Quote:
I'm confused about speaker ratings and amplifier ratings and how they interact.

Some pages say your speakers should be rated higher than your amp. Other pages say that your amp should be rated higher than your speakers. Both of these cannot be true.



When in doubt I'd with the Basic car audio electronics site. Haven't seen anything there that goes against my education and experiences while here's a lot of BS and voodoo spouted out there.

Quote:
Let's start with why we're concerned about this at all. I think we all understand that a higher rated amp is going to be able to produce a louder sound than a lower rated one (assuming their ratings are equivalently determined).


Yes

Quote:
And a higher-rated speaker probably performs the same way. So in the best of all worlds, we'd all get amps and speakers with massively high ratings and be done with it. But economics kicks in.


Not the power ratings - they just state how much power the speaker can handle. How loud it'll be depends on the sensitivity of the speaker, expressed as a dB rating per watt and meter (dB measured at 1 meter distance with 1 watt power applied).
That and how much power the speaker he how handle will give how loud it can be. THeoretically every doubling of power will give you +3dB. For a perceived doubling in volume, a factor 10 in power is required.

Quote:
So, what causes damage?

The obvious one is clipped waveforms, which have much more power than unclipped waveforms. What causes clipping? Assuming I don't go over 0dB on the empeg, the source isn't clipping, so it must happen somewhere in the amp or speakers. But it's unlikely that your amp (unless it's really cheap) is causing clipping on its own, I think, so it must be some interaction of the speakers and amp (since the speaker itself is obviously not able to cause it).


Background: The amp has an internal power supply that pumps up the voltage above 12V in order to be able to give more power. Power is proportional to voltage squared and inverse proportional to resitance/impedance P=U^2/R.

I see clipping happening from 2 causes:

: Asking the amp to deliver a voltage on the output higher than the supply voltage. This happens from the combination of input signal voltage and gain setting. While normally labelled in sensitivity (ie how much input voltage is required for max on the output) it really sets the multiplication factor between input and output voltage.
When input*gain > max possiblevoutput, the utput will simply remain at the max possible...

Second is asking the amp to deliver more current han the supply can give. Typically running a too low impedance load. Tooo much current drawn will likely cause the voltage from the suppply to also drop - then see above.
Quote:

Some would argue that an underpowered amp causes this, as the high-powered speakers try to draw more power than the amp could handle, like hooking up a bunch of space heaters to one power outlet in your home. This could be a bad analogy because the amp varies voltage while your house does not (not to mention the whole AC/DC thing).


No, a higher powered speaker, of the same impedance (typical 4 Ohm), will NOT draw any more current than a lower power rated one. urrent is solely determined by amp output voltage and the speaker impedance I=U/R. Voltage times current gives power P=U*I. Combine those and yuo see that P=U^2 / R (or P=I^2 * R) A higher rated speaker can just withstand/dissipate more power. It's just like resistors - despite being the same Ohm value, depending on construction and size they can handle different amounts of power.

Hooking up to many/too large heaters in your house will result in a a current draw larger than your fuses (because the total resistance of the heater(s) is too low) and/or house wiring can handle. Hooking up a higher power rated speaker (with the same impedance) will make no difference at all. Hooking up multiple in parallell will lower the total impedance the amp 'sees' and the current will increase, posibly above what the internal power supply can handle (clipping) or what the output transistors can handle (going up in smoke)

Quote:
Others would argue that an overpowered amp causes it, like a power surge hitting your delicate computers.



An underpowered amp is likelier to run into the limitations above, if you ask too much from the amp, in a desire for more volume...
ncw
Quote:
Maybe both of those are true. One page argues that if your speakers are rated way higher than your amp, then even if the signal is clipped that it won't be able to produce enough power to actually damage them.


I'd agree. The difference in power between a sine wave and a square vawe of the same amplitude (top output voltage) is a factor 2 (Calculate the areas (integrate) of a halfperiod and you'll see a factor of sqrt(2), and since power is a function of voltage squared, we get sqrt(2)^2=2)
If the speaker can handle way more than that, I see no cause for alarm - except for it sounding like crap due to the speaker playing a square vawe signal..

Quote:
I suppose now is where we get into quality, as that begs the question that even if the speakers won't be damaged, they won't sound good, and is it the high power rating of the speaker contributing to the clipping?


And the fuzzy marketing factor in ratings of car audio stuff, either for the imponator effect "Look at my 1600W amp" (a right bargain at only 89.95 on eBay...) or brand reputation - Acme speakers can handle more than rated / Acme amps delivers more than rated - must be good stuff...

But the power rating does NOT contribute to clipping. A too low impedance could.

Quote:
What other causes are there for speaker damage? Overdriving, causing mechanical damage to the coil, surround or cones? What causes this? It would seem that that would be caused by the speaker trying to generate sound louder than it is able to. That would seem to be caused by an overpowered amp. (Obviously the amp would have to be "told" to produce that much volume by turning up the volume on the source.)


Yes, to high a voltage will mean the speakers tries to move further than the designer anticipated. Too much and something will break. Keping within the outright breaking things limit, stuff can get too hot over a period of time, typically causing glue to fail, plastics to weaken/deform/melt, insulating laquer on the coil wire to burn off, causing shorts between turns of the coil -> lower impedance -> more current -> more heat...

Quote:
Here's another question: what causes a change in volume? Is it voltage? Variation in voltage is what describes the waveform, right?


Yes. and variation in voltage is what eventually (by causing a current to flow, causing a magnetic field in the coil) causes the coil/membrane to move. More chaange in voltage gives larger movement of the coil/embrane, meaning more air moved, meaning your eardrum moves more, ie louder.

Quote:
And speakers are designated to be of specific impedances.


Yes. Car speakers typically 4 Ohm, though subs vary more, for higher power ability and flexibility in wiring multiple subs.

Quote:
Is there a set voltage maximum for audio systems or is it that 4V specifies the max deflection of the speaker cone in one speaker, but only halfway in another?


A given voltage will cause different deflection in different speakers. Somewhat described by the sensitivity rating of the speaker (a dB number measured at 1 meter with 1 watt power to the speaker. Power and voltage relates according to P=U^2 / R.
Then again different speakers have differing max deflections.
The space in a HU is rather limited, giving litle/no room for a power supply to crank up tthe internal voltage, limiting most HUs to bout the same real power, despite various claims.

Quote:
Or is current a variable, too?


For a given speaker, current varies according to amp output voltage and speaker impedance.
Power is ultimately given by voltage times current, but given voltage and impedance we need not explicitly calculate/measure current as it's a given from voltage and impedance.
A lower impedance speaker will tickle more power from an amp (within limits). An amp might be rated at 75W @ 4 Ohm and 150W @ 2 Ohm. Theoretically running a 1 Ohm load would give 300W, but we're likely to run into limits of the power supply and/or the output transistors.

Quote:
But the waveform voltage doesn't specify volume directly. It's the degree of variation in the waveform that's volume. That is a sinus waveform that travels between 3V and 4V wouldn't be very loud, but one that travels between -4V and 4V would be, despite that its additive value would be zero.


Yes, the variation in voltage drives the motion of the coil/membrane. The variation is called amplitude, in your example 1V and 8V (peak to peak) the average from 0 is the DC offset, 3.5V and 0.

DC offset moves the 'resting point' of the coil/membrane but and causes heat to be developed in the coil but, doesn't contribute to volume.

Quote:
Of course, that -4V to 4V ssinus wave at 200Hz isn't going to seem as loud as a -4V to 4V sinus wave at 5000Hz.


Because the ear is less sensitive to lower frequencies. The coil/membrane still moves just as much (possibly ignoring any limiting effects from mass/inertia).

Quote:
I see now that audio signals are actually AC, aren't they?
Yes

Quote:
And you get into power factors with all of this, and not even linear ones, which makes all of this numbering even more complicated than I thought.


For most calculations power factor /phase angle can be ignored...

Quote:
Okay, now I'm in too deep, and I just want someone to tell me if I need speakers more powerful than my amp, or speakers less powerful, or that they should be the same, or that I should just go stick my head in a Gramophone just to keep my brain from turning to goo with all this confusing information.


I'd say go with roughly the same ratings and be sensible with the gain setting, trust your ears.

'm using a 6x75W amp; front system is rated to 50W, rear speakers to 80W and the sub 250W (runs of 2 channels bridged, which should give 300W)

Quote:
And please show your work. There's more than enough conjecture without any basis. Hell, there's a lot of inconclusiveness pretending to be an answer. To be honest, I don't really care that much about the math, but there are obviously a lot of people out there just repeating what they've heard, so I need some basis for a conclusion now. (That is, if everyone said "your speakers need to be rated at 1.414 times your amps rating", I wouldn't really care about any basis, but as it is, that 1.414 can be replaced by anything from one-tenth to ten and it seems that someone will agree with it.)


I'd be happy to go deeper, but it's a lot easier in person/on paper than using a keyboard.
If there's anything specific to something I've said, ask and I'll try try to expand.
_________________________
/Michael