Do Speaker Drivers Create Heat?

Has anyone who works with amps or speaker systems ever known the back of a speaker (where the driver is) to get noticeably warm? hot? or, burning hot? From those few I’ve asked, no one has noticed this…and yet speakers are supposedly only 10%-40% efficient, at best, so why aren’t they cooking after cranking for awhile?

Also, when wattage is quoted for a speaker, is that assumed to be at max volume? Perhaps this is one reason, since no one can listen at max volume…unless a roadie can share his/her experience here!

Please help! I am trying to figure out (a) where all the heat goes without forced convection, fins, or heat sink…if speakers are so inefficient.

…And, ultimately (b) how my vendor’s ratings back-calculate to being 150% efficient (given 108 dB SPL @ 100 ft @ 400W which adjusts to 136 dB SPL @ 4 ft @ 400W actual…which is greater than the ideal which is 133.5 dB SPL @ 4ft @ 400W ideal …adjusted from 107.47 dB SPL @ 4 ft @ 1 W ideal).

Creating energy here,

• Jinx

Yes, speakers get hot.
I personally have seen tweeters burst into flame from being driven too hard.

Never seen one catch fire but they do get hot. At high performance you’re certainly pumping a lot of power through those things and therefore there’s going to be some loss to heat.

That wattage number represents a measure of how much energy the speaker can handle and still produce sound within a given fidelity measurement (ex: 25W RMS @ .1 THD).

It can handle more power but will distort more and at much more power it will reach a thermal damage threshold.

And speakers are designed to dissipate heat. The movement of the diaphragm causes air movement over the coils and helps to cool the speakers.

The amount of energy a speaker accepts is proportional to the volume level, so say a speaker is rated at 25 watts it may only be using 3 at a low volume.

What you’re probably missing here is that speakers are directional. Your calculations for the energy in the sound probably assume that the sound is being emitted isotropically from the source, instead of just in a cone in front of the speaker.

Chronos, good point. However, I am modeling outdoor speakers for tornado warning systems which are specially built to throw sound 360 degrees in the horizontal plane of the speaker stack. Granted, I am sure it does not generate a perfect sphere, but without going to the vendor just yet for more data…what shape might you WAG the “sound envelope” would be for these? I am sure you’ve seen these around. They look like a stack of pancakes on a pole.

You could try assuming a cosine(elevation) variation, just to get an idea of the impact. It’s almost certain that the variation with elevation itself varies with the sound frequency. Since audio covers many octaves, there won’t be one variation to look at.

Chronos and Zen, what do you think of this reasoning:

Books begin with the spherical sound envelope because it is nice and ideal. As such, they explain that if such a speaker were placed against a wall, the sound energy blocked by the wall has to go somewhere. This manifests as a +3dB gain in SPL. This is known as the directivity index (DI).

So, IF it is safe for me to assume sound wants to radiate in all directions despite the actual geomety of a loudspeaker, then I would argue that whether I have really have a cone or not is not important. The acoustical energy has to go somewhere, and a sphere would be most conservative with DI= 0dB gain in SPL. A cone would use that same acoustical energy in some other way…it doesn’t mean it is any less efficient, does it?

Or…eureka…maybe I should look at my logic the other way? Perhaps Chronos is trying to say that I am trying to apply vendor’s db rating without knowing how to adjust for DI (because I really have cone and not a sphere). The dB rating, reduced to account for the credit gained by DI, would yield a beliveable value for the acoustical watts that went into producing the cone.

I bet I just reasoned it out for myself, huh?

I’ll just point out that all electrical circuits generates heat… well, a superconducting one wouldn’t but you don’t see many superconducting speakers around.
Some more than others, but as long as there’s any electrical resistance somewhere, that resistance will turn electricity into heat.

But geometry is important. The driver only moves along one axis, and the back of the speaker is usually closed. I expect the actual shape of the dB surface will be similar to that of a cardioid microphone.

Maybe what I should really be asking is where I can find a table of Q values for different shape cones? For example, if Q = 1 for a perfect sphere, then there must be a table of Q values for cones of different angles of width. (FYI: I believe DI = 10 log(Q)). I have yet to find a text that addresses this.