Speaker damage/degradation from sudden & short relatively high volume?

[Andreas Lilja]

Greetings.

So basically I recently turned on my audio system (Edifier S730) without remembering that Windows' sound was at around 85-90 percent or so (due to a previous video with really poor audio). It was really loud (especially the bass from the woofer), at least compared to what I'm customarily used to. (not really into loud stuff) I quickly turned it down.

System volume knob was at 30 on 50; don't know how far these puppies can theoretically go. Any chance of damage/decreased quality... or I'm being overly paranoid and/or seriously underestimate the manufacturer's stress testing? I haven't noticed any change so far personally.

Thanks.

[HPWebcamAble]

Those are pretty good quality, I imagine they are designed to be fairly loud without being damaged.

 

If you don't notice anything sounding strange, they're fine.

[Tenelia]

Audio engineer here. Unlikely. All speakers can handle short bursts of relatively high sound decibels for perhaps a maximum of 4 seconds. Obviously, the smaller speakers will tolerate proportionately lower levels of high sound bursts.

[Mark77]

Speakers are burned out not because music is played "too loud" on them, but rather, because the amplifier attached to them cannot deliver enough power to the coils without going into saturation.  Sometimes this is called "clipping" because basically the amplifier "clips" at a level implied by the amplifier's MOSFET or BJT transistors' power supply.

 

If you look at the Fourier Transform (ie: frequency spectra) of an amplifier output that is saturated, there is a large DC component.

 

Speaker coils are inherently inductors.  Inductors have zero impedance when frequency = 0, ie: DC current.  So the coils heat up, melt, and bye-bye speaker.

 

This is why it is important to always use a large enough amplifier that will not go into saturation, and to immediately shut down music if there is saturation/distortion observed.

[KaminKevCrew]

1 hour ago, Mark77 said:

Speakers are burned out not because music is played "too loud" on them, but rather, because the amplifier attached to them cannot deliver enough power to the coils without going into saturation.  Sometimes this is called "clipping" because basically the amplifier "clips" at a level implied by the amplifier's MOSFET or BJT transistors' power supply.

 

If you look at the Fourier Transform (ie: frequency spectra) of an amplifier output that is saturated, there is a large DC component.

 

Speaker coils are inherently inductors.  Inductors have zero impedance when frequency = 0, ie: DC current.  So the coils heat up, melt, and bye-bye speaker.

 

This is why it is important to always use a large enough amplifier that will not go into saturation, and to immediately shut down music if there is saturation/distortion observed.

Do you have a source for this being what occurs when clipping happens (@SSL, is this a thing)? I don't think that's actually what happens. Also, nothing has Zero impedance... 

 

Melting generally occurs when you have an amp that is too powerful for the speakers and you turn the amp up too high (source: I've done it.) In my experience, there are kind of two kinds of 'clipping' one is where the loudness war happened, and the recorded audio is at too high of a volume, in which case, you will experience clipping no matter what (and you won't experience any damage to your speakers) and the other is what happens when the speakers have too far of an excursion and can't adequately return to their resting, or as the case may be, opposite position. This is where damage can occur, but again, that's  caused by the speaker being played too loud.

 

@OP, those speakers are powered. Your computer only outputs a certain loudness of signal (generally something around line-level at most) which is very quiet. Powered speakers are designed to take line level sound, and amplify it (actually, all amplifiers are designed to do that). Basically, the amp in those speakers is paired to them for the reason that it cannot damage your speakers, even when turned up all the way. 

 

If I were you, I would actually leave windows volume on 100% all the time, since the OS will actually clip bits when you turn down the system volume, which can result in an audible degredation in sound quality. There's a reason that your speakers have a volume knob. Use it.

[Andreas Lilja]

3 minutes ago, KaminKevCrew said:

@OP, those speakers are powered. Your computer only outputs a certain loudness of signal (generally something around line-level at most) which is very quiet. Powered speakers are designed to take line level sound, and amplify it (actually, all amplifiers are designed to do that). Basically, the amp in those speakers is paired to them for the reason that it cannot damage your speakers, even when turned up all the way. 

 

If I were you, I would actually leave windows volume on 100% all the time, since the OS will actually clip bits when you turn down the system volume, which can result in an audible degredation in sound quality. There's a reason that your speakers have a volume knob. Use it.

Thanks everyone for your inputs!

 

So basically I'm worrying for nothing? Excuse my misunderstanding of the jargon. 

 

Will definitely follow your Windows tip, tho. Plus it prevents the replication of such events.

 

Seems like I'm OK for this time, just received the following by e-mail from the manufacturer:

 

" The speakers will be fine for working in high power in a short period of time." 

Edifier North America Service Team

[Mark77]

19 minutes ago, KaminKevCrew said:

Do you have a source for this being what occurs when clipping happens (@SSL, is this a thing)? I don't think that's actually what happens. Also, nothing has Zero impedance... 

 

Impedance through a coil, Z = j*2*pi*f*L.  For DC, f=0, hence, Z = 0.  There is a small amount of resistance ("R"), but it is negligible compared to the impedance of the coil during normal "AC" operation. 

 

21 minutes ago, KaminKevCrew said:

Melting generally occurs when you have an amp that is too powerful for the speakers and you turn the amp up too high (source: I've done it.)

 

Sure, if there's literally so much voltage in those coils that the insulation is damaged.  But that's not usually the case.  Clipping, where you're trying to drive speakers with not powerful enough of an amplifier, causing the power transistors in the amplifier to saturate, is the most common problem. 

 

23 minutes ago, KaminKevCrew said:

Your computer only outputs a certain loudness of signal (generally something around line-level at most) which is very quiet. Powered speakers are designed to take line level sound, and amplify it (actually, all amplifiers are designed to do that). Basically, the amp in those speakers is paired to them for the reason that it cannot damage your speakers, even when turned up all the way. 

 

 

There are plenty of amplified speaker sets out there which can be driven into saturation, and hence, speaker damage if done on a prolonged basis by hitting them with high magnitude low-frequency signals. Not to mention it sounds like total sh*t when clipping occurs.  But give me a set of amplified "computer" speakers to destroy, and I don't doubt for a moment that I could.

 

 

[KaminKevCrew]

1 hour ago, Mark77 said:

There are plenty of amplified speaker sets out there which can be driven into saturation, and hence, speaker damage if done on a prolonged basis by hitting them with high magnitude low-frequency signals. Not to mention it sounds like total sh*t when clipping occurs.  But give me a set of amplified "computer" speakers to destroy, and I don't doubt for a moment that I could.

 

 

But could you do it with normal music?

[rhyseyness]

6 hours ago, Mark77 said:

 

Impedance through a coil, Z = j*2*pi*f*L.  For DC, f=0, hence, Z = 0.  There is a small amount of resistance ("R"), but it is negligible compared to the impedance of the coil during normal "AC" operation. 

 

Sure, if there's literally so much voltage in those coils that the insulation is damaged.  But that's not usually the case.  Clipping, where you're trying to drive speakers with not powerful enough of an amplifier, causing the power transistors in the amplifier to saturate, is the most common problem. 

There are plenty of amplified speaker sets out there which can be driven into saturation, and hence, speaker damage if done on a prolonged basis by hitting them with high magnitude low-frequency signals. Not to mention it sounds like total sh*t when clipping occurs.  But give me a set of amplified "computer" speakers to destroy, and I don't doubt for a moment that I could.

 

 

 

4 hours ago, KaminKevCrew said:

But could you do it with normal music?

Electronic engineer here, and just wanted to back up what Mark77 said.

DC voltage through a speaker coil will cause massively high current and kill your speaker.

Also, Kamin, you're right about 0 impedance only being theoretical, but I believe Mark already corrected you here.

The resistance of the wire is significantly lower (basically 0) compared to the inductor (coil) reactance under normal operation (higher frequency).

 

The only part I disagree with is where you (Mark) say "if there's literally so much voltage in those coils that the insulation is damaged".

Too much voltage through a coil is less likely to cause damage, especially when we're talking about audio signals, but the current definitely will.

Most insulation doesn't break down until upwards of 500V is dropped across it, which is highly unlikely in home audio applications.

[Dackzy]

6 minutes ago, rhyseyness said:

 

Electronic engineer here, and just wanted to back up what Mark77 said.

DC voltage through a speaker coil will cause massively high current and kill your speaker.

Also, Kamin, you're right about 0 impedance only being theoretical, but I believe Mark already corrected you here.

The resistance of the wire is significantly lower (basically 0) compared to the inductor (coil) reactance under normal operation (higher frequency).

 

The only part I disagree with is where you (Mark) say "if there's literally so much voltage in those coils that the insulation is damaged".

Too much voltage through a coil is less likely to cause damage, especially when we're talking about audio signals, but the current definitely will.

Most insulation doesn't break down until upwards of 500V is dropped across it, which is highly unlikely in home audio applications.

I only know a little about this, but I agree.

[Mark77]

2 hours ago, rhyseyness said:

The only part I disagree with is where you (Mark) say "if there's literally so much voltage in those coils that the insulation is damaged".

 

So what's the reason for an excessively large amplifier killing a speaker, as claimed by the prior poster?  Or is the previous poster mistaken?

 

Are we back to excessive coil heating because the amplifier is able to force so much energy in that things melt?  Do the magnetics in the speaker itself saturate, and eddy currents cause heating and non-linearity? 

 

Just interested (EE/CompE/CS here!)

[rhyseyness]

17 hours ago, Mark77 said:

 

So what's the reason for an excessively large amplifier killing a speaker, as claimed by the prior poster?  Or is the previous poster mistaken?

 

Are we back to excessive coil heating because the amplifier is able to force so much energy in that things melt?  Do the magnetics in the speaker itself saturate, and eddy currents cause heating and non-linearity? 

 

Just interested (EE/CompE/CS here!)

Too much current through the coil causes the cone to oscillate outside of its limits and thus become damaged.

Alternatively, as you describe, too much current/energy through the coils will cause excessive heat and damage to the coils.

I'm not an audio expert but this is how I understand it from an electronics standpoint

[Lutkeveld]

Oh wow, I don't even know where to start here.

It's cool to see some actual technical knowledge on this forum, but a lot of it is applied wrong.

 

Clipping is harmful for speakers because it decreases the crest factor of the output (music) and produces harmonic content, which can stress the tweeter.

 

A speaker is far from an ideal coil. It has a high DCR (2-8ohms). There is no "DC" in a clipped signal, it's high frequency content (-> fourier). 

 

The output often isn't saturated when clipping, it just reaches the max of the available voltage (rail).

 

I'm always open to discussion of course.