Calculating power handling capabilites...

[CADman_ks]

How does one go about figuring out what the power handling capailities of a given speaker design?

Let's say that we have an MTM configuration, like the Ochocinco's.

Mids are 40W RMS each, but the total is only 40W, since they are in series (resistance is 8ohm since they are in series) and tweeter is 50W RMS for a total of 90W, since the tweeter is in parallel with the mids. These numbers are somewhat typical for a 5" driver, and the tweeter in this case is maybe a little bit higher than most, but definitely not out of line.

But, when you go and look at some of the commerical MTM's that have 5" in them, like the Polk LSi9, they will be around 200W, and some are higher than that. Granted there are some higher power capability mids out there like some TB's, but even then, in a system like the Ocho's, they wouldn't get to 200, if my math is right.

Now I may not be comparing apples to apples here since those are pretty pricey speakers, but how does anyone come up with the power handling numbers for a given speaker design? Can that be done like I did by just adding up the values, or do you have to add in some of the resistor and inductors values as well to figure the total? Or, can you only do this by actually running a pair, and seeing that the total value is for the maximum?

CADman_ks

[AdelaaR]

I personally think that "Watt" is a very nondescriptive term, for speakers at least.
Compare it to TT cycling (one of my other hobbies) and I can assure you I can put out over 1000 Watt with my legs!
... but I didn't tell you for how long I could keep that up, did I?
... and I also didn't tell you how much muscle damage I would suffer, did I?

I have a 1978 amplifier here that has 80W per channel. In those days they used "program power" and I can assure you it's way more powerful than many crappy "1000W" things I've heard.

Also ... making low frequencies consumes way more power than making high frequencies, so the numbers in your example don't compute that easily.

[CADman_ks]

...

Also ... making low frequencies consumes way more power than making high frequencies, so the numbers in your example don't compute that easily.

True, and I realize that. But, they still rate speakers for some type of a power input. I suppose that this could be a fictitious number that they've pulled out of the air, but when it comes to the higher end commercial offerings, they are putting a number on there. With the higher end commercial offerings, I would have to think that they would also be putting a somewhat realistic number on there, so that consumers don't blow them up, since they do carry quite a bigger price tag.

Maybe my logic is all wet here...

CADman_ks

[cjd]

I don't really know where they get their numbers, but it's definitely wonky and relatively useless data. Generally, that's the amount of power the voice-coil will take before melting down at a given frequency. If you pick a frequency in the middle of the crossover, I bet you could feed quite a bit before either coil melts.

At any rate, I usually look at the excursion limits for power handling.

Also, clean power will always allow more than dirty power, if that makes any sense.

[AdelaaR]

I would have to think that they would also be putting a somewhat realistic number on there, so that consumers don't blow them up
[CADman_ks

You shouldn't be able to "blow up" a speaker by driving it with a huge amp.
If my logic is right, it isn't the amp that is driving the speaker ... it is the speaker that is in fact "pulling" the amp.
All electricity works this way.
So ... any speaker can and will have huge power spikes that only last fractions of a second ... and any amp should be able to deliver these huge power spikes.
So the question isn't "how much power?" but "how much power for how long?".
Speakers, like amps or legs or cars or anything, shouldn't be operated near their maximum anyway.

Take a car for instance ... do you know exactly how long it takes for your engine to break down at a given number of rotations per minute?
Do you ever drive your car at 8000?
The only way to get a real answer to your question is to take a multimeter and measure the amps driving a certain speaker ... then drive up the power until the speaker breaks down ... then you'll know

[cjd]

Do you ever drive your car at 8000?

I wish I could rev that high - mine made peak power (and still climbing) at the revlimiter (6750 in this case) but the bean-counters didn't want to let it rev higher (can't entirely blame them, cam has pretty steep ramps so the valve-springs would need to be quite good... but then, people DID take the same engine on an older computer to 8k without trouble... ) It was already close to 5% away from rated power, except that was at the wheels, and teh factory rates at the crank. Have a good build, I did. Though the clutch isn't rated for that, and was slipping in the dyno rather clearly...

And when I raced, I hit revlimit a lot. Or close.

This just illustrates even more clearly how complex such a seemingly simple question can be.

[CADman_ks]

...

At any rate, I usually look at the excursion limits for power handling.
...

OK, that makes sense. But what number or graph should I be looking at? These have an Xmax of 20mm at excursion, but I don't see any number that indicates what the load will be at 20mm excursion.

...

Also, clean power will always allow more than dirty power, if that makes any sense.

I figure that I'm going to be on the dirtier side of power. I'm driving these speakers that I'm building with a Pioneer VSX-821-K. Not the cleanest power, I'm sure, but probably not the dirtiest either...

CADman_ks

[CADman_ks]

...
Do you ever drive your car at 8000?
...

I don't own a vehicle that even get's close to that! Mine all have redlines in 4,000 range OR LESS! (I have diesels. )

...
The only way to get a real answer to your question is to take a multimeter and measure the amps driving a certain speaker ... then drive up the power until the speaker breaks down ... then you'll know

LOL. I'll get right on that.

CADman_ks

[cjd]

OK, that makes sense. But what number or graph should I be looking at? These have an Xmax of 20mm at excursion, but I don't see any number that indicates what the load will be at 20mm excursion.

I'll make it simple.

1W

I have no idea what frequency that will be though.

How loud, how low, how long.

More power than they can handle is actually safer than less.

[---k---]

Those "Watts" numbers are a bunch of hooey. I'm stepping beyond knowledge comfort zone here, but The correct way to answer your question is to model the drivers in Unibox (or similar software). With computer models, you will see for a given enclosure and a given driver at what amount of watts the driver reaches xmax and what the spl is at xmax. That will be the correct way.

Here is a graph I made when making the Ocho's. You see in a 14L box, the drivers will reach xmax with only 20W. Step up to 21L and you're down to 15W. BUT, you're putting out 100db with only 15W. And, if you have a subwoofer in play, all this is somewhat moot. A different box size or different drivers gives different results. It just isn't straight forward, easy question to answer.

CJD and Curt discuss it a little in the Ocho thread.



Oh and the other thing is that some drivers are much more tolerant of being abused than others. I believe the ND140's are in that category.

[cjd]

Having trouble getting this to work right with a sealed box modeling, but this does give you the data you want, and shows what happens when you add a sub to the mix.

Green should roughly match Ryans graph. Black is with a 2nd order LR filter applied at 80Hz (basic THX spec does this I believe.) Gets you about another 9dB, which is a significant amount louder, and that still within xmax not xmech.