Thread: Mythbusting

I recently saw a post about speaker failure, and it hinted that distortion or clipping will cause speaker failure, even if considerably under the power rating of the speaker. I do know that some waveforms have higher power content due to higher crest factors and different shapes; some consider clipping bad because it starts to flatten out like a square wave, or even turn to DC (impossible).

My response was something like, "The power rating of a speaker is its thermal capacity. The speaker will not blow if operated within its thermal capacity and mechanical limits, regardless of waveform."

As far as I know, speakers can only be blown two ways: too much heat (average power), or mechanical failure (peak power).

Thoughts?

Let sleeping dogs lie.... :-(

JR

If waveforms destroyed speakers, how did Bob Moog ever hear one of his creations?

One answer is that it depends on what is clipping. Obviously, a distorted guitar isn't going to blow a speaker just due to the distortion. But, if your sub amplifier is clipping, you may well induce a thermal failure of the subwoofer. This is because clipping starts to produce significant amounts of high frequency energy. The sharper the corner that is induced, the more high frequency content that is introduced. The significant detail is that this energy is introduced well after the crossover, so it is not filtered and sent to the proper place. This leads to lots of out-of-band voltage and current being sent to the sub. This energy just turns into heat in the voice coil, since the driver can't reproduce it. It may take a while, but failure will happen if the amplifier is kept in clipping all night.

When the distortion or different waveform is introduced prior to the crossover, for instance as an intended source like a square wave from one of Moog's synthesizers, there is less chance of speaker failure because the crossover has a chance to send the energy where it is supposed to go.

Does that make sense? This is how underpowering a system can lead to failure, especially of the low frequency components of a system.

Originally Posted by Andrew Welker

...produce significant amounts of high frequency energy...lots of out-of-band voltage and current being sent...turns into heat in the voice coil

What you're saying is that a distorted signal may unintentionally send more power to the speaker causing a failure, right?

But if the output power (even if it's more than expected) is still within the power rating of the speaker, nothing bad will happen

Not immediately. The speaker motor structure is still heating up, and is probably getting more heat than it is designed to handle and dissipate. It isn't a situation that I would want to regularly be in, as it will more likely than not lead to failure eventually. Like I said, it all depends on where the distortion/clipping is occurring and how long it occurs.

Originally Posted by Andrew Welker

One answer is that it depends on what is clipping. Obviously, a distorted guitar isn't going to blow a speaker just due to the distortion. But, if your sub amplifier is clipping, you may well induce a thermal failure of the subwoofer. This is because clipping starts to produce significant amounts of high frequency energy. The sharper the corner that is induced, the more high frequency content that is introduced. The significant detail is that this energy is introduced well after the crossover, so it is not filtered and sent to the proper place. This leads to lots of out-of-band voltage and current being sent to the sub. This energy just turns into heat in the voice coil, since the driver can't reproduce it. It may take a while, but failure will happen if the amplifier is kept in clipping all night.

um no.... The increase in average power from turning the amp up past clipping is far more significant, than the HF distortion products which will be a small fraction of that heat power increase.

There are some secondary effects related to wide band clipping in combination with modest power tweeters.

When the distortion or different waveform is introduced prior to the crossover, for instance as an intended source like a square wave from one of Moog's synthesizers, there is less chance of speaker failure because the crossover has a chance to send the energy where it is supposed to go.

Huh? Woofers generally have far more power handling capability that tweeters. If the clipping related energy were steered to the tweeters, failure would be more likely, not less. Heat is heat and woofers can tolerate a lot more than tweeters. Note: If the woofer is inductive at the clip distortion energy frequencies, the current and associated heat impact will be even less.

Does that make sense? This is how underpowering a system can lead to failure, especially of the low frequency components of a system.

No, under powering a system DOES NOT lead to failures. Pushing levels TOO HIGH does.

[edit] Note, pushing the level past clipping is arbitrary and incidental to the insult, it is pushing the average level higher that does the damage. Inexperienced system operators will sometimes turn underpowered systems up higher (average) than fully powered systems, that get louder on transient peaks. A larger amp turned up too loud will eat speakers too, but the operator is less inclined to turn it up that high when it plays louder transiently. Yes, this sounds a little paradoxical. [/edit]

JR

PS: Arghhhhhhhhh This topic (matching amps to speakers) is sufficiently complex to support old myths among those unfamiliar with the actual failure mechanisms.. Beware simple answers for complex questions (except for my simple answers) :-).

Don't we get one of these threads like every 4 months?

Speaker go boom boom. Turn it up, it go boom boom louder. Turn it up more and it go boom boom until no more boom boom is left. As long as you stay within the boom boom operating specs, you'll be fine.



Evan

Originally Posted by Evan Kirkendall

Don't we get one of these threads like every 4 months?

Speaker go boom boom. Turn it up, it go boom boom louder. Turn it up more and it go boom boom until no more boom boom is left. As long as you stay within the boom boom operating specs, you'll be fine.


Evan

+10^1,000,000

John, you are correct, in that under-powering a system in and of itself does not lead to failures, if it is operated correctly. What generally happens though, is that whoever is driving the system doesn't understand this and pushes the level too high, as you mentioned. Wouldn't this lead to clipping in places that could potentially lead to failures?

I think I was wrong in using the sub amp as an example, since you are correct that they will generally handle much more power than a tweeter. However, won't the same issue cause failure of a tweeter? I am learning here too, so I apologize for the simplistic explanations.

Originally Posted by Andrew Welker

John, you are correct, in that under-powering a system in and of itself does not lead to failures, if it is operated correctly. What generally happens though, is that whoever is driving the system doesn't understand this and pushes the level too high, as you mentioned. Wouldn't this lead to clipping in places that could potentially lead to failures?

I think I was wrong in using the sub amp as an example, since you are correct that they will generally handle much more power than a tweeter. However, won't the same issue cause failure of a tweeter? I am learning here too, so I apologize for the simplistic explanations.

As I already said, the clipping is incidental to turning it up, a consequence of the smaller power amp's PS and limited headroom, not the cause of the power increase which it turning up the voltage gain and generating higher "average" power (even with transients clipping).

If you inspect the voltage gain of an underpowered system turned up too far trying to compensate for lack of more BOOM on peaks, and a properly sized (larger amps) system operated normally even while making louder BOOMs on peaks, the smaller system will be turned up to a higher voltage gain and generate more "average" power than the larger system.

This is not our first time discussing this particular misunderstanding.... If this seems repetitious to Evan (the kid) imagine how it seems to some of us. I wrote a column called Audio Mythology back in the 1980s. Old myths never die (dammit), maybe we need copper bullets or a neodymium stake through it's heart :-(

JR

Originally Posted by Andrew Welker

John, you are correct, in that under-powering a system in and of itself does not lead to failures, if it is operated correctly. What generally happens though, is that whoever is driving the system doesn't understand this and pushes the level too high, as you mentioned. Wouldn't this lead to clipping in places that could potentially lead to failures?

I think I was wrong in using the sub amp as an example, since you are correct that they will generally handle much more power than a tweeter. However, won't the same issue cause failure of a tweeter? I am learning here too, so I apologize for the simplistic explanations.

Andrew, forget the whole idea of "under-powering". The concept had at one time a tiny grain of fact, but has been overblown to be so incorrect that it's a worthless way of thinking.

Speakers die in one of two ways: Overexcursion - too much voltage that causes the diaphram to be physically damaged by stretching or banging against something, and OVERPOWERING. Overpowering happens when the voice coil exceeds its thermal capabilities and breaks, melts, or deforms.

It's true that a square wave has more average power than a sine wave with the same maximum amplitude, but as long as the driver in question can handle the heat load of the square wave, no damage will result.

The logical followon to the deplorable concept of "under-powering" causing speaker death is the equally deplorable concept that "a larger amp is less likely to blow a speaker than a smaller one". The basic premise is that the same signal fed into a larger amp may not cause clipping, which would theoretically reduce the average power delivered to the driver. The practical reality is that any moron who allows their system to run into clip with a small amp is clearly looking for more than their system can deliver, and when given a larger amp, will only run that into clipping too - destroying the driver even faster than before.

I thought this article explained things rather well and is good at dispelling the many myths:

Speaker Failure Analysis

Good read. Thanks for that.

Originally Posted by Greg Cameron

I thought this article explained things rather well and is good at dispelling the many myths:

Speaker Failure Analysis

My god, Phil Allison is the biggest *sshole on the internet. It is amazing you can run into him at every corner.

Originally Posted by TJ Cornish

Andrew, forget the whole idea of "under-powering". The concept had at one time a tiny grain of fact, but has been overblown to be so incorrect that it's a worthless way of thinking.

Speakers die in one of two ways: Overexcursion - too much voltage that causes the diaphram to be physically damaged by stretching or banging against something, and OVERPOWERING. Overpowering happens when the voice coil exceeds its thermal capabilities and breaks, melts, or deforms.

It's true that a square wave has more average power than a sine wave with the same maximum amplitude, but as long as the driver in question can handle the heat load of the square wave, no damage will result.

The logical followon to the deplorable concept of "under-powering" causing speaker death is the equally deplorable concept that "a larger amp is less likely to blow a speaker than a smaller one". The basic premise is that the same signal fed into a larger amp may not cause clipping, which would theoretically reduce the average power delivered to the driver. The practical reality is that any moron who allows their system to run into clip with a small amp is clearly looking for more than their system can deliver, and when given a larger amp, will only run that into clipping too - destroying the driver even faster than before.

That basically says it all.

From the Wile E. Coyote school of thought:

Likelihood of driver failure using Small Amp = (small amp headroom + speaker killing distortion) / user experience

Likelihood of driver failure using Large Amp = (higher headroom + speaker killing distortion and then some) / user experience

Fallacy: user will not push larger amp into speaker killing distortion region

Problem: User experience is gained as likelihood of failure increases.

Resolution: User experience is gained as likelihood of failure increases.

Originally Posted by Mark DeArman

My god, Phil Allison is the biggest *sshole on the internet. It is amazing you can run into him at every corner.

I'm curious, what part of Mr Allison's article do you think qualifies him as an "*sshole?

Originally Posted by Chuck Simon

I'm curious, what part of Mr Allison's article do you think qualifies him as an "*sshole?

It's not that article, he really gets around the net social sphere. Mainly, what I would say qualifies him as an *sshole, is his ongoing spew of bile on sci.electronics.design. I'm sure you can find many examples if you search on google. Although, my saying he is an ass doesn't mean he is stupid, he just completely lacks decorum of any kind.

Example link posted below:
https://groups.google.com/forum/?fro...M/YaZeG5BYvDgJ

He is technically accurate in what he said about the PS circuitry. Without knowing (and I don't really want to know) what the standard is for decorum on that forum, he does not appear to suffer fools graciously (like we do). :-)

I would not discount his technical postings because he is not polite if I needed that particular information.

I have run across his web pages often in technical searches and his writing is more often factually correct than wrong from my passing inspection. This is well above average for the WWW that high on noise and low on signal.

JR

Originally Posted by John Roberts

He is technically accurate in what he said about the PS circuitry. Without knowing (and I don't really want to know) what the standard is for decorum on that forum, he does not appear to suffer fools graciously (like we do). :-)

I would not discount his technical postings because he is not polite if I needed that particular information.

I have run across his web pages often in technical searches and his writing is more often factually correct than wrong from my passing inspection. This is well above average for the WWW that high on noise and low on signal.

JR

Totally agree, just don't bother getting into an argument with him.