heat in loudspeakers

[Deleted member 21219]

Good evening. I am not an engineer or scientist, and I have 2 questions for those people here who are engineers or scientists. 1) in loudspeakers that are relatively insensitive, needing more-than-average power, how hot can the voice coil get? 2) in loudspeakers that are relatively insensitive, needing more-than average power, how hot do some crossovers get?
Thank you for your time and consideration. Jim Taylor

 

[mhardy6647]

Crossovers can get hot enough to immolate components therein.
VCs can, and do, "burn" open when stressed beyond their limits.

A scorched crossover (burned up power resistor)

source: https://forum.polkaudio.com/discussion/138722/lsi15-crossovers-burned-through

 

[Doodski]

As for high temps. I opened drivers to inspect the voice coils and found totally blackened carbon coils and others I've seen have been blued from the heat but not burnt. So I am guessing if the metal can be blued then that is near red hot. The same goes for crossovers being burnt and crispy when abused.

 

[Doodski]

The answer for #1 is yes although the air movement in a cabinet could exceed the flow around a driver mounted on a open baffle. The design/venting of the driver matters here.

The answer for #2 is a yes too although there are many peeps out there using too small of power amps and resultingly they have clipping issues.

 

[mhardy6647]

another toasted XO

https://electronics.stackexchange.c...ntifying-replacement-components-on-pcb/425457

and a scorched voice coil.

https://sound-au.com/articles/speaker-failure.html

So -- I am a high sensitivity loudspeaker kind of guy (my 'real' hifi amplifier is capable of about 3.5 watts per channel). A high-sensitivity loudspeaker will burn out just as readily -- in fact, more so, if over driven. Conversely, the XO parts in mine (at least) are not likely to be threatened by the output of a single-ended 2A3 amplifier.

DSC_0938 (3) by Mark Hardy, on Flickr

OB -- well, the speaker loading (or lack of same) puts a woofer or "full range" driver at considerable potential risk around (and below) Fs.

PS -- I assume you're not my next door (down the hill) neighbor, @Jim Taylor -- are you?

 

[Zvu]

Burned voice coil is a sign of abused driver. In people with normal hearing, distortion due to overload will be obvious enough. Voice coil can get 250C with no problem, it is all about the varnish used for wire insulation and glue that holds windings to the voice coil former.

When i use high power drivers, i use multiple resistors in parallel to get to the value i need. If i need 8 ohms, i'll use 4x32 ohm resistors in parallel. It will dissipate same amount of heat but the emiting surface will be 4 times single resistor and that means lower component temperature. That being said, i always start with single component value if i can and then i test it under hard conditions with iR thermometer. If the temperature is too high, i double them up and check again. I also don't use pcb's but a direct connection between components. I've had zero problems in last 20 years.

 

[Zvu]

......
The answer for #2 is a yes too although there are many peeps out there using too small of power amps and resultingly they have clipping issues.

Whole clip is great but look at 6:38

 

[tomtoo]

Low efficience drivers and high spl dont fit.
Thats easy to explain. Imagine a very low efficient driver 80db/w and a high efficient 100db/w. Now lets calculate.

Low efficient:
80db 1w
90db 10w
100db 100W
110db 1kw

High efficient:
80 db 0.01w
90db 0.1w
100db 1w
110db 10w
120db 100w
130db 1kw

You can easy see that low efficience means a lot of heat. Thats why PA speakers have to be highly efficient, even you could push with the amps enough power the speakers need to get ride of this heat.

Edit says: Maybe i should add nearly nothing of the power pushed into a speaker makes sound. No it's heat. So with a low effizient speaker you have a 1kw heater while you have at same sound level with a high effizient you only have a 10watt heater. And so its not hard to see thats much more easy to disperse 10w heat vs. 1kw.

 

[ctrl]

1) in loudspeakers that are relatively insensitive, needing more-than-average power, how hot can the voice coil get? 2) in loudspeakers that are relatively insensitive, needing more-than average power, how hot do some crossovers get?

1) This is difficult to predict in a generalized way. A good reference value is the "Rated power handling" of the chassis. This limit should not be exceeded permanently.
A voice coil can get as hot as you want - until the voice coil is destroyed. A voice coil heats up after only a few seconds.
It can take hours until the entire system consisting of the chassis, inner volume and speaker cabinet has warmed up evenly.

2) As others have already shown, very hot. The problem are usually undersized resistors (regarding their electrical power rating) in the midrange branch or impedance linearization with a passive series notch in the low frequency branch of the crossover.

 

[mhardy6647]

The... umm... fragility of crossover components is, to my way of thinking, one of a plethora of good reasons to think about implementing crossovers in the low level circuits rather than at the loudspeaker.


(although I haven't taken that plunge -- yet)

 

[texbychoice]

A properly designed crossover is not fragile. The pictures of failed crossover components are examples of cheap marginal designs using low quality components. There were likely abused by distortion from an overdriven amp.

It is not difficult to select quality components with adequate design margin to survive clean power producing ear splitting sound levels.

 

[FrantzM]

A bit OT as I am wont to

We often do not realize how inefficient our speakers are.
120 dB is about One watts acoustic ...
Taking the example above from post
High efficient:
80 db 0.01w
90db 0.1w
100db 1w
110db 10w
120db 100w
130db 1kw


A 100 db/1watt/1 meter, by all aspects a very high sensitivity speaker with a perfect crossover with no loss requires about 100 watts to output 1 watt acoustic . Most of the remaining power is loss to heat. This is a case for an inordinately efficient speaker 100 dB/w/m. Most speakers on the market are 90 dB or less efficient ...

Carry on...

 

[mhardy6647]

A bit OT as I am wont to

We often do not realize how inefficient our speakers are.
120 dB is about One watts acoustic ...
Taking the example above from post
High efficient:
80 db 0.01w
90db 0.1w
100db 1w
110db 10w
120db 100w
130db 1kw


A 100 db/1watt/1 meter, by all aspects a very high sensitivity speaker with a perfect crossover with no loss requires about 100 watts to output 1 watt acoustic . Most of the remaining power is loss to heat. This is a case for an inordinately efficient speaker 100 dB/w/m. Most speakers on the market are 90 dB or less efficient ...

Carry on...

I'd opine that you're absolutely on point, actually (and fwiw).
And my loudspeakers are on the order of 100 dB/watt @ 1 meter in terms of sensitivity -- because I do like dynamics.
High sensitivity loudspeakers are very much out of fashion in audiophile circles because amplifier power is cheap (for better or for worse).

 

[helom]

Out of curiosity, would the SEAS A26 speaker kit sidestep some of these problems? Jim

What problems specifically?

I think the important consideration with the A26 kit is the tweeters have a 1st-order crossover. This means only a 6db reduction for each lower octave. If you opt for the recommended 12 ohm resistors, the crossover frequency of the tweeters becomes ~2650Hz. As I understand crossovers, 1st-order implies that the tweeters will be playing only 6db lower at 1325Hz, and only 12db lower at 662.5Hz, which is quite close to the diaphragm’s free air resonance —not good. Seas’ recommended crossover point for this tweeter is >/=1500Hz, though I would guess that is intended for 2nd or 3rd order filters.

Luckily, the tweeters have a large diaphragm, large voice-coil and thick aluminum frame, so they should be able to handle more heat than the typical tweeter.

I own a pair of A26s. For my preferred balance I use the 15 ohm resistors. This puts the crossover frequency around 2200Hz IIRC. These speakers can play quite loud before causing listener fatigue, and as such, I often find myself listening at higher levels then I realize. That in turn has caused me to suddenly stress the drivers into audible distortion on a few occasions. I’m honestly not sure if it’s the tweeters or woofers that are distorting, but since I high-pass the speakers and cross them over to my subs at 160Hz, I suspect it’s most likely the tweeters distorting.

This isn’t to take away from the excellent value of the A26s. From about 70Hz on up, the pair I built (slightly altered from the standard kit) sound like speakers you’d buy for $6K/pair from a retailer.