ICEinLAVA's Domain

I learned about speaker sensitivity and power handling by example long ago. Back in the late ‘70s, I installed my $29.99 below-dash cassette player from Radio Shack and needed a couple of 6" by 9" speakers for the back dash of my trusty ’71 Chevy Nova. Because I was a high school senior saving most every penny for my future electrical engineer school tuition, the then popular 30-watt co-axial 6" by 9" speakers were both pricey and over-spec for my lo-fi 4-watts per channel at 4-ohms player. So I managed to scrounge up two 6" by 9" speakers from a junker Magnavox console TV set.

The speakers were stamped on the Alnico magnet assembly with 5-ohms and 5-watts, and looked to be a better match for my 4-watt amplifiers. My intuition told me that these speakers should be more efficient (sensitive) since they would give all they had at about five watts, while I suspected the 30-watt version would not necessarily give me six times the acoustic power, even if I purchased a 30-watt power booster unit. The auditionable result confirmed my suspicions, and I had plenty of thump coming from the back dash. So much that the image in my rear view mirror blurred to every kick drum and bass note played through the system.

From frugal student to soundman curmudgeon, I still get incensed about speaker cabinet specsmanship when I look at a broad range of sound reinforcement speakers and their sensitivity ratings. In general I see low to upper 90s (dB SPL at 1-watt and 1-meter) for the MI offerings, and upper 90s to mid-100s for the touring rigs.

I know it is not all black and white when it comes speaker sensitivity for MI or touring grades. Obviously long throw cabinets and horn loading will add a couple dB in the calculation. But even when you level the field on enclosures with direct radiating low-frequency drivers and wide coverage angles, there still is a difference. After all, the touring guys can afford super-sized amplifiers and can tolerate a little less speaker cabinet efficiency, compared to the weekend warrior musicians and the ankle-biting club-dwelling soundcos.

Taking the other side of the argument, I know that specsmanship and driver design trades all effect the final published speaker sensitivity. Even when I shop for speaker cabinets, I give a bit of leeway to the poor cabinet designer who accidentally found a dB or two of dip in the frequency response curves at 1kHz, manifested as a lower sensitivity number when comparing apples to apples.

What Is It All About?

When you get down to it, the speaker comparison game comes to quality and quantity. When you ask a group of sound engineers for speaker recommendations, they typically gravitate to brands of known quality and reliability. The history of pro-sound speaker manufacturing is littered with sound companies that started by trying to find bits and pieces of drivers, cabinets, crossovers, hardware, and cover materials—anything that would likely survive the bumps of the road and still sound decent. The sound companies who succeeded in assembling the “good stuff” into road-worthy enclosures are today’s pro-sound speaker manufacturers.

But today we have a pro-sound speaker-buying populace that has reached sufficient “critical mass.” Several dozens of manufacturers all over the globe offer reasonably complete lines of speaker cabinet choices. Each vendor has done plenty of market research and knows what configurations sell how many units per year, resulting in many similar speakers with different logo-plates. The quality has risen to the point that even the lesser known brands field road-worthy gear. So are we back to specsmanship?

Where’s the Beef?

Turning to the quantity side of the equation, available Sound Pressure Level is usually rated in decibels at a standard 1-meter reference standoff. Most manufacturers will offer both continuous and peak values, and some will only offer the peak. Besides relying on only the raw SPL measurements, both manufacturers and buyers should know that the final SPL can be calculated pretty closely by combining the sensitivity and the power handling numbers.

For example, a dual 18" subwoofer cabinet could contain two drivers, each with 400-watt continuous power handling ratings. If the pair in a ported, front baffled box provide a sensitivity of 101dB SPL at 1-watt and 1-meter, then an 800-watt signal in the center of their frequency response should provide about 130dB SPL continuous at 1-meter. Mathematically, the reason for this is that 10 times the logarithmic value of 800 is 29, or 29 decibel Watts. A reputable manufacturer may give both the 130dB SPL continuous rating and the 136dB SPL peak rating, but a less scrupulous vendor may just give the 136dB peak, hoping that you space out on the peak suffix and compare 136 to other vendors’ continuous rating numbers.

How do you get 136dB peak SPL at 1-meter, you ask? In our example, our 400-watt drivers may indeed have 800-watt program power ratings and 1600-watt instantaneous (peak) power ratings. The same example subwoofer may then advertise a 3200-watt peak power capability. Ten times the logarithm of 3200 is 35—or 101 plus 35 gets you 136db peak SPL.

More Power, Scotty

This brings us back to what speaker cabinet manufacturers have traditionally hyped—power ratings. If you are a speaker designer, you know that about 80% of the amplifier watts delivered to a low-frequency driver are lost in heating the voice coil resistance. As a result, speakers are tested with pink noise signals and pumped up until the voice coil wires reach the fusing (melting) point. This becomes the thermal limit of the voice coil in watts; backing off a touch to survive the long-term testing gives us the continuous ratings.

But in reality, music signals are not short-term continuous test signals. To reach an equivalent voice coil thermal power input using music signals, you would typically double the music power levels (+3dB). This is defined as the program power rating. Also, this is around the maximum power you want to use to size an audio power amplifier to a speaker. If you play the sensitivity and power numbers game, use the program levels to best depict the acoustic capabilities of your system. In the example, the drivers are rated at 800 watts program, and together, create 133dB SPL at 1-meter given an amplifier that delivers 1600 watts.

Peak power ratings are more of a driver physics proposition than a power issue. Most driver designers will aim for a continuous power rating +6dB over the peak power capability. This limit is what the driver can mechanically (acoustically) perform before the driver component damage occurs. This damage scenario could include the voice coil popping out of the motor assembly, the coil former banging against the back of the magnet or magnetic motor assembly, folding of the cone, or tearing of the cone surround or spider. With today’s ultra-high-power amplifiers, sound companies are beginning to see more driver mechanical damage than fused voice coil windings.

One other point to make is that driver design has progressed so much over the recent years that the thermal design of voice coils and coil formers has improved to the point that secondary thermal issues have become primary considerations. The heat dissipated by the voice coil now efficiently transfers to the nearest thermal mass, the magnetic motor assembly, and heats the magnets. Unfortunately, magnetic materials have a bad property of losing magnetic field strength if they surpass a certain temperature, known as the Curie temperature. This results in speakers that actually get quieter as the concert goes on, and may even spiral into driver thermal destruction if the person at the console continues to crank up the volume to compensate.

My Soapbox

The point of this blabbering about specifications and physics is that I want speaker buyers, especially entry-level buyers, to really pay attention to sensitivity ratings when doing apples-to-apples product comparisons. Next, I want speaker manufacturers to stop foisting these trashy, flashy entry-level cabinets with bad specs on the consumers. If I choose a cabinet with 3dB less sensitivity than the others, that means I need 3dB more power—twice as much—to get the same loudness. Sensitivity improvement is not necessarily only for the “rich and touring,” but a design aspect that can be brought into cost-effective products as well.

In numbers, this means that single-driver boxes should be able to meet or exceed 98dB SPL at 1-watt, 1-meter. I feel put off by long-time and reputable manufacturers offering 95 to 97dB SPL designs, and even trying to market them in a professional context. I am even more insulted by seeing these “DJ-market” crumble boxes sporting a piezo tweeter or two with 92dB SPL sensitivities. If I want that kind of power efficiency, I know where the hi-fi store is. Yes, quality of sound is important, and should command premium dollars for premium performance. But the business of sound reinforcement is also getting the job done efficiently.