High efficiency speakers vs Low, pros and cons

Let me shoot this question. Theoretically, why would the first engineers to design a low sensitivity/efficiency (???) (like 84dB/w/m) when ligh sensitivity/efficiency (??) speakers were available? Considering the fact that the first viable amplifiers were diminuitive (3W SETs, etc) naturally speakers that designed around those amps. Since all things being equal (and they never are) a 50W or 150W amp generally costs more than a 3W or 8W one. What would be the driving (pun not intended) reason to reduce sensitivity/efficiency (??) in a new loud speaker design.

Flatter response over wide frequency range?
Cheaper construction or cheaper R&D costs?
A different "sound?"

For example consider a $75-200 loudspeaker from Circuit City. (Essentially a shoe box with some drivers). Why should these designs be 87dB/w/m. Why not make then 97dB/w/m? Since someone that price conscience about loudspeakers probably can't afford a decent amplifier (either SQ wise or power output wise). You'd think the HT crowd would be all over high sensitivity/efficiency speakers, since you could drive 5 our 7 speakers in a big room with just about any A/V receiver.

Tubes vs. SS. Analog vs. digital. MC vs. MM/MI. Active vs. passive. 3way vs. 2way vs. 1way. I realize everything is a trade-off. And no matter what the technology or features, there are generally great executions of all of the above. And it's more about the end result than any specific technology. There are no absolutes. No "magic bullets."

This is just something I've always wondered. What do low sensivity/efficiency speakers "bring to the table" so to speak?

This is a VERY technical question and involves the engineering trade-offs between transient response, bandwidth, driver excursion, cone break-up ( distortion ), power handling, etc... Suffice it to say that if one could achieve "perfection" i.e. all of the desirable traits that we look for in hi-fi with high efficiency, the designer would be a genius that knew how to circumvent the laws of physics. As such, engineers and manufacturers weigh the balance in what they are looking for and design according to what they have available to achieve those goals at a given price point. As such, a medium efficiency speaker offers the best compromise in several areas. On top of that, they are also the most commonly available in terms of selection of drivers, so that is what we end up with as a majority. As others have mentioned, either design can work well in a system IF that system is set up to work as a team. Like anything else though, the team is bound to have a few "star players" ( high points ) and a few sore-spots ( weaknesses ) regardless of the efficiency.

My suggestion is to look for speakers that are at least "reasonable" in efficiency ( 88+ dB's ), offer a stable impedance that is 6 - 10 ohms nominal and will cover the frequency range that you want in a room the size that you'll be using them in. Then again, finding all of these features in one package at a price you can afford is "almost" like having your cake and eating it too : ) Sean
>

Aroc asks an interesting question. I'm hardly an expert but as I've come to understand it, the first speakers were actually extremely efficient. The early years of vacuum tube amplification required efficient single driver baffle mounted speakers. Musical, but pretty crummy frequency response. Later came the "great" multidriver high efficiency horn speakers (Klipsch's, Altec's, JBL's, etc).

The less efficient sealed box designs utilizing multiple drivers and power robbing filter components happened to come at a time when we saw higher output tube and transistor amplifier designs. Efficiency was no longer a big deal. The culmination of this was probably the small closed box mini-monitors that utilized complex power hungry crossovers and a sealed box to create "bass" from much smaller than expected drivers. (I played with these designs for awhile in the early 80's. The crossovers were complex and the resistor networks in them often became quite warm to the touch at moderate listening levels. Lots of the amplifier power was used to warm the room! I recently acquired a blueprint copy of the KEF 101 crossover from the factory and that extremely complex 2-way crossover included 25 components including a relay and an LED!)

Vented box designs promised better efficiency with even lower bass. (Unfortunately it's not always well executed.) That has sort of made them the predominant design for the last couple of decades. Now, with audiophiles falling into various camps that favor everything from SET's with single digit wattage to solid state designs with nearly 4-digit outputs, speaker designers are free to create what sounds best to them with the knowledge that proper amplification can be found for nearly design.

Sean is right, unless you're running a SET amp, choose any reasonably efficient speaker. Other than that there don't seem to be an absolute pros and cons to high efficiency. If your amp is unhappy with low impedances, checking the impedance curve of a speaker is important before buying. Again, I'm certainly no expert but those are the only numbers I pay much attention to. I buy what I enjoy listening to.

Richard, everyone above makes good points.

There seem to be three advantages to seeking high sensitivity during the design of any speaker system:

--to use lower-powered amplifiers (= a simpler amp, which is often more linear/more transparent)
--to reduce voice coil temperature swings (sensitivity decreases as the voice coil's impedance climbs with temp, a form of "power compression")
--to be able to play it more loudly (assuming it can handle the power AND has the excursion).

Personally, I think #2 is the main "hi-fi" reason a speaker designer should seek the highest sensitivity raw driver for a given task.

There are several disadvantages/obstacles to increasing a driver's sensitivity:

--no matter how large/powerful the magnet, we can only get a finite amount of magnetization into the pole piece, located inside the voice coil.

--about half the moving mass of a cone or dome driver is in the voice coil. All the high-sensitivity woofers and mids have very short voice coils, to reduce their moving mass. For a woofer, that shorter coil reduces the maximum stroke available, which limits loudness and low bass excursion (think Lowther). A mid driver that has a short stroke forces the use of a higher-order crossover to keep it from running out of stroke, and that crossover always screws up the time coherence.

--the lower the moving mass, the more compliant that driver's suspension must be, to keep the driver's frequency response flat before a crossover is even applied. And to let the driver go just as low as its lower-sensitivity competition. Except there's a limit as to how compliant the suspension can be manufactured, for consistency from unit to unit and to keep the voice coil centered.

Most high-sensitivity speakers do not go below 45Hz, flat, because of their suspensions' intentional lack of compliance- done to limit the stroke of their short voice coils. They do not have to be "boomy" however- that's the result of poor engineering somewhere in the design process, not sensitivity.

If multiple drivers are used to increase sensitivity, there are many problems:
-little chance of time coherence.
-room positioning becomes critical, as off-axis, the tone balance and time-coherence (if any) are unpredictable.
-more reflections off the larger cabinet face.

The highest sensitivity, high power-handling drivers must be carefully assembled. One of the biggest decreases in sensitivity comes from widening the voice-coil gap for sloppier, faster assembly and out-of-round voice coils. This is why there are a lot of 83-87dB designs out there.

The best woofers, made by larger manufacturers, with high sensitivity and low distortion are from PHL Audio and Volt Loudspeakers. However none of those are "ideal"- some are not flat in tone balance, and others do not respond to small signals.

There is the 100dB-sensitive Stage Accompany 6" ribbon tweeter, if one could afford it (>$600 each). Regardless, finding a mid driver to mate to it seems impossible, without loading the SA down with resistors to turn it down (reducing its sensitivity).

In a complete speaker system that uses only a single driver per frequency range, the upper limit to sensitivity is now 90-91dB, if you are to have any useful bass response, a cabinet that's not too large, and wide dispersion.

Sealed-box woofer boxes are smaller than ported woofer boxes, for the same low-bass extension. To go that low, they have longer voice coils- they need the stroke since there's no port. Which means more moving mass, thus less sensitivity.

Sealed-box woofers have softer suspensions than ported woofers, so the sealed box itself can become the primary "spring"- a more perfect suspension than any man-made one, for lower distortion and better response to very soft signals.

Sealed box designs were necessitated by the onset of stereo reproduction, as no one really wanted TWO big Altec or E-V speakers in the living room back in the '50's, or could afford two. When those sealed boxes started selling in big numbers, that was one impetus for amplifier companies to increase power outputs. The other was "marketing".

Yes Richard, you did open a Pandora's box. Shame on you.

Best regards,
Roy
Green Mountain Audio

I talked to a horn speaker mfgr. He told me his speaker contributed at least 80% of the sound because amps sound virtually identical at the 1 watt used to run the high efficiency horns. Meaning, the lower power, the less effect amp qualities can have on make / breaking the system.