Audio & Video Basics: Power Amplifiers

[Andrew Robinson]

[Dark Ranger]

I'm going to get royally flamed for saying this, but here goes.


The advice in the video states that having too much amplifier power is not what kills a speaker, but it's actually damaged by having too little power. Specifically, loudspeaker damage occurs from over-driving an amplifier into clipping and sending a clipped signal (and sometimes a square wave) into the loudspeaker.

I disagree completely. Loudspeakers can be damaged in one of two ways: mechanical stress and/or thermal overload. The first, mechanical stress, is common for woofers and occurs when the woofer system is driven beyond its excursion limits. Damage occurs due to brutal force, which only occurs with too much power. The second method of damage, thermal overload, is common for tweeters and occurs when the tweeter system is delivered more power than it can effectively dissipate. The delicate components simply overheat and burn up.

In both cases, the too-much-power overload I'm talking about here refers to average power. All loudspeakers can handle more short-term power than they can long-term (often referred to as continuous power). Sending 400 watts to a loudspeaker rated for 200 watts continuous is generally OK provided it's not for a long time. However, if the amplifier sends an average/continuous power of 400 watts to a loudspeaker rated for 200, then damage will likely occur.

Just to clarify, I completely agree that a clipping amplifier is not a good thing. However, as long as the average/continuous power remains below the loudspeaker's average/continuous rating, the loudspeaker can play square waves all day with no damage. Sure, it'll sound absolutely horrible, but the speaker will still work fine.

I just don't feel comfortable with the recommendation of getting monster power amplifiers to avoid loudspeaker damage. The point of a high-powered amplifier should be to recreate transients with accuracy and a non-clipped signal. If too little power actually damages speakers, why on earth do manufacturers sell AVRs and amplifiers with "low" power? They should just release 500+ wpc monstrosities and cull the entry-level stuff.

[DYohn]

Dark Ranger, you are 100% correct. Bravo. As an engineer, this is one of the audio myth-understandings that grates on me more than almost any other.

I will reiterate: IF a voice coil can handle the thermal energy, you can run a 100% clipped signal into it ALL DAY LONG without any ill effects other than bad sound. If the power exceeds it's capabilities to dissipate, clipped or not, then out comes the magic smoke. It has nothing to do with the size of the amplifier: it has to do with the thermal capabilities of the VC mechanism. It is HIGH power that fries speakers, not low power.

[yves]

Dec 9, 2013 16:43:31 GMT -5 Dark Ranger said:

why on earth do manufacturers sell AVRs and amplifiers with "low" power?

[biggerwoofer]

yepper, the AUC is only gonna grow when the top is not sliced off.

[mgbpuff]

Dark Ranger and DYohn - It's not as simple as that if the speaker is a multi transducer unit. The tweeter, being lightly built, does not have the power handling that the mid or woofer does. When a given frequency signal clips - the watts delivered at the fundamental square wave frequency is 62% over the sinusoidal power at the same frequency. This is bad, but, and this is the important thing to consider, the clipping causes harmonic components (3 x fundamental) and (5 x fundamental) and (7 x fundamental) that may now be above the crossover cutoff design causing much more power to go to the much more delicate tweeter. You two can bet all you want on trying to get by on less amplifier power, but the safe bet is having more than enough amplifier power. You will not turn the volume on a powerful amp up more than your normal listening level because it will be uncomfortable. Any clipping sounds bad and it will take out your tweeter easier than you would think. Go with more power - not just barely adequate power!

Here is a nice article on the subject - www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/clipping/page1.html

[DYohn]

Dec 9, 2013 20:40:13 GMT -5 mgbpuff said:

Dark Ranger and DYohn - It's not as simple as that if the speaker is a multi transducer unit. The tweeter, being lightly built, does not have the power handling that the mid or woofer does. When a given frequency signal clips - the watts delivered at the fundamental square wave frequency is 62% over the sinusoidal power at the same frequency. This is bad, but, and this is the important thing to consider, the clipping causes harmonic components (3 x fundamental) and (5 x fundamental) and (7 x fundamental) that may now be above the crossover cutoff design causing much more power to go to the much more delicate tweeter. You two can bet all you want on trying to get by on less amplifier power, but the safe bet is having more than enough amplifier power. You will not turn the volume on a powerful amp up more than your normal listening level because it will be uncomfortable. Any clipping sounds bad and it will take out your tweeter easier than you would think. Go with more power - not just barely adequate power!

Here is a nice article on the subject - www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/clipping/page1.html

Actually all that is right (and I understand it well) but it remains "as simple as that." If the transducer can handle the power, then it simply does not matter. It is TOO MUCH power, regardless of how it is delivered, that fries voice coils (or overdrives a suspension.) Period.

[mgbpuff]

Dec 9, 2013 20:53:12 GMT -5 DYohn said:

Dec 9, 2013 20:40:13 GMT -5 mgbpuff said:

Dark Ranger and DYohn - It's not as simple as that if the speaker is a multi transducer unit. The tweeter, being lightly built, does not have the power handling that the mid or woofer does. When a given frequency signal clips - the watts delivered at the fundamental square wave frequency is 62% over the sinusoidal power at the same frequency. This is bad, but, and this is the important thing to consider, the clipping causes harmonic components (3 x fundamental) and (5 x fundamental) and (7 x fundamental) that may now be above the crossover cutoff design causing much more power to go to the much more delicate tweeter. You two can bet all you want on trying to get by on less amplifier power, but the safe bet is having more than enough amplifier power. You will not turn the volume on a powerful amp up more than your normal listening level because it will be uncomfortable. Any clipping sounds bad and it will take out your tweeter easier than you would think. Go with more power - not just barely adequate power!

Here is a nice article on the subject - www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/clipping/page1.html

Actually all that is right (and I understand it well) but it remains "as simple as that." If the transducer can handle the power, then it simply does not matter. It is TOO MUCH power, regardless of how it is delivered, that fries voice coils (or overdrives a suspension.) Period.[/quote

O.K. I agree, BUT, a non clipping amp signal will give you more safe, louder, and less distorted music than a clipping amp signal and IT IS THAT SIMPLE! PERIOD.

[DYohn]

Dec 9, 2013 21:03:34 GMT -5 mgbpuff said:

O.K. I agree, BUT, a non clipping amp signal will give you more safe, louder, and less distorted music than a clipping amp signal and IT IS THAT SIMPLE! PERIOD.

Oh, of course. But the topic here is not what sounds good but what will damage a loudspeaker voice coil.

[mgbpuff]

Just like guns, amplifiers don’t kill speakers,
people kill speakers

[Darksky]

I believe you guys are saying the same thing. Heat destroys the coil, that is clear. Heat is heat, whether it comes from distorted signal or not. I think where the rub here is: Which more commonly causes over heating? A low powered amp driven to clip or a high powered amp delivering too much power. I am going to guess, under powered, only because I believe it happens more commonly at the hands of unknowing users.


I would like to think that with quality speakers, the pressure created by extremely loud listening levels would become unpleasant before damage would occur when using high powered amplifiers.

Well done, Andrew. If you follow up with another video. In this series, could you touch on how ohm ratings are dynamic and how the amplifier responds to these changes?

[deltadube]

Dec 9, 2013 21:50:48 GMT -5 mgbpuff said:

Just like guns, amplifiers don’t kill speakers,
people kill speakers

mostly with under powered avr ... 9 out of 10 speakers I bet..

paradigm recommends to over amp your speakers right in the speakers owners manual..

cheers..

[boomzilla]

Dec 9, 2013 21:50:48 GMT -5 mgbpuff said:

Just like guns, amplifiers don’t kill speakers, people kill speakers

Yep - Crank it to "ll," and you can fry just about any speaker, regardless of amplifier power rating. Mr. Yohn is undoubtedly correct - it is power in excess of the driver capabilities that fries speakers whether from a clipping AVR or a cranked up XPA-1 pair. If you feed the speaker more power than the motor can handle, it fries. I watched a salesman once use a signal generator set to 20KHz on a large Advent speaker - since he couldn't hear it, he cranked the amplifier up. Smoke literally came from the tweeter (which I promptly pointed out to the salesman, who turned as white as a ghost).

The bottom line is: Don't clip your amp or overdrive the speaker. Either is likely to cause damage UNLESS you're using heavy-duty "pro audio" drivers that will tolerate the power you're sending.

[mgbpuff]

"Mr. Yohn is undoubtedly correct - it is power in excess of the driver capabilities that fries speakers whether from a clipping AVR or a cranked up XPA-1 pair. If you feed the speaker more power than the motor can handle, it fries."

Of course - that's like saying buildings fall down because of gravity - no more discussion needed - the expert has spoken!

[boomzilla]

I don't claim to be the "expert" here at all. It is true that mechanical damage can "blow" speakers, but it is also true that clipping is a more likely cause of tweeter damage. More speakers seem to blow their tweeters than their woofers (except in live sound use), so it seems logical to me that clipping is the likely culprit. You're the electrical engineer, mgbpuff - and I'm not trying to be a smart one when I ask, but your statement "a non clipping amp signal will give you more safe, louder, and less distorted music than a clipping amp signal" is absolutely true so far as it goes, but it it doesn't really address the "more power than the driver was designed for" scenario. Comment?

[mgbpuff]

I posted a rather rigorous discussion about clipping above ( www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/clipping/page1.html). Short of a speaker built like a battleship or a very expensive electronic speaker protection circuit that no doubt would deteriorate SQ, I do not know how to easily solve a problem that common sense and a little knowledge can avoid.

[boomzilla]

Agreed, and good article, by the way!

[KeithL]

You must remember that, while this is true for high frequencies, it is not true at all for very low bass frequencies.

Our ears are relatively insensitive to very low frequencies, generating loud low frequency content requires that we move a lot of air, and many speakers aren't very efficient at generating very low frequencies. (Unless you have a very powerful subwoofer, playing pipe organ music loud enough to be unpleasant would require a lot of power, and could quite easily blow even the heaviest duty speaker before it got excessively loud.) Also, most speakers dissipate heat "mechanically" - heat is conducted to the magnet structure or carried away by air flowing around the voice coil. Because of the way this works, most speakers can withstand a lot of power for short periods of time, but may heat up gradually to the point of damage, and eventually fail, if subjected to a sustained overload; this will vary widely for different speakers. This is one reason why home theater systems usually have "specialist" subwoofers - designed for just that sort of thing.

I would also disagree strongly with one statement you made; do NOT assume that "quality speakers" will be able to produce very loud listening levels before damage will occur. Speakers are designed with different purposes in mind, and there are many high quality speakers that are simply unable to produce loud listening levels. While it is true that most speakers (including most of ours) can play quite loudly before being at risk, this is another instance where common sense is important - and knowing a bit about your particular speakers.

The truism about tweeters most often being destroyed by low-powered amplifiers comes mostly from the simple fact that inexperienced users (the kind with poor judgment) tend to NOT consider most of these common sense issues.... and to simply "turn it up until it's as loud as they want it" (often with no notice of how distorted it is) - and doing so is most likely to end up with a heavily clipped output if the amp is underpowered.  

Dec 9, 2013 22:09:09 GMT -5 Darksky said:

I believe you guys are saying the same thing. Heat destroys the coil, that is clear. Heat is heat, whether it comes from distorted signal or not. I think where the rub here is: Which more commonly causes over heating? A low powered amp driven to clip or a high powered amp delivering too much power. I am going to guess, under powered, only because I believe it happens more commonly at the hands of unknowing users.


I would like to think that with quality speakers, the pressure created by extremely loud listening levels would become unpleasant before damage would occur when using high powered amplifiers.

Well done, Andrew. If you follow up with another video. In this series, could you touch on how ohm ratings are dynamic and how the amplifier responds to these changes?