Speaker Efficiency

mcgsxr · « **on:** 7 Mar 2005, 12:26 am »

I have a couple of questions about speaker efficiency, and I am hoping that the collective minds here at the Circles can help me out.

Recently I built a set of speakers that I really like the sound of, at and beyond, a certain volume. Below that volume, the soundstage seems to compress, and collapse. The speakers, for the record, are based around the WR125 full range driver available from DIY Cable. The raw drivers are rated at 86db, and the finished speaker is rated at 90db, from what I can discern from the pdf file on DIY Cable that outlines some tests that were done.

Now, I know that higher efficiency speakers will allow for the use of a lower powered amp, among other things, but I love my Bolder modded Teac (at a modest 30wpc) so I am looking to improve my sound, and am wondering WHAT ELSE does having higher efficiency speakers do?

I suspect that it will mean that my volume control will become more coarse (ie less travel for more increase).

What else can I expect?

I am thinking about building up a set of Visaton full rangers, in an open baffle, but want to hear about experiences others have had with increasing speaker efficiency, but retaining the rest of the system.

Anybody want to chime in on this for me?

Thanks,

Jon L · « **Reply #1 on:** 7 Mar 2005, 12:49 am »

30 WPC is a TON of power IMO. I used to run 1.5 WPC pretty successfully with 95dB sensitive speakers (not driving bass). I also used to use 30 WPC with 88dB sensitive "normal" speakers, and it got as loud as I wanted.

With 30 WPC available, all you would need is really >90dB sensitivity with stable impedance. Just because one speaker rated at 90dB doesn't come alive, it doesn't mean another speaker with 90dB won't.

Just remember, there's almost always a tradeoff between sensitivity and bandwidth. The more sensitive the driver, the less frequency extension it's likely to have (practical FR, not what's on paper). This is why a really top-end high-sensitivity speakers (Edgarhorns, Avantgarde, etc) use many drivers, each covering a narrow range. You can find two-way speakers that have good extension, but usually at 87-88 dB type of sensitivity.

I know full-range drivers are popular now, but most of them should really be used to cover much less bandwidth than measurements/paper says to Really come alive, IMO. For example, those Horning speakers that use the "full range" Lowther driver only for broad midrange after cutting off the whizzer cone.

Gordy · « **Reply #2 on:** 7 Mar 2005, 12:57 am »

My thinking is that lose of dynamics at lower levels is more a funtion of the preamp/volume control than speaker efficiency. I was amazed at the life (and clarity) to be found at lower levels once I switched to a tranny based pre! Could you possibly borrow one for a trial?

Gordy

SET Man · « **Reply #3 on:** 7 Mar 2005, 02:03 am »

Hi mcgsxr.I have 94dB Single Driver Fostex 167E. I preiously have Magnepan 1.5 at 85dB. And now also have single driver 3" the "pipe dream" at 88dB.

From my expreince with higher ef speaker the sound is more dynamic than the lower one. This have to do with the higher eff driver seem to respond quicker to the single than the lower eff driver. And they do retained the dynamic better at lower volume.

Overall I feel that higher eff driver have better dynamic then the lower one. Than there are speaker system that use many drivers to gain higher eff. But with each driver at lower eff I think they still won't respond as quick as higher eff driver.

Oh! of cause the volume will more coarse for sure.

Buddy

roymail · « **Reply #4 on:** 11 Mar 2005, 08:30 pm »

At moderately loud listening levels, is it true most of the time that the sonics start to compress and sound coarse loosing the purity of the sound especially in vocals? I'd really appreciate some insight from some of you guys who have been doing this awhile. Thanks.

avahifi · « **Reply #5 on:** 11 Mar 2005, 09:05 pm »

Dynamics, clarity, compression, distortion, all are interrelated and speaker efficiency, amplifier power, amplifier linearity, and more all enter into the equation.

All audio reproduction devices are essentially non-linear and become more non-linear as their physical and electrical limits are approached. If you plot the input vs: output of almost any device, the plot is not the diagonal straight line you would hope for, but is an "S" curve, with the curve flattening out the most as limits are approached. Thus one wants to run any audio device as far within its limits as possible as that is where the best linearity exists.

Speakers become non-linear as they near the mechanical limits of their suspensions. Amplifiers become non-linear as they approach their power supply rail limits, and even your ears become non-linear as they shut down to protect yourself from very loud noises.

Thus a reasonably good approach would be to have a relatively efficient speaker with good power handling capability driven by a very good amplifier with more than adequate power for it, fused to avoid accidental damage.

In general, horn loaded speakers, while gaining great efficiency, do this by trading off one form of non-linearity for another. The efficiency is gained by creating an acoustic amplifier, which is done by generating an underdamped resonant peak in the range the horn is tuned to. The problem with this approach is that underdamped resonators resonate, generating output by responding to harmonics of the tuned frequency, and with a frequency response that falls off very rapidly outside of the tuned horns range. A good "worse case" example is the honking sound of a cheerleaders megaphone. Of course a very low powered amplifier matched with a horn loaded loudspeaker may sound better than that same amplifier running at very high distortion trying to drive a less efficient speaker.

So as your system starts to compress, it is because the speakers, the amplifier, the room, or even your ears might be reaching their linear limits.

The real task is to carefully figure out which part is the worse case problem and fix that first.

Frank Van Alstine

roymail · « **Reply #6 on:** 11 Mar 2005, 11:16 pm »

Frank, thanks for the great post and insight. Since I'm a bit technically challenged, I think I understood most of it. OK, you said, "The real task is to carefully figure out which part is the worse case problem and fix that first." So how and where do I start this process? Since I've got questions, I know you've got answers. Thanks for the advice.

-- Roy