Could a speaker cause a fire?

[Welsh Anorak]

Here's an odd one for you.
A normal mid-range Technics midi-system is in for repair. The STK output device was dry-jointed and the feed resistor to one half had blown. No big deal, of course, a new resistor and device and all is well. However the customer only brought in one speaker (genuine, part of the system) as the other, according to the fire brigade, had caused a fire in the other room.
We both think this isn't possible. The IC works on around 30v and is connected via the usual thin cable to the speaker. So if, as is probable, the dry joints saw the IC fail and briefly present 30v at the speaker connections, surely this wouldn't be enough to start a fire before the R22 fusible resistor failed? Or would it?
The speaker is apparently 'burnt to a crisp', but we've asked to see it out of curiosity.
There's no litgation or anything like that involved, just our interest.
We suspect a faulty lamp, candle or similar left on top of the speaker. What do you think?

[Nickthedentist]

I'm sure if you shoved mains up a speaker it could potentially ignite, but that's not what's happened here.

[Refugee]

Those STK things are awful and can oscillate to destruction at the slightest whiff of a drying out capacitor.
You can end up with a 40W medium wave radio transmitter pumping its output into the speaker.
They stopped making transistors that only worked on audio at about the time those STK things appeared and the film caps that prevent amplifiers destroying themselves if an electrolytic cap gets a bit of ESR were not yet being fitted.
Some owners of old transistor amps have an awful lot of trouble finding 2N3055s that don't oscillate.
Modern amplifiers need an HF cut off built into the NFB components as should the ones with those STK things in them.
Those STK things appeared just before the need for the HF cut off mod began to get discovered.
To us there might be some value in seeing if it was the tweeter or bass voice coil that melted and set fire to the cone. I would consider the tweeter to be where the fire started.

If we could get full data sheets for those STK things the mod could be retrofitted subject to the individual internal transistor dies having there RF bandwidth quoted.
Modern amp chips either have the comp caps inside or there is a pair of pins to attach a cap.

[AdrianH]

I would have thought that if 30 Volts went down cable to a coil if say 8 ohms static 112 Watts (OK some lost in the wires) would be enough to heat the coil enough to ignite a paper cone, possibly before fusible resistor blew?

No I have no idea of the circuit in the amp.

Adrian

[paulsherwin]

I can't see how a fault condition could cause an actual fire in the speaker, as opposed to a bit of sizzling and smelly smoke. Stranger things have happened though. It will be interesting to see what you discover.

[mhennessy]

In a very general sense, yes, a speaker could emit a flame because of an overheated voice coil - I've seen it happen. But as all the materials used in the construction of a speaker should be self-extinguishing or treated to resist combustion, the flame should be short-lived. The potential problem comes from combustibles like paper being left in close enough proximity to catch the flame.

Older speakers might not benefit from modern regulations, so there might be concerns about any damping foam or similar. Others will know more about the relevant regulations...

Yes, a DC fault puts a lot more energy into a voice coil. As has been said, over 100 watts - into something designed to cope with about 25 watts of music (and the average value of that might only be a couple of watts in practice).

Some pictures - first 4 are a PMC that was somewhat over-driven, and the 5th is a Harbeth that was connected to a HH VX150 which "went DC". Obviously this Technics is less powerful than either amp in these examples, but then these speakers will have a higher power rating than the Technics one. The question here is time; how long did the fault condition persist before the fusible resistor opened (if it did indeed interrupt the current to the speaker). Even if the remains of the speaker were soaked by the fire brigade, it should still be possible to inspect the woofer's voice coil to see if it was the source of the fire.

[Lloyd 1985]

I bought a Goodmans amp that had a faulty STK in it back when I was at college, and that was shoving 40V out the left speaker terminal, didn’t seem to be any protection in that amp at all (not surprised really, it was a Goodmans after all!!). When I bought it the guy in the shop said he ruined 2 sets of speakers with it before he realised it was faulty, and that on one set of speakers he could see the coil arcing as it burnt out, the amplifier still happily playing merrily out of the right speaker!

It’ll certainly be interesting to see the burnt remains!

Regards
Lloyd

[Sideband]

Originally Posted by mhennessy

In a very general sense, yes, a speaker could emit a flame because of an overheated voice coil - I've seen it happen. But as all the materials used in the construction of a speaker should be self-extinguishing or treated to resist combustion, the flame should be short-lived. The potential problem comes from combustibles like paper being left in close enough proximity to catch the flame.

We've tested powered speakers for audio manufacturers and the front cloth should be self extinguishing. We test them by placing small pieces of firelighter on the cloth (speaker unit laying on its back) and just light the piece of firelighter. The cloth is only allowed to ignite briefly at the source of flame and it should extinguish within a few centimetres of the source. The test is to simulate a cigarette burn.

The test also applies to non-powered speakers.

Of course it depends if the manufacturer has had the speakers (or equipment) tested in the first place.

[G6Tanuki]

I wonder how much ultrasonic/RF power could be fed to the speaker if the amplifier chip went into oscillation? Ultrasonic energy wouldn't move the coil/cone so there would be no air-currents to help dissipate the heat; you'd essentially have a RF induction-heater.

[Sideband]

Originally Posted by Nickthedentist

I'm sure if you shoved mains up a speaker it could potentially ignite, but that's not what's happened here.

Ha ha.....I've done that...in my very early days when an extension speaker was fitted with a 5 amp two pin plug.....you can guess what happened! In our old flat we had two pin sockets (back in the 60's). Apart from a brief flash from the plug and a grumbling pop from the speaker, nothing untoward happened. I don't think it even took the house fuses out. It was only a relatively small speaker though. Still didn't do it much good.....

[Refugee]

The presence of the fusible resistor makes me think that the amp oscillated keeping the current just below the level that would open it.
If the fire started in the tweeter I would go for oscillation.

[John10b]

Glyn can you clarify please, in your post you say the stk amp is connected to the speaker by a thin cable, is the stk amp enclosed in the speaker cabinet ? Thanks.
John

[Roger Ramjet]

Originally Posted by Nickthedentist

I'm sure if you shoved mains up a speaker it could potentially ignite, but that's not what's happened here.

When I was a lad I connected up speaker to the mains via a flex & old fashioned BC plug into a lamp holder. When I switched on there was an almighty bang but thankfully no fire.

The things you do when you are young ! Rog

[kevinaston1]

I have seen loudspeakers which have been on fire; but these were on very high powered pro audio systems, usually rental systems.

Typical scenario was to turn every control to 13, and let it rip. Between 10 minutes and a coupe of hours later, smoke and flames from the loudspeaker and often the amp; but they would continue to thrash the system until it went silent.

Invariably, these would be returned to us under a "warranty" claim.

[pmmunro]

Originally Posted by Roger Ramjet

When I was a lad I connected up speaker to the mains via a flex & old fashioned BC plug into a lamp holder. When I switched on there was an almighty bang but thankfully no fire.

Rog

At the primary school I was at in the early 1950s, there was a baffle board radio, probably a Murphy A122 or similar on a trolley which was wheeled around classroms by the janitor. It was plugged into an extension speaker system which used 2-pin 5A plugs and sockets to allow the sound to be heard in more than one room.

Even at 8 years old, this looked very dangerous to me so I pointed it out to one of the teachers, even then not expecting a very favourable response. Fortunately, I picked one of the more begnign ones but she still dismissed my concerns, telling me "I'm sure they know what they're doing"!

Later, we were visited by a careers advisor or similar and I was asked what I liked doing in my spare time. I told him I liked playing with a 4.5V (126) battery, a bell and some other electrical bits and pieces which my father had set up for me. "Oh no, you can't do that - far too dangerous"!

PMM

[m0cemdave]

Originally Posted by kevinaston1

I have seen loudspeakers which have been on fire; but these were on very high powered pro audio systems, usually rental systems.

Me too. I used to work for a loudspeaker systems manufaturer.

The fire in this one started with a frayed tinsel lead-in wire and spread to the paper cone. Fortunately it didn't go any further.

But I have heard tales of sustained fires in loudspeakers in the 1970's when there was less appreciation (and regulation) of the flammability of wadding materials.

[SeanStevens]

I've heard (from a radio ham) that living near somebody who transmits at high power can do this sort of thing.

I'm guessing the coil would need to be at a length resonant at the transmission frequency.

Could this be the case?

SEAN

[lesmw0sec]

Originally Posted by SeanStevens

I've heard (from a radio ham) that living near somebody who transmits at high power can do this sort of thing.

I'm guessing the coil would need to be at a length resonant at the transmission frequency.

Could this be the case?

SEAN

Highly unlikely, although it is common for RF transmissions to break through into audio equipment and cause interference. I've tested RF kit to 1.5KW (well over the legal limit) in my shack, without any damage to anything. Even the wired internet was unaffected.

[cmjones01]

Originally Posted by G6Tanuki

I wonder how much ultrasonic/RF power could be fed to the speaker if the amplifier chip went into oscillation? Ultrasonic energy wouldn't move the coil/cone so there would be no air-currents to help dissipate the heat; you'd essentially have a RF induction-heater.

This is indeed possible. A friend of mine who works in the professional audio industry, developing digital audio processing software, told me about something that happened in their lab. They had a pair of Wharfedale Active Diamond speakers hooked up to some test rig or other and were puzzled as to why no sound was coming out. After a few seconds, there was a loud "clonk" from one of the speakers and a smell of burning, so they switched off. Upon closer inspection it turns out that a bug in the software had generated a full-amplitude ultrasonic signal, which the speakers obediently amplified and sent to the tweeter. It was inaudible to humans but the heat generated in the tweeter was enough to melt the glue that held the magnet on. The "clonk" was the sound of the magnet falling off and hitting the woofer below it!

I happen to have a pair of Active Diamonds here which appeared to have suffered a similar fate when I got them, also from a (different) lab environment. One of the tweeter magnets had fallen off and the voice coil left behind was open circuit. I replaced the tweeter with a random one from the scrap pile and still have the speakers, which served in my office for many years.

Based on these anecdotes it's not hard to imagine that instability in a power amplifier could melt or burn things inside the speaker, but I'd hope that the materials used were at least self-extinguishing.

Chris

[russell_w_b]

Originally Posted by Sideband

We've tested powered speakers for audio manufacturers and the front cloth should be self extinguishing.

What if there's years of accumulated dust and fluff in there?