Home Theater Forum and Systems banner

1 - 14 of 14 Posts

·
Registered
Joined
·
185 Posts
Discussion Starter #1

·
Elite Shackster
Joined
·
1,506 Posts
Re: Amp wattage greater than speaker rating

Myth. Clipping is potentially injurious to tweeters and midranges, never to woofers or subs. Explained here:
http://forum.carstereos.org/clipping-test-results-t47441.html?
And here:
http://www.gmtruckhq.com/sound-security/clipping-killed-my-speakers-did-959.html
Interesting articles. I didn't exactly understand what the first article was saying about normalized power vs square wave power. Aside from that I'm not sure I agree with everything stated in them. I'd also like to know how the 2nd author came to his conclusions regarding cooling of a woofer in a clipped signal and I'd like to look a little further into his claims about the sums of the fundamentals being the reason the woofers don't stop.

I'd like to see the tests from the first article repeated using high xmax woofers and large power amplifiers like most of us are using.
 

·
Registered
Joined
·
185 Posts
Discussion Starter #3
Re: Amp wattage greater than speaker rating

Interesting articles. I didn't exactly understand what the first article was saying about normalized power vs square wave power. Aside from that I'm not sure I agree with everything stated in them.
Another interesting document is this one:
http://www.jblpro.com/pub/technote/lowpower.pdf

I've seen literally dozens of sources quote this article in justification of the underpowering myth, every one of them somehow oblivious to the fact that it specifically deals with the potential for damaging high frequency components with clipped waveforms, as clipped waveforms have higher than normal power density at high frequencies.
While the myth has its proponents in the audio field, all of whom should know better, none of them are transducer engineers. I've even seen customer service reps from driver manufacturers repeat the myth, but that's why they're in customer service and not in R&D. My friends in the R&D departments at various driver manufacturers find it all rather amusing.
 

·
Elite Shackster
Joined
·
1,506 Posts
Re: Amp wattage greater than speaker rating

Ya the high frequency harmonics are the primary danger in a clipped signal. In fact when the second author from your first post talks about this indirectly when he speaks of harmonics riding the top of what appears to be a square wave.
In a subwoofer system it is not these signals are not harmful to the woofer. As the author notes the harmonics as a source of movement of the woofer, he fails to mention that the sum of the harmonics look just like the waveform, which is the motion the woofer follows. He also fails to remember that these harmonics are riding at the top of a full power wave. I agree that you could argue the square wave as simply having double the power and that the woofer would remain unruined if it were rated to handle that thermal load. But this neglets that the woofer essentially stops at these peaks and we have no idea what kind of cooling is taking place. The woofer is doing no mechanical work and all energy is lost as heat.

Under these coniditions it makes sense to me that a clipped signal into a woofer would be more damaging than a non clipped signal of the same average power. Maybe the effects are just really insignificant? :huh:
 

·
Registered
Joined
·
185 Posts
Discussion Starter #5
Re: Amp wattage greater than speaker rating

Ya the high frequency harmonics are the primary danger in a clipped signal. :
Harmonics are all that differentiate a clipped signal from one that isn't. A very effective distortion detector is a simple piezo tweeter. Run a low frequency tone through an amp hooked up to a piezo. If the amp is severely clipping you'll hear a tone in the piezo. That tone is harmonics that aren't there if the amp is running clean.
But this neglets that the woofer essentially stops at these peaks.
That never happens. Part two of the above scenario is that when a woofer is fed a severely clipped signal those harmonics are rolled off by the rising impedance of the woofer voice coil.

The proponents of the myth assume that the signal a driver 'sees' looks like what appears on an oscilloscope trace of a square wave, and that that voice coil/cone move in the same fashion as that trace. It isn't, and they don't.
 

·
Elite Shackster
Joined
·
1,506 Posts
Re: Amp wattage greater than speaker rating

Ok, so I'm really trying to understand what happens to a woofer when a clipped signal is sent to it. I've attempted to model what a clipped waveform would look like in excel. For these charts I assumed the clipping level to be .5v, and sent the amplifier a 1v signal at 20hz, thus resulting in a full clipping senario. I calculated harmonics at 40,80,160 and 320hz, with each half as strong as the previous. This is where I'm unsure if I've modeled correctly. Is it safe to assume that 40hz is half as strong as 20 hz and so on? Here are the graphs.

This is a graph of the fundamental plotted with the harmonics for demonstration.



This graph is the summation of the harmonics only, no 20hz signal.



This graph shows the clipped fundamental (blue) with the summated harmonics added to it only above and below .5v.



And last, is the same graph with any output above and below .5 cut out as the amp has no capability there. This is what I would assume to be the final signal and path of the subwoofer.



Would this be a good model of the motion of a clipped 20hz signal as the subwoofer would see it?
 

·
Registered
Joined
·
185 Posts
Discussion Starter #8
Re: Amp wattage greater than speaker rating

Would this be a good model of the motion of a clipped 20hz signal as the subwoofer would see it?
Not really. First off, the only time any woofer ever sees either a sine or square wave is when the source is a tone generator. Music program bears little resemblance to either, though if you do look at a music signal with its preponderance of harmonic content it looks a lot more like a square wave than it does a sine.
What matters is power density, which with a normal music signal drops by 3dB per octave, just like pink noise. Above 5kHz, for instance, on average only 5% of the amp's power normally goes to a tweeter. With a clipped signal power density increases in the harmonics. Assuming a 100 watt amp a pure square wave will have the same power density at 100 Hz as it does at 20 Hz, but that's still only 100 watts and it won't bother a 100 watt woofer. OTOH a moderately clipped signal at 5kHz might double the power density to 10 watts in the tweeter bandwidth, and that would kill a 5 watt tweeter.
 

·
Elite Shackster
Joined
·
1,506 Posts
I totally agree with you on increasing power density for higher frequencies that would end up in a tweeter thus frying it.

Applying to the subwoofer only, in the case of a pure 20hz sin wave, would those graphs closely approximate the signal and movement of a subwoofer? If so, it appears that during clipping, the power sent to the woofer is actually decreased the worse the signal is clipped, and that the woofer does in fact (inertia omitted) stand still for fractions of seconds. These are interesting results
 

·
Registered
Joined
·
185 Posts
Discussion Starter #10
Applying to the subwoofer only, in the case of a pure 20hz sin wave, would those graphs closely approximate the signal and movement of a subwoofer?
No.
it appears that during clipping, the power sent to the woofer is actually decreased the worse the signal is clipped, and that the woofer does in fact (inertia omitted) stand still for fractions of seconds.
It doesn't. If you want to see what the driver 'sees' get a generator that can produce both sine and square waves and do an RTA of both.
 

·
Registered
Joined
·
185 Posts
Discussion Starter #12
I'm sorry Bill, I'm trying to understand this. Would you mind elaborating?
I did, in the second paragraph. Waveforms viewed on an oscilloscope, or a chart, don't reflect what any driver sees as an input signal. RTAs are far more accurate in that respect.
 

·
Elite Shackster
Joined
·
1,506 Posts
Hmm, I guess that's what doesn't make sense then. My understanding is that an RTA takes samples of given frequencies in a signal and displays their amplitude via mathematical transform of the input signal. I guess it doesn't make sense that in an analog system such as a loudspeaker, that there would be any transformation taking place. I guess I'll have to look into this further.

I can see that it would be easier to interpret an RTA analysis for design purposes, but I am trying to understand the mechanical motion of the loudspeaker and relate it to power input and cooling to find out just how much of this myth is myth, and how much may be true. I suspect the answer has components of both.
 

·
Elite Shackster
Joined
·
1,506 Posts
Here's another good article from Rane pertaining to blowing tweeters with a clipped amplified signal. Interestingly enough, it is not the harmonics induced by clipping which actually blow the tweeters.

The clipping, of course, produces harmonics but not nearly as much as the square waves discussed earlier. The amplitude of the high frequencies went up by 3dB in relation to the low frequency fundamental. (3dB compression).

If you overdrive the amplifier by 10dB the high frequency amplitude goes up by 10dB. This goes on dB for dB as you turn up the volume, until the high frequency reaches the 100 watt level. Meanwhile the peak level of the low frequency portion can not increase above 100 watts. This now represents nearly 100% compression (no difference between HF amplitude and LF amplitude).

Now it is easy to see how the high frequency portion exceeds the 5 or 10 watts tweeter rating. Sure, clipping is producing extra harmonics but they never approach the levels of the amplified high frequency source signals.
 
1 - 14 of 14 Posts
Top