Opinions on speaker/tweeter protection?

[elipsey]

Hi,

I'm building some bookshelf hi-fi speakers, and considering protection mechanisms. Some people think fuses are too slow to be good speaker protection, while polyswitches are only a little better (~3 seconds) [1]. I'm considering bulbs, which seem like they ought to be faster (but I guess I don't really know), and would like to hear your opinions about how to appoach this.

Parts express sells bulbs for this, which are rated for 12.8V, .97A [2]. My naive supposition would be that ~1A across an 8 Ohm load (my tweeter) implies that the bulb will begin to illuminate (and the signal will begin to compress) when dissipating about 8W. This doesn't seem like very much, in spite of the fact that these bulbs seem to be intended for high powered PA horns and such.

How does this actually work and how should I choose a bulb (or something else instead)?


[1] How to calculate fuse value for speaker protection? | Audiokarma Home Audio Stereo Discussion Forums
[2] Speaker Crossover Tweeter Protector Lamp

[The Dude]

Commonly, book shelf speakers are not run that hard and therefore I wouldn't think you'd need tweeter protection. I rarely see tweeter protection in consumer electronics products. Is this some sort of exception? In other words, do you plan to run the hell out of them?

[elipsey]

Thanks Dude.

It's not an exceptional case except in the sense that I'm building these to learn more about speakers. If I can add a cheap insurance policy (and learn how this works), all the better.

[g1]

The idea with the series bulb is that it increases resistance as it heats up, so acts as a dynamic fuse, or limiter.
So, even if it doesn't blow, it reduces power to the horn (or speaker) during high power use.
You seem to be hinting at this when you mention "compress", but I thought I'd make it clear that this is an intended effect.

[The Dude]

What g1 says is correct. I suspect, however, that with smaller wattage bookshelf style tweeters, you won't get anywhere close to the range that the lamp style limiter/fuses work and I doubt they will offer much protection. I agree with your statement in the first post, "This doesn't seem like very much, in spite of the fact that these bulbs seem to be intended for high powered PA horns and such."

[elipsey]

In general then, how might we choose a bulb wattage rating, given a tweeter wattage rating?

[The Dude]

There are many variables- tweeter impedance, crossover frequency, tweeter power rating, etc. I have, in rare instances, been asked to fuse tweeters. I find it best to simply insert a current meter in line with the tweeter and figure it from there.

[tedmich]

some speakers need more protection than others



but if sound companies renting their stuff out to meat headed DJs don't think this is worth it, maybe it isn't.

Eminence sells D-fend boards to build into speakers but they are not cheap, or widespread.

D-fend? Loudspeaker Protection Technology | Eminence Speaker

[Leo_Gnardo]

Parts express sells bulbs for this, which are rated for 12.8V, .97A [2]. My naive supposition would be that ~1A across an 8 Ohm load (my tweeter) implies that the bulb will begin to illuminate (and the signal will begin to compress) when dissipating about 8W. This doesn't seem like very much, in spite of the fact that these bulbs seem to be intended for high powered PA horns and such.

Bulbs/lamps in series with the tweeter not such a bad idea. Don't forget, you can parallel lamps for more current if you need. A single lamp such as PX sells, or you can source from an auto parts store because that's what it is, should be well enough for any bookshelf speaker unless you plan to run it at disco volumes. Community has used lamps to protect speakers for decades in their MI gear. And the "ol' reliable" UREI 813 large studio monitors have 'em too. That's how they stay "ol' reliable."

There's also Raychem auto-reset fuses. They look like disc capacitors, fairly cheap, and are included in Peavey's crossover for their coax speakers. TomCarlos put up a photo on his Peavey ECoustic thread a couple days ago. They do have some resistance - they're positive temperature coeffient resistors anyway, that's how they work. So plan on having the "cold" resistance in series with your tweeter network, generally not such a bad thing anyway as many tweeters "out shout" their woofer partners. With Raychems, if you really slam you speakers and they sound soft in the highs, should return automagically a couple seconds later.

[Enzo]

The bulb draws the 1 amp when there is 12.8v across it, and that gives me about 13 ohms effective resistance when hot. So to get 1 amp to flow through the whole thing, you'd need about 20v across it all.

[elipsey]

The bulb draws the 1 amp when there is 12.8v across it, and that gives me about 13 ohms effective resistance when hot. So to get 1 amp to flow through the whole thing, you'd need about 20v across it all.

OK, so the edge cases are 1) the cold bulb has a resistance close to 0 Ohms up to some voltage between 0V and 12.8V at which it begins to illuminate, and 2) the illuminated bulb and the tweeter have a combined impedance of 20.8 Ohms. Between cases 1 and 2, R for the bulb varies continuously causing compression.

For V less than case 1, the bulb has very low R (but perhaps still considerable?) so the tweeter operates at about the same volume as with no series bulb.

For V between case 1 and 2, the tweeter operates at lower volume than with no series bulb. (Also, the tweeter crossover frequency goes up, and Q changes).

For V greater than case 2, the bulb has R is at least 12.8 ohm. Does R keep going up, or just stay the same untill the bulb melts?

It sort of bugs me is that the light bulb isn't matched to the power rating of the tweeter. Parts Express sells exactly one lightbulb for this, and 1A ought to be enough for anyone, right? Since I have a 50W tweeter, it seems like it would be good if the bulb came on at a little higher power than 8W dissipation by tweeter. Let's say I want the bulb to illuminate when the tweeter dissipates 25W. This implies about 1.75A across 8 Ohms, and about 22.4W across 7.31 Ohms for 12.8V light bulb. Maybe this is one of those things where the perceived difference in volume (for starting to limit at 8W vs. 25W) is small because it's logarithmic w.r.t. power, so it doesn't matter very much. Or maybe it's just that you can't buy a 22.4W light bulb.

I get the impression that this works OK in practice, so I figure I'm doing the math wrong and/or over thinking it...

[Leo_Gnardo]

You can put your lamp or other PTC resistor between the amp and crossover, then your XO frequencies aren't much altered.

I get the impression that this works OK in practice, so I figure I'm doing the math wrong and/or over thinking it...

And that, too!

[elipsey]

So I did the three dollar solution -- soldered a pair of automotive lightbulbs in line on the wires between the crossover and tweeter.

Ironically, one of my tweeters already didn't work -- the coil is open. Fortunately it has a replacable diaphram. Anyway, just currious if the brown goo on the voice coil is ferrofluid or some other revolting substance that should not be in there.

[g1]

Looks like ferro fluid. You'll probably also find it in the gap.