should be OK if it was a pink noise. Try to measure distortion with a sine sweep at much lower amplitude, like 2 - 3 Vrms. You might like to check the amp output (without a speaker) with a DMM before measuring.

I hope so too! I'll measure again at 1m, and post updated plots. I have some plots from a measurement after the loud pink noise, but the speaker and mic are in listening positions. I'll double check the output of the amp also.

 

[underthetable]

So, this was

Thankfully, I didn't kick off a sweep at -12 dBFs, it was just 3 seconds of pink noise from REW's "Check levels" button in the "Measure" panel.

Good idea with regards to the distortion measurements, would that just be the distortion plots you get when you perform a REW measurement? I have those plots (attached) from measurements (at much lower level) I did after the loud pink noise. However, the measurements were done with the speaker far away from mic (>3m), and sitting high on top of a bookshelf. The second plot shows the measurement with some EQ applied via REW, and distortion is increased.

Unless the increased distortion is from the fact that REW doesn't take into account the applied filters when calculating distortion?

 

[staticV3]

The second plot shows the measurement with some EQ applied via REW, and distortion is increased.

The second plot looks like no preamp was applied to the EQ settings, and so the digital signal clipped.
Definitely not natural.

 

[Sokel]

So, this was

Thankfully, I didn't kick off a sweep at -12 dBFs, it was just 3 seconds of pink noise from REW's "Check levels" button in the "Measure" panel.

Good idea with regards to the distortion measurements, would that just be the distortion plots you get when you perform a REW measurement? I have those plots (attached) from measurements (at much lower level) I did after the loud pink noise. However, the measurements were done with the speaker far away from mic (>3m), and sitting high on top of a bookshelf. The second plot shows the measurement with some EQ applied via REW, and distortion is increased.

If everything was correct at the second measurement, result is alarming.
Either what @staticV3 said, or some strange filter structure (you don't fill dips, do you? )

Try again closer to the speakers and compare them.

Edit: looking at the response it seems that you DO fill dips.
Better not do it, either the speaker or the amp can't take it (you need way too much power to fill dips) or you have to reduce pre-gain by a great amount.

 

[underthetable]

Here are the response and distortions measurements at 1m, the R measurement has a distortion peak around 484Hz which is not present in the other speaker.

 

[underthetable]

I took some more measurements with a slightly different position in the room (still 1m away), and increased the sweep length. Updated plots attached, now with the legend.

 

[voodooless]

Given the usage of pink noise, the volume level, the shortness of the burst, and the fact that the speaker has extensive countermeasures to prevent over-excursion and over-temperature, I would assume that you’re fine.

 

[Sokel]

I took some more measurements with a slightly different position in the room (still 1m away), and increased the sweep length. Updated plots attached, now with the legend.

They seem fine.
As the highlighted curve (brown) is the noise floor I wouldn't worry about that either for the elevated at the second chart, probably some sound source is near there, PC, fridge or something.

 

[underthetable]

They seem fine.
As the highlighted curve (brown) is the noise floor I wouldn't worry about that either for the elevated at the second chart, probably some sound source is near there, PC, fridge or something.

Cheers. When you say elevated noise in the second chart, do you mean L of v2, or the v2 set (both L and R)?

 

[Sokel]

Cheers. When you say elevated noise in the second chart, do you mean L of v2, or the v2 set (both L and R)?

I mean Lv2 at post #10.
Nothing to worry about.