Tweeter damage - contact-coil connection ribbon broken - looking for a root cause

[raphsieniu]

Hello everyone!

I tried to find any post on the forum discussing similar issue but with no success.

I am currently analyzing the root cause of damaged contact-to-coil connection in a treble speaker (link to the speaker: https://stx.pl/en/d-9-500-8-ti.html), which is a part of industrial acoustic signaller. The speaker is being driven with a 1kHz sinusoidal signal with half of its rated power. There is no crossover in the circuit - just an LC LPF. That might seem clear - the speaker is driven with too low frequency thus the damage but I look for the answer on why that really happens - only trace I found in the Internet is that such conditions may lead to excessive diaphragm movement.

Is anyone able to link some sources describing this phenomenon or know any other possible explanation?
If such thing was discussed before please direct me to the topic.

Thanks in advance,
Regards
Rafal

 

[Roland68]

Hello everyone!

I tried to find any post on the forum discussing similar issue but with no success.

I am currently analyzing the root cause of damaged contact-to-coil connection in a treble speaker (link to the speaker: https://stx.pl/en/d-9-500-8-ti.html), which is a part of industrial acoustic signaller. The speaker is being driven with a 1kHz sinusoidal signal with half of its rated power. There is no crossover in the circuit - just an LC LPF. That might seem clear - the speaker is driven with too low frequency thus the damage but I look for the answer on why that really happens - only trace I found in the Internet is that such conditions may lead to excessive diaphragm movement.

Is anyone able to link some sources describing this phenomenon or know any other possible explanation?
If such thing was discussed before please direct me to the topic.

Thanks in advance,
Regards
Rafal

Recommended minimum crossover frequency: 1900 Hz / 18 dB / octave is a very clear statement.

Even at 6 dB, this frequency would shift significantly upwards, so you've deliberately murdered him.

The coil consists of a thin solid-core wire that moves back and forth at the corresponding frequency. Without an appropriate crossover, the movement becomes too large.

What happens if you repeatedly bend a solid-core wire? Exactly, it breaks.

 

[dfuller]

Recommended minimum crossover frequency: 1900 Hz / 18 dB / octave is a very clear statement.

Even at 6 dB, this frequency would shift significantly upwards, so you've deliberately murdered him.

The coil consists of a thin solid-core wire that moves back and forth at the corresponding frequency. Without an appropriate crossover, the movement becomes too large.

What happens if you repeatedly bend a solid-core wire? Exactly, it breaks.

flexible tinsel wire is used to connect the voice coil to the connectors, so not super likely it's that. More likely 1khz at half power may well have overheated and killed the voice coil.

 

[Roland68]

flexible tinsel wire is used to connect the voice coil to the connectors, so not super likely it's that. More likely 1khz at half power may well have overheated and killed the voice coil.

Most tweeters use the enameled wire used to wind the voice coil, i.e., simply the excess of the winding.
I've repaired more than a hundred such damaged tweeters under a magnifying glass or microscope.

For midrange drivers (except domes) and woofers, a flexible connector is used, of course, but even here, the wire of the oscillating coil can break at the connection point due to mechanical deflection. I recently had this happen again with a high-quality woofer.

Of course, it could also be burned through, but you can tell by the discoloration, and this usually happens within the coil.

 

[raphsieniu]

flexible tinsel wire is used to connect the voice coil to the connectors, so not super likely it's that. More likely 1khz at half power may well have overheated and killed the voice coil.

The coil itself is alright - confirmed by measurement.

Recommended minimum crossover frequency: 1900 Hz / 18 dB / octave is a very clear statement.

Even at 6 dB, this frequency would shift significantly upwards, so you've deliberately murdered him.

The coil consists of a thin solid-core wire that moves back and forth at the corresponding frequency. Without an appropriate crossover, the movement becomes too large.

What happens if you repeatedly bend a solid-core wire? Exactly, it breaks.

And you don't refer to resonance frequency, right? Do you have any source describing the mechanism? It seems to be the case you point out - if I see it correctly the thin copper ribbon connecting the coil is broken in the point where it is bent - see attached picture.

Thank you for your replies!

 

[JaapD]

You’re mentioning a LPF, which sounds strange to me. I’d expect a HPF with a cross-over frequency of let’s say 250-500Hz (1-2 octaves distance from working frequency). Since a HPF consists of a series capacitor this has the benefit of excluding any DC to the tweeter. Speaking of this, in finding the root cause can you exclude any DC on the amplifier’s output of this industrial design?

 

[DVDdoug]

with half of its rated power.

What power rating is that? The specs are all over the place! The AES power rating is the most meaningful but it does NOT mean it can handle 50W continuously.

How long did the speaker last?

Speaker ratings are tricky because they are for program material peaks. That is, a 100W amp that's not clipping should be safe with a 100W speaker. If you push the amp into clipping, the peaks are clipped/limited but the average power goes-up and you can burn-out the speaker.

Similarly, you can also burn-out a 100W speaker with a continuous 100W tone.

And... If you clip a sine wave the power goes up. If you clip it so badly that it becomes a square wave you get double the power (with the extra power in harmonics).

And tweeters can't handle a much power as the woofer (or midrange). With normal program material most of the energy is in the lower frequencies.

Speaker power ratings are "statistical" and of course not all program material is the same. And it doesn't help that a lot of speaker companies are not completely honest.

This driver has an AES rating of 50W with a 1900Hz crossover. That means you can put it into a speaker system with MOST of that 50W going to the woofer/midrange. The tweeter cant handle 50W by itself. And, in a "normal" speaker system with "normal" program material, you can give it a higher power rating if you specify a higher crossover frequency!

I don't have the AES paper but this should be similar.

just an LC LPF. That might seem clear - the speaker is driven with too low frequency

You mean HPF (high-pass filter). That's the high-frequency half of a crossover. It's good if you are feeding it music. Of course you don't need crossover or filter if you are feeding it 1kHz and if there is no woofer. If that's a 1900Hz high-pass, you're loosing some signal but you're frying the speaker so the last thing you want is more power!

 

[Mnyb]

Stupid question this application does it include the horn ? It’s a compression driver and supposed to be coupled to a horn to work properly.

Horns or waveguides change the acoustical impedance ? and driver movement .

Suppose that’s why there is the 18dB per octave requirement the horn helps , but outside its range it does not work . The corresponding horns you’re supposed to use also has a frequency range.

You better ask one the horn speakers design guys here at ASR as I’m not really qualified .

Using anything in a way it’s not intended to be used may require a lot of thinking and understanding how it really works in its intended use case to then be able to bend the rules a bit .

 

[voodooless]

This driver has an AES rating of 50W with a 1900Hz crossover. That means you can put it into a speaker system with MOST of that 50W going to the woofer/midrange. The tweeter cant handle 50W by itself. And, in a "normal" speaker system with "normal" program material, you can give it a higher power rating if you specify a higher crossover frequency!

I don't have the AES paper but this should be similar.

Are you sure? From what I’ve seen the power rating is for the band-limited signal, not the pre-filtered source.

It also includes a 6dB crest factor, so contains power peaks of 4x, so 200W in this case.

Nevertheless, pink noise is rather different from a continuous sine wave way below the operating area of the driver.

 

[Roland68]

The coil itself is alright - confirmed by measurement.


And you don't refer to resonance frequency, right? Do you have any source describing the mechanism? It seems to be the case you point out - if I see it correctly the thin copper ribbon connecting the coil is broken in the point where it is bent - see attached picture.

Thank you for your replies!

This indicates a typical fatigue fracture, especially at the bend where the material is already weakened.

The resonant frequency is irrelevant; it's only used for parameterization and calculations.
The specified 1900 Hz at 18 dB clearly tells you the limit, not a recommendation. Normally, you keep a minimum distance of at least a few hundred Hz from that, and at 12 or 6 dB, much more.

What kind of source do you want for the mechanism?
It's a coil that oscillates back and forth with two connecting wires. This is the case with almost all loudspeaker drivers, but also in many places in industry, basically anywhere you have moving coils.
Try searching on YouTube for "slow motion recording voice coil" or just "voice coil."

 

[Arindal]

it could also be burned through, but you can tell by the discoloration, and this usually happens within the coil.

And you usually notice smell of burnt voicecoil or voicemail former, the moment you open the compression driver unit.

For many compression drivers, the voicecoil unit can be bought for little money as a spare part. Wonder, if this is the case here?

 

[Roland68]

And you usually notice smell of burnt voicecoil or voicemail former, the moment you open the compression driver unit.

For many compression drivers, the voicecoil unit can be bought for little money as a spare part. Wonder, if this is the case here?

Probably not, since the whole unit only costs €29.

 

[voodooless]

Probably not, since the whole unit only costs €29.

Yes, there is a refine kit:

Recone kit RD.9.500.8.TI

Recone kit RD.9.500.8.TI Technical data: Recone kit D-500-8-Ti-S ; D.9.500.8.TI ; CD.STX.1.1,5.500.8.F.T.TI ; CD.STX.1.1,5.500.8.F.B.TI Diaphragm material - titanium

stx.pl

Under € 12

 

[Roland68]

Yes, there is a refine kit:

Recone kit RD.9.500.8.TI

Recone kit RD.9.500.8.TI Technical data: Recone kit D-500-8-Ti-S ; D.9.500.8.TI ; CD.STX.1.1,5.500.8.F.T.TI ; CD.STX.1.1,5.500.8.F.B.TI Diaphragm material - titanium

stx.pl

Under € 12

It's not worth repairing.
I couldn't find it despite a quick search.
Actually, something like this belongs to the article, or should be found with the article in a Google search.

 

[voodooless]

It's not worth repairing.

It’s probably not worth buying the whole damn thing in the first place Definitely not for this use case.

I couldn't find it despite a quick search.

Just put in the model number in the website search, and it shows up immediately. Took me less than a minute to find

Actually, something like this belongs to the article

Yes, I agree, but with things in this budget range, expectations are low. I was also surprised to find the replacement part