I have seen speakers blown up being under powered and speakers fried being over powered.
I wish you guys would stop digging this hole - this is beyond tedious.
Form 1 of the "under-powering" myth: Running an amp into extreme clipping increases HF content of the signal, contributing to HF driver damage.
This one is half-true - a square wave does have high-order harmonics that a sine wave doesn't. HOWEVER: These harmonics are at much lower energy than the fundamental, and beyond the 3rd harmonic (which is likely still at least partially being reproduced by the LF driver) really don't represent that much power. See this link for a graph:
http://www.informit.com/articles/article.aspx?p=1374896&seqNum=7 The more likely reason for tweeter failure is that as you continue turning the volume up, you start increasing the amount of legitimate signal in the HF band, which can exceed the relatively modest power capacity of the tweeter. A 100w woofer may be paired with a 5w tweeter - something that a small amp - say 50w - can easily OVER-power. With extreme HF content in the source, it's very possible to blow the tweeter even without clipping the "too small" amp. In any case, a larger amp makes this problem WORSE, not better.
Form 2 of the "under-powering" myth: The shape of the waveform can cause speaker damage, square waves are bad for speakers, a too-small amp can blow LF drivers, however you want to phrase that.
This one isn't even half true, as there isn't some counter-intuitive factor buried here that confuses the situation.
The shape of the waveform doesn't matter to the driver. A common misconception is that the driver follows the shape of the waveform - if the signal is a square wave, then the driver will magically move infinitely fast from in to out, causing damage from "slamming". This is not the case. The driver and coil assembly have mass, and therefore have inertia to overcome, and can't move infinitely fast. The waveform acts like a gas pedal in each direction - send a positive signal and the driver starts receiving a force that begins to change the direction and/or speed of the driver out. Send a negative signal and the driver starts building energy to go in. The greater the amplitude, the faster the driver movement. The steeper the slope of the line, the faster the direction change is begun, subject to the driver's ability to follow the signal.
The mass of the driver assembly directly relates to the frequency range the driver can reproduce. If the frequency of the signal is high, a heavy driver never finishes moving before getting orders to go the other way, and therefore little movement and little sound is produced. This is the natural low-pass behavior of the driver. The HF energy that hits the woofer, yet is too fast to translate into movement, is lost to heat; however that HF signal is low energy compared to the LF energy (see chart in above link), and isn't very significant. Additionally, the impedance of drivers rises as frequency increases, so the driver accepts progressively less of the HF energy offered as frequency increases.
As an amp starts clipping, the power does continue to increase; a hard-clipped amplifier producing a square wave produces up to (depending on the capabilities of the amp) twice the power - 3dB more - than a sine wave of equal amplitude - the amp operating right up to clipping. As most of this extra energy is dominated in the 3rd harmonic (3X the fundamental frequency), this new energy is likely still in the operating range of the LF driver, and is reproduced (and sounds like crap), and contributes to heat and excursion of the driver.
So, therefore then under-powering is real, then right? Not really. If we use a "correctly-powered" amp that's 3dB larger than the "under-powered" amp running to just before clipping, we're still putting 3dB more signal into the woofer; however the energy now is in the form of the intentional frequency rather than harmonic distortion.
Whether either this clipped signal or the unclipped louder signal from the larger amp will damage the driver depends totally on the capabilities of the driver. A 100w amp run into full square wave into a driver rated for 3200 watts will run forever. A 3200 watt amp run at -10dB into a 100 watt driver will let the smoke out.
At the end of the day, this is all that matters about powering speakers:
- For best performance, an amp that equals or slightly exceeds the program rating of your speakers is a good idea. This is about utilizing the full capabilities of your speakers, not about speaker protection.
- Depending on the relative capabilities of your speakers and your amps, extreme clipping may contribute to driver damage, or it may not. Either way, clipping is bad. Please don't do this - it makes your system sound like crap.
- If a person is dumb enough to ignore all warning signs and fry their system, giving them larger amps is like giving your nephew who totaled his Civic a Porsche. The only difference is death will come sooner and/or more spectacularly than before.
- Bring enough rig for the gig.