Speaker specifications: A buyer's guide
Dynaudio C-1 speaker with a 6.7-inch woofer and 1.1-inch tweeter
(Credit: Steve Guttenberg)Listen, if you're buying speakers, the best thing to do is listen to them. If you can't, you'll learn a lot perusing specifications.
Online shoppers need the most info, but even brick-and-mortar buyers can predict quite a lot about a prospective speaker's performance and system compatibility by looking at the numbers. Tech talk gets you woozy? No problem, the main points of interest would be the dimensions of the speaker and how much the thing weighs.
Hint: bigger and heavier speakers almost always sound better than small, lightweight ones.
So far, so good, but woofer size also provides clues about how much bass a speaker will produce. Bigger woofers make more bass than smaller ones, but the cabinet's size is almost as important a factor as woofer size in determining bass quality/quantity. Multiple woofers may look impressive, but they're less reliable predictors of bass oomph; we've tested lots of skinny speakers with two, three, four, or more drivers that didn't make much bass. The cabinets were too damned small, four 3-inch woofers don't make as much bass as one 6-inch woofer in a larger box.
You might hope the frequency response specification, "38 Hertz -- 22,000 Hertz," would predict how much bass the speaker would make. The first part of that spec, "38 Hertz" refers to the bass output--the lower the number--the deeper the bass. Ah, but since there wasn't any reference to the tolerance of the spec, it's just about useless. Some manufacturers include a "+/-" after the frequency response number, something like 38 Hertz - 22,000 Hertz +/- 3 dB, and that helps a bit. Any speaker that reaches 50 Hz or lower can be considered "full range" and doesn't necessarily need to be partnered with a subwoofer--for music. Multichannel (5.1, 6.1, 7.1) home Theater applications almost always require subwoofers.
I have less to say about tweeters, other than to recommend buying speakers with tweeters; a fair number of pricey "sound bar" speakers are tweeterless. That's not to say they will sound bad, just not as detailed and clear as speakers with decent tweeters.
The sensitivity (or efficiency) spec can be helpful in predicting how a speaker will work with low or moderately powerful amplifiers/AV receivers. An average sensitivity spec is something around "88dB/1watt." Higher sensitivity speakers, 90dB/1 watt are "easier to drive," meaning they will play louder with fewer watts than less sensitive speakers. If you have a $500 or less expensive AV receiver, even if it's rated at 100 watts per channel, I'd still recommend going with speaker rated at 90dB/1 watt. 95dB/1 watt speakers will rock your world with 50 watt per channel receivers.
Next, the impedance spec. This one is a little tricky because it's so misleading. You might think an "8 ohm" speaker is an 8-ohm speaker, but depending on the frequency the speaker is playing at any given instant it might be 14 ohms or 3.6 ohms (impedance constantly varies with frequency). Yes, the impedance may average 8 ohms, but that doesn't help your receiver when it's attempting to reproduce the thundering sound of dinosaur footsteps. Be that as it may, try to go for higher rather than lower impedance speakers if you're using budget- or moderately priced electronics (the lower the impedance, the more power the speaker needs).
I've covered power handling specs before, typically stated as "up to xxx watts," but the spec is really pretty useless. For example, it implies that you'd be living dangerously if you hooked up a 200-watt-per-channel amplifier to a speaker rated for 100 watts. Not so! In fact, you'd be more likely to harm said speaker with a 50-watt-per-channel receiver, it you played it loud. Why's that? Because the 50-watter would be way more likely to clip (distort) when playing loud than the 200 watt amp.
The lower power receiver's distortion is way more likely to "blow" a driver. Yes, if you went nuts with the 200-watt beast it would also cause damage, but at a much higher volume level than the 50 watt receiver. Either way, a little common sense restraint will save the cost of repairs, but the point here is the power handling spec doesn't provide any real guidance. Ignore it.
Hint: If you want to play loud, even if only now and then, buy bigger speakers and more powerful electronics.
Did we miss anything? Please submit your speaker spec questions here.
Steve Guttenberg is a frequent contributor to magazines and Web sites including Home Entertainment, Playback, and Ultimate AV. He is a member of the CNET Blog Network, and is not an employee of CNET. Disclosure. 
Impedance (complex resistance). It's easier to drive more power into a speaker with lower impedance. Suppose, you have an amplifier with maximum output amplitude 50 V. If you have 8 Ohm impedance, you can drive 156 W of sine wave (50*50/8/2). For 4 Ohm, it will be 312 W.
The drawback is that your amplifier efficiency would be lower at a given output level for lower impedance. Suppose your amplifier is powered from +/- 55V. Then, at 50 Watt sine output, its last stage will consume 160W on 4 Ohm load, or 113W on 8Ohm.
Impedance is mostly defined by the coil wire gauge and number of turns. DC resistance only depends on the wire length, while impedance on a given frequence depends on more factors. But, given the same construction, if you have twice as many turns with wire that is 0.707 (sqrt(0.5)) as thin, the speaker efficiency (and most other characteristics) will stay the same, while its impedance will be 4 times as high.
That assumes that the coil is wound with the same size.
You don't seem to understand power rating concept. If your amplifier can produce 200 W, and your speaker is rated to 100, you're more likely to fry it, than a speaker rated for 200W. Assuming their efficiency is similar.
Said that, the most danger is for tweeters, because they can't stand the full power; their actual rating is quite lower. If you happen to feed "speaker killer" sound to them, they're fry.
But you'll be very hard pressed to produce enough distortion to damage them by the overdrive. Suppose your amplifier can drive 150W of sine wave when it starts clipping. If you overdrive it so much the output becomes a square wave (you would never do that), full output power is 300W, of that about 255W is main harmonic, 29W third harmonic, 11W fifth harmonic, 5W seventh harmonic. Make your conclusions.
Your efficiency numbers are 2dB off, too. With a 90dBa on W speaker, your 100W amplifier will give you 110 dBa (113 if both channels). With a 95dBa on 1W, a 50W amplifier will give you 112dBa (115 with both channels on).
One more thing, you are definitely more likely to damage your 100W speakers with a 50W amp than you are with a 200W amp. Steve is correct about that in the article and his reasons are correct as well. Think about it this way - when you're listening to your 200W amp you likely aren't trying to annoy the neighbors. A 200W amp does not output a continuous 200W, it depends on what the speakers are doing. What frequencies (bass is harder to produce than treble because the waves are much larger) and what volumes you're trying to reproduce is the determining factor here. On the other hand, a 50W amp is much easier/more likely to be driven into clipping due to the fact is has less power to handle the transients (explosions in movies, that loud part of your favorite song), which does damage the speakers. It's easy to do, everyone has done it at some point in their lives - and your lack of understanding that speaks volumes.
Now, can you tell us all... which brand of aftershave is most impressive to young, nubile women?
Do you agree with Steve, that "the lower the impedance, the more power the speaker needs"? Do you have any proof for that? Can you point to any flaw in my arguments about impedance?
Just because the amp is rated to 200W doesn't assume the listener will be driving it that hard, its is just allowing headroom that the 50W amp just would not have.
It's largely in the pairing, matching speaker and amp. No primer that tells you that you can purchase a pair of speakers without hearing them powered by a amp with the same sound signature as what you have in your listening room is worthless. Ideally, you take your amp with you. But even then the listening room will have different acoustics.
Curses.
Dalkorian is correct that you will damage a speaker MUCH MORE readily if you use an underpowered amp driven to high volumes, than using a high powered one with superior reserves, the current is simply cleaner and less corrupted. Also, the reason impedance is germane to Steve's point is not for the SPEAKERS, but for the AMP. You're reversing his point. You will stress an amp and potentially damage it, if the speakers have a lower rated impedance, because of the heat generated from the higher current. The rest of your point is, sorry to say, not at all relevant.
Took the following CD: Gustav Mahler's Symphony No.3, with Benjamin Zander and Philharmonia Orchestra, cut 40 seconds, starting from time 33:07 of the first movement; this fragment is pretty much tutti ff. Increased volume by 3 dB, to make it close to clipping; at 33:43 it would actually briefly clip. Assume that's the output of a 200W amplifier. Made a copy of that fragment and clipped it at 70.7% level. That would be an output of a 100W amplifier. Clipped another copy at 50% level; that would be an output of 50W amplifier. Separated each copy with single-pole crossovers, at 300Hz and 3000 Hz, to make bass, mid-range and tweeter inputs. Calculated RMS at peak levels.(where the most distortion would be). For mid and high ranges, the difference was less than 0.5 dB between no clipping, 70% clipping, and 50% clipping. So distortion doesn't bring enough harmonics to effectively fry your tweeter.
Said that, an underpowered transistor amplifier is much more likely to suffer a thermal breakdown. If one side of the last cascade breaks, it gives you full DC on the output; that could really fry your subwoofer, though not midrange and tweeter. Another consideration is that power cascade could suffer parasitic ocillation on high load, which your tweeters would not like. A follower is prone to that, especially with a Sziklai pair. This is most likely with an underpowered amp, too.
The original claim was that you're more likely to fry a 200W speaker with a 50W amplifier than with a 200W amplifier. This is what you should call "too much generalization". 200W speaker with what impedance? 200W amplifier rated for what impedance? Is it thermal ratng, or -20dB THD rating? 50W amplifier rated on what impedance? What are those amps using: tubes, bipolar transistors, or FETs?
Without considering all the real variables, Steve's claim becomes a piece of BS.
And by the way, nowhere in that claim (about 50 vs 200W) he mentioned impedance.
Took the following CD: Gustav Mahler's Symphony No.3, with Benjamin Zander and Philharmonia Orchestra, cut 40 seconds, starting from time 33:07 of the first movement; this fragment is pretty much tutti ff. Increased volume by 3 dB, to make it close to clipping; at 33:43 it would actually briefly clip. Assume that's the output of a 200W amplifier. Made a copy of that fragment and clipped it at 70.7% level. That would be an output of a 100W amplifier. Clipped another copy at 50% level; that would be an output of 50W amplifier. Separated each copy with single-pole crossovers, at 300Hz and 3000 Hz, to make bass, mid-range and tweeter inputs. Calculated RMS at peak levels.(where the most distortion would be). For mid and high ranges, the difference was less than 0.5 dB between no clipping, 70% clipping, and 50% clipping. So distortion doesn't bring enough harmonics to effectively fry your tweeter. "
This may be the most ridiculous thing that I've ever seen being passed off as an "experiment." Why would someone test the hypothesis "at the same [loud] volume, a 50 watt amp is more likely to fry a speaker rated for 100 watts than a 200 watt amp driving the same speaker" with this, um, method?
"Next, the impedance spec. This one is a little tricky because it's so misleading. You might think an "8 ohm" speaker is an 8-ohm speaker, but depending on the frequency the speaker is playing at any given instant it might be 14 ohms or 3.6 ohms (impedance constantly varies with frequency). Yes, the impedance may average 8 ohms, but that doesn't help your receiver when it's attempting to reproduce the thundering sound of dinosaur footsteps. Be that as it may, try to go for higher rather than lower impedance speakers if you're using budget- or moderately priced electronics (the lower the impedance, the more power the speaker needs)."
- by hk2000 December 19, 2008 12:37 PM PST
- The lower the impedance, the higher the current, since the power is proportional to the current and voltage (signal at speaker terminals), So yes, the lower the impedance, the higher will be the power needed. Power= current x voltage.
- Like this Reply to this comment
-
-
- by alegr December 31, 2008 10:13 AM PST
- Man, you think the same way as my science teacher 30 years ago. She was trying to convince me that the more powerful electric heater is, the more resistance it has.
- Like this
-
- by alegr December 31, 2008 10:57 AM PST
- More on that.
- Like this
-
(20 Comments)Speaker efficiency (how much electric power converted to actual audio wave power) doesn't depend on the coil resistance, assuming that the coil and diffusor size and mass is the same.
If you have a speaker with less impedance, your amp will have to pump less voltage, but more current, to get the same power. It's just different load for the amp, nothing more.
Suppose you have one speaker with 4 Ohm impedance, and another with 8 Ohm. We'll assume that its mid-range and tweeter also have twice as much impedance, and crossover circuit is also changed to match that. If the first one has, say, 96 dBa on 1W, another one will also have the same efficiency.
Now suppose you have an amplifier that can give you maximum amplitude 50V. WIth 4 Ohm speaker, you can get 312W out of it; with 8 Ohm speaker, you get half as much. Assuming the amp thermal design is rated for 4Ohm load, you get 3dB more with it, than on 8 Ohm speaker.
Force that the voice coil makes, is proportional to the current and number of turns. As the coil moves, voltage is also induced in it, which adds to the impedance. This voltage depends on the number of turns.
Suppose we have a coil wound with 0.2mm wire. Suppose its DC resistance is 4 Ohm. If you take 0.237mm wire, and wind 0.707 as fewer turns, its DC resistance will be 2 Ohm. Wire mass will be the same. Now, all other equal, assume on certain frequency the first speaker will have 8 Ohm impedance. Now feed the second speaker certain voltage that the current is 1.414 more than in the first case. It will produce the same force and the coil will move with the same amplitude, producing the same acoustic power. Its movement will induct 0.707 as less voltage back (because of fewer turns), which corresponds to half as much impedance (remember the current is 0.707 of the first case). Now we've got a speaker with half as much impedance, but with the same efficiency.