Is it damaging to clip an amp?

So long as you keep them cool, they should be OK.

A quibble, though: if the subs cost so little that you don't mind destroying them, then I have to question the quality of the sound they produce.

Hard clipping would be a very bad thing, soft-clipping not so much...
As long as the both stay cool it should be OK.
It's not really a good thing for either of them, and will shorten the lifespan of them both. By how much exactly? Only god knows that.

It can be catastrophic to drivers too.

It is best to avoid it as much as possible, that's why we build in headroom.

Not at all. Guitar amps are always clipped, and never damaged by it. And there's nothing special about them to allow that. But with subs it won't sound good, especially with SS. Tweeters yes, midranges sometimes. Woofers and subs, never.

True. I have, when I was a youth, burned out a tweeter by clipping. Woofers/subs will not be damaged by clipping but can be damaged by over excursion (not clipping) and overheating (thermal overload). I'm certain you know all this.

We should, but the myth of under powering just won't die. It's not referenced often here, on musical instrument and pro-sound forums it pops up with surprising regularity from people who should know better.

I agree with you, but since thermal overloading and over excursion can be a direct result of hard clipping shouldn't it be said that clipping could damage a woofer if one doesn't hear it and know when to back off?

It isn't damaging until it is. It is real simple, avoid it and you won't worry about it.

As long as the driver can thermally take the power, you can clip your amp all day long and it's not going to hurt anything. The DC current is the tweeter killer. I just recently blew 2 x sI 18's with less than 1000 watts power. I was using the dayton 1000 plate amp on both and both went up in smoke. was it clipping? probably, theres no reason the SI should have crapped out in 3CF boxes with 1000 watts. Both coils were smoked. I can't think of any other reason they would crap[ out as both amps still work perfect on my UXL's. Turn them up the same deal and they clip all day long and the UXL's just laugh at them.

I believe it is the high frequencies produced from the clipping, ie the square wave, that burns out tweeters.

It's the DC current the amp produces when it clips that is the killer.

That's one of the myths in question. There is no DC.
http://forum.qscaudio.com/forum/viewtopic.php?f=29&t=2736
High frequency harmonics, to be exact. With a normal signal power density drops by 3dB for each 1 octave rise in frequency, so normally a tweeter might see 5% of the total system power. The added harmonics of a clipped signal can result in 10dB or more power density in the highs than normal, so the tweeter might be called on to handle 50% of the total system power, burning it out. Since the woofer is already expected to handle at least 75% of system power and its voice coil inductance acts as a low pass filter it's unaffected unless the clipping is accompanied by the low frequency output being more than the driver is rated for. Keep in mind that amps are rated typically at 1% or less THD, and when driven into clipping they can deliver much more than their rated power.

Well, I'm happy I wasn't wrong.

I can quote several websites that say the exact opposite. I'm not saying they are 100% correct but I'm only doing wjhat you are Bill, quoting a link just for reference.
http://www.formaudio.com.au/technical-articles/83-technical-clipping-and-under-powering-speakers

http://www.focal.com/en/content/420-why-do-voice-coils-burn-out

Although the voltage can't actually hold the woofer in place, it does attempt to until the input falls within the voltage limits. There is quite a bit of evidence both ways. An amp can certainly produce more wattage that it's rated for for short bursts, which is the result of most failures but thermally, the coil can't cool itself and bingo, we have a cooked coil.

If that's what they say they're wrong. Nope. A woofer never sees a true square wave. It only sees that portion of the square wave that remains after it has passed through the crossover, or if there is no crossover that which remains after being filtered by the inductance of the voice coil. If your links show an oscilloscope trace of a square wave then totally clueless they be, for no speaker ever sees a waveform that even vaguely resembles that.
As far as speakers are concerned there's nothing special at all about a clipped waveform, it's just another complex waveform, and they treat it just like every other complex waveform that they deal with with every pulse they receive.
If clipping bothered woofers in the least then every guitar player would be replacing drivers as often as they replace strings.
Edited by Bill Fitzmaurice - 2/3/13 at 3:02pm

The Yoda saying of the day.

I only put the links up as an example. Simply the same thing you did, just to show different points of view. I'm only trying to show others. There is lots of papers saying it does. It is what it is. This will be my last post on the subject as I'm sure it will just turn into a typical debate we've all seen time and time again lol

http://sound.westhost.com/clipping.htm

I think part of the problem is misused terminology. In this reference: The author says:

" 1. When an amp is clipping, it will output a DC signal. "

This is not only wrong, but is clearly contradicted by the image on his oscilloscope. I think he is describing the flat-top as DC, which it is not. The energy delivered is less than it would have been had the sine-wave been allowed to continue its normal curve rather than flat-topping.

It is the abrupt edges when the signal goes flat that causes the higher-frequency harmonics. As Bill said, those would then be routed to the midrange and tweeter by the crossover, and possibly cause damage there.

That second reference has too many problems to refute.

I'm only pointing out examples of different papers. I'm not saying that they are right or wrong. Just pointing out that there are lots of differing views on the subject. I personally have never blown a tweeter, I've blown up a few subs lately that shouldn't have been nowhere near cooking with the power they were getting but they did. If I'm wrong, I'm fine with that but I've not seen any evidence to prove otherwise. Like what was just said, just linking a paper or some numbers isn't evidence because as we all know, it can be wrong.

Agreed, and note that I stated that this is a persistent myth that won't go away, partly because parties that should know better continue to spread it. Were you monitoring the power? If not you can't know for sure that overpowering didn't occur, and if you cooked drivers then the only logical conclusion is that it did. I've blown exactly one driver in the last 40 years, an EVM15B that I was testing. I was using a 50 watt amp and an errant transient pulse took out the 400w rated voice coil in the blink of an eye.

Yes, I see your point.

A poster needs to do more than just link to what may be a dubious source. This is one of the reasons I read AVS, as there are many people like Bill who are speaking from in depth experience. If someone here misstates the facts, it will always be corrected. I've learned a lot in this forum, in particular.

They should not have blown, they should have been able to handle way more power. I don't know why they fried but both subs cooked at the same time on "scary christmas" dub step song I play to make sure they have no problems before I ship them away. They were SI 18's in 3 CF boxes. They might have been getting 650 watts, if that. It just surprised me I guess. I've been putting quadruple power to subs for 10 years now and never had one smoke a coil. I've had damage from over excursion but never thermally. 2 seperate dayton 1000 plates amps. Crazy stuff. I can get a VC made no problem and I've got a few re-cone kits laying around that should be ok.

People who write, or perhaps proof, such papers should be required to take at least one basic circuits class.

I love debating "marketing" papers on the Internet, proves just how useless three degrees and decades of experience can be. Helps keep me humble. Or maybe not, just annoyed and befuddled at how readily physics can be redefined...

Life would be easier if pi = 3 instead of 3.141592653..., but...

There are debates on opinions that cannot be resolved. This is math and physics, not opinion.

The author of the info at the westhost link offers the relevantly titled article "Why Do Tweeters Blow When Amplifiers Distort?" The introduction of which laments this is: The rest of the article then provides measurements that show the added harmonics from clipping do in fact increase the signals in the tweeter. Just that the ponderous explanation misses the point. So the author is correct: the question is rarely answered properly.

The article also makes this statement about what happens when a sine wave just below clipping is pushed 3 dB into clipping: Except it doesn't. The power increases even though the peak voltage of both waveforms is limited by the power supply. As a sine wave becomes a square wave, the power ultimately doubles. Where does all that added power exist spectrally? At every harmonic above the fundamental -- many of which heat the tweeter.
Edited by Roger Dressler - 2/4/13 at 9:10am

"As a sine wave becomes a square wave, the power ultimately doubles. Where does all that power exist spectrally? At every harmonic above the fundamental -- many of which heat the tweeter."

i thought the average of a sine wave was 0.707 of peak, therefor at full square wave where the average increases to ~1.000, the increase in power would be 1.000/0.707 which is around 1.414 or a 41% increase in power.