The nominal rating is the impedance the speaker manufacturer wants you to think the speaker is. The minimum impedance is closer to what really matters: what's the toughest load a speaker presents to an amplifier? (Impedance varies with frequency, as you may already know.) But when I see a minimum impedance of 4 ohms, I generally assume that means "3 ohms."
As for amps, it's not so easy. What you'd like to know (among other things) is how much power the amp can produce in a short burst at the frequency at which your speaker's impedance dips to its lowest point. But your power spec is for a continuous tone (by government fiat, by the way), usually at 8 ohms.
As James Boyk once wrote, the only two specs that mean anything are the dimensions and the weight. The dimensions tell you whether it will fit on your shelf, and the weight tells you if your shelf can hold it.
There is no link between the db and the ohm rating.
The dB rating basically tells you how loud the speaker will sound when driven with a certain amount of power.
The ohm rating, the impedance, can tell you how difficult it is for an amplifier to jamn a certain amount of power into the speaker. Neither a very low nor a very high impedance is desirable ... most amps are happiest around 8ohms ... they're designed that way.
Read on if you want to get technical (though it's 10 years since I did any amplifier design).
Amplifiers are limited by the current supply (usually down to size of transformers, capacitors, and duration of the loud passage), and by maximum output voltage, which is an inherent part of the design (usually near the power rail voltage, switched by the power transistors for SS amps).
From Ohm's law, V=I*R, and from the power law P = V * I one can derive that the power delivered to the speakers can be expressed in terms of voltage OR current :
power = V*V/R OR
power = I*I*R
Therefore, given the maximum voltage of the amplifier, or the maxiumum continuous current (or both), and given the worst case R, you could calculate the maximum power which could be driven into the speakers. (This is theoretical and the real value is a little lower).
So power delivery to the speakers when the speakers are higher impedance is usually limited by the maximum voltage of the amplifier (before clipping occurs).
Conversely when speakers have very low impedance the power delivery is limited by the current driving ability of the amplifier.
If you can find out the rail voltage and maximum continuous current of your amplifier and the minimum and maximum impedance of your speakers you could calculate a rough maximum power transfer, then use the dB efficiency to come up with the SPL. Of course, they're all AC signals, so you need to multiply by 1/root PI, or something like that .. it's too long ago for me to remember.
You have to look at ALL of the specs, inch by inch. If they spec "at one watt", the spl levels are directly comparable regardless of impedance. If they spec "at 2.83 volts" and are a relatively lower impedance speaker, they are playing games. This takes into account that they are "averaging" the spl level over the "midband" frequency range and not just picking some arbitrary frequency that may produce a slightly higher peak SPL.
With a speaker that is "nominally" 8 ohms, 2.83 volts is equivalent to one watt of power. Applying the same amount of voltage into a "nominal" 4 ohm speaker is NOT one watt of power. That is because, even though the voltage is the same, current is doubled due to the speaker having half the resistance. The 4 ohm speaker may appear to be more efficient or as efficient, but in reality, it is sucking up measurably more power to obtain those spl levels.
Another little known trick is to spec the speaker at 3 ft instead of at 1 meter. While this is only a matter of 3.5" or so, taking the measurement a little closer and "fudging" the figures can help them out a bit. Combine the distance factor with a higher current level into a low impedance load can make things seem quite a bit "sweeter" than they really are.
So, to compare apples to apples, you have to have the same reference level i.e. X amount of wattage into the speaker and measured at the same distance. Otherwise, you can end up with more power ( same voltage but more current ) measured at a closer distance.
Even with all of that in mind, you need to realize that different types of speakers ( dynamic "cone" drivers, planars / ribbons / e-stat's, horns, etc...) will produce different spl levels at various distances EVEN if they may produce the same spl at 1 meter. This all has to due with how they load into the room, their individual dispersion characteristics and radiation patterns, etc...
Like anything else, there are TONS of ways to create specs, interpret specs and "bamboozle" the unsuspecting. That is why so many "old timers" say things to the effect that specs are relatively useless. Even with the few rules that the FTC has created to "protect the consumer", those rules have some BIG loop-holes in them. Your best bet is to keep your eyes and ears open and try to experience as much for yourself as you can. First hand experience can be a HELLUVA teacher... Sean
Not going to say very much except emphasize that this is REALLY an important question you ask. One of the most important issues when it comes to matching equipment. When an amp lacks for power for a given speaker/circuit it "clips". I think clipping is the most important distortion issue in stereo. It's not a matter of playing loud - the ability of your amp to deliver power(voltage & current as discussed in previous posts) will determine how you music sounds when it peaks - even at moderate levels.
The power output of the amp is determined by the load (speaker) impedence and the available voltage and current of the amp. Voltage, current and impedence work together.
Some speakers present a difficult load. If your amp is a one ton truck you do not want to put a two ton load in it. Many great "2 ton" speakers get poor reviews by people trying to move them with "one ton" amps.
Nominal impedence does not tell much. Impedence changes with frequency and certain resonant frequencies of the speaker itself. As noted above, an impedence curve of a speaker often looks like a mountain range.
My experiece is that if you are in the normal range of spending (under 10k) its best to get 1) an effecient speaker and 2) one that presents a decent load throughout the frequency range. This makes the amps job a lot easier. High impedences offer little chance for real power delivery. For example, an amp rated at 200W @ 8 Ohms will typically deliver only 3 Amps of current at best into a 12 Ohm load.
If you are new to audio/electricty this all may sound strange. If you are starting out think of it in terms of water instead of electricity. Electric potential is a pool or water near a hill. The hill represents the impedence of you speaker. The pressure needed to pump the water up the incline of the hill is voltage. The size of the pipe will obviously controll total amount of water that can go throught at a pressure (current).
The steepness of the incline may change just like music. Dynamic music causes steep inclines. The water (electric potential) does no good it you do not have the pressure (voltage) to get it up the incline. If your amp doesn't have the voltage it will clip. Sounds bad. Likewise, when there is no incline (low impedence) your amp may have design limitatioins on the amount of current it can handle when the water runs free so to speak. Current and Voltage abilities are built in design parameters of the amp.
Sincerely, I remain
In many cases(i will try to bring my point) the low efficiency speaker can be successfully driven with low-wattage amp with almost no compromise meaning "2 tone speaker with 1 tone amp" and also the other way arround: some of the efficient speakers realy need a helluvapower to drive them?...
As far as I know that's where the load plays the biggest role in the speaker. Despite all efficiency ratings tests, different freequencies will create a different load to the speaker. The speaker impedance has mainly inductive nature which has the relation X=wL where X is inductive reactance and w is natural freequency and L is inductance respectively. So on the higher freequencies the impedance is higher and respectively on the lower freequencies the driver(s) are pasing higher current i.e. create a heavier load to the amp.
To my opinion, the dB efficiency of the speaker isn't realy important parameter. Does anyone listens music higher than 100dB in their homes? I believe that there are only a few.
If you're not one of them than on the simple calculations to bring 86dB efficient speaker to the level of 100dB you will just need 35Watts.
It is possible and many of us know that it's possible to successfully build up a front end using inefficient speaker and low-powered amp(Not sometimes helpful advices such as "just listen and decide for yourself") if the speaker load is relatively stable and has only <=2Ohms load difference.
I've been experimenting with my 86dB/W/m rated Totem Forests driving them with 35W/ch Pathos towers which are 15Watts bellow the recommended power -- simply no compromise!
The different case comes when you need to "pump" a large or huge room(still I wouldn't pay attention to the dB).
BTW the main advantage of low-efficient speakers is load stability or at least a higher possibility for the load stability.
And finally I dare to agree with Bomarc about the 2 dimensions that you can trust about the speaker.
Marakanetz: Actually, Boyk said it about ALL spec sheets. The only way to know if a particular amp has enough oomph to drive a particular speaker is to put both in your listening room, and crank up the volume (slowly!).
Bomarc, With all due respect, the only dimensions that matter are not if it will fit on your shelf and whether the shelf will hold it. Specs can't tell the whole story but they have a lot of important things to say and some basic mistakes can be avoided by paying attention to them. Listening is not the whole story. I agree that ultimately the ear has final say but the ear cannot tell if the speaker is bad or if the amp is underpowered or if the room is acoustically poor or if there is some other design problem involved. You need to use and respect both. You cannot rely on specs in the marketing material.
Sincerely, I remain
Clueless, First, we need to distinguish between measurements and specs. Specs ARE marketing material, and I've yet to see a spec sheet that really told me anything I would trust.
Now, the right kinds of measurements can tell you plenty about the performance of a speaker, but those measurements are represented by 2- and even 3-dimensional graphs, not numbers. And very few manufacturers publish such graphs.
As for underpowered amps, leaving aside the extreme cases (5-watt SETs driving huge sealed boxes), the best test is to listen for clipping distortion.
Bomarc: Well, as usual, it's hard to say clearly all you mean in these short posts and, when you make the distinction between "specs" and "measurements", I tend to agree with you. Sometimes I think some folks around here are measurement adverse and so I periodically go off the deep end.
Sincerely, I remain
T, If you still need a few internet sources let me know.
Bomarc, that is a good point. Unfortunately, we tend to HOPE that the "specs" are actually based on REAL measurements of the equipment, not just marketing hype that paints a pretty picture. I guess that this is why i still like reading Stereophile, as they are about the only ones that really do provide hard facts and data in their reviews. Giving sonic impressions is fine, but as we all know, how something sounds in a different system to someone else could be COMPLETELY different. Sean
Yes, Clueless, I kinda suspected you and I were more in agreement than not. "Measurement" sometimes seems to be a dirty word around here. It needs people to stick up for it.
Sean: I haven't read stereophile for awhile because I shook my head over a couple reviews a couple years ago. You find them helpful and among best nowadays? Maybe I should take a second look. The only thing i been reading regularly is audioXpress.
Sincerely, I remain