I had an idea. It seems so obvious to me that I would be surprised if no one has done it before (assuming it actually works) but I didn't find anything with google.com.
How about this:
1. Use a separate amplifier for the subwoofer, with an active crossover/lowpass filter. (No, that's not the idea, just a requirement for it to work.)
2. Add a passive two-way crossover to the sub as if there were going to be a speaker for higher frequencies. The crossover point would probably be higher than the active crossover for the amplifier, but at least not less than it. (I'm guessing. This would require experimentation.)
3. Add a DUMMY LOAD to absorb high frequencies, of the proper resistance. (Maybe some power resistors on a heatsink?)
The theory being, that if you push the amplifier into clipping, it introduces high frequencies into the signal being sent to the sub. The passive crossover routes the highs to the dummy load. This "rounds off" the signal the sub sees, so it doesn't pause as long at its maximum excursion.
Does that sound like it would work? Has it been done before?
How about this:
1. Use a separate amplifier for the subwoofer, with an active crossover/lowpass filter. (No, that's not the idea, just a requirement for it to work.)
2. Add a passive two-way crossover to the sub as if there were going to be a speaker for higher frequencies. The crossover point would probably be higher than the active crossover for the amplifier, but at least not less than it. (I'm guessing. This would require experimentation.)
3. Add a DUMMY LOAD to absorb high frequencies, of the proper resistance. (Maybe some power resistors on a heatsink?)
The theory being, that if you push the amplifier into clipping, it introduces high frequencies into the signal being sent to the sub. The passive crossover routes the highs to the dummy load. This "rounds off" the signal the sub sees, so it doesn't pause as long at its maximum excursion.
Does that sound like it would work? Has it been done before?
The theory being, that if you push the amplifier into clipping, it introduces high frequencies into the signal being sent to the sub.
Is this what is causing the woofer to blow up or is it something
else, perhaps exceeding thermal or mechanical ratings of
the woofer ?
Would your woofer still be subject to high voltage because
it's wired across the amplifier ?
Is this what is causing the woofer to blow up or is it something
else, perhaps exceeding thermal or mechanical ratings of
the woofer ?
Would your woofer still be subject to high voltage because
it's wired across the amplifier ?
thylantyr said:
From what I understand, a clipped signal causes the subwoofer to pause at the peak for longer than it was designed to withstand, and this causes a heat buildup, because the coil is supposed to be cooled by motion. So, yes, I guess that exceeds its thermal rating.
As far as exceeding its mechanical rating, that could happen even without clipping if the amplifier is powerful enough.
thylantyr said:
I'm not sure what you mean by that.
Datoyminaytah said:
That's nonsense.
As to the original question, yes, a filter set-up as you describe will keep some of the clipped signal from getting to the woofer and clipping won't sound so bad.
Bill Fitzpatrick said:
OK, so I've been told nonsense. 🙂 I'm still trying to figure it out, but there's a lot of nonsense going around.
No, it isn't. If a driver is going to faithfully reproduce a true square wave then the cone is going to 'sit' at the end of its travel range for a bit before returning to center, but that action is not likely to hurt the driver half as much as the resulting sound quality will hurt the listeners ears.
While the posters scenario does account for the killing a lot of tweeters with small voice coils and limited XMax it propbably doesn't blow 1 in a thousand woofers that die; woofer failure is usually purely thermal in nature, toasted voice coils in particular, resulting from being fed a lot of power below their passbands.
No, it isn't. If a driver is going to faithfully reproduce a true square wave then the cone is going to 'sit' at the end of its travel range for a bit before returning to center, but that action is not likely to hurt the driver half as much as the resulting sound quality will hurt the listeners ears.
While the posters scenario does account for the killing a lot of tweeters with small voice coils and limited XMax it propbably doesn't blow 1 in a thousand woofers that die; woofer failure is usually purely thermal in nature, toasted voice coils in particular, resulting from being fed a lot of power below their passbands.
Speaker Failure
Contrary to what many of you feel, I design powered speakers for a living and an intimately familiar with real world failure modes of speakers.
It is true that clipping can kill tweeters by generating far more high frequency energy in a signal than was there originally. However, woofers have a natural HF limiting action and woofers are not killed by clipping.
If you look at the impedance vs frequency curve for a bass driver you can see that it peaks at the resonant frequency, then as you go up in frequency the impedance drops back to nominal for a bit and then as you still go further up in frequency, the inductance of the voice coil takes over and the impedance rises with frequency, reducing the amount of energy the voice coil needs to dissipate. Esentially, the crossover the person is talking about is already there.
When we rip open blown speakers, the speaker is damaged by one of two things. Over excursion or thermal failure.
Over excursion is from slaming the voice coil past the ends of save travel. Thermal failure is from too much energy being put into the voice coil for too long a period of time. Over excursion is the biggest killer of poorly made speakers, ones that are not glued together properly, specially a lot of Chinese made speakers.
In our powered professional speakers we do not get either of these failures. We only see them in out unpowered speakers where others set them up with other amplifiers.
To prevent thermal failures, we use a long time constant limiter. This limiter does nothing unless the set point power is exceeded for several seconds. Bsically we use a limiter with an attack time of several seconds. This allows the ful dynamic range of the music to pass unaffected. The limiter only cuts in it the set point is exceeded for several seconds. It never cuts in with only mild overdrive.
The second thing we do is excutsion control. We found that when driven at the speaker's resonant frequency is when the voice coil is most likely to hit the end stops. At other frequencies this id not a problem. So what we do is have an EQ filter in the circuit that is set to the speaker's resonant frequency with a Q of about 2 and about 4 db of cut. This stays out of circuit and is engaged through a resistive opto coupler when the power level reaches a level about 4 to 6 db lower than the overall limiting threshold.
With these measures in our powered speakers, we have stopped losing drivers. We use this in our flagship product which puts out 1 Kilowatt into the LF. We also use excursion control on our MF drivers which get 1/2 kilowatt of power.
And I know some theoretical purists are going to insist that clipping kills bass drivers too. Yes, I can make up theories too. In the real world, clipping can kill HF drivers. Mids and bass drivers are killed by using too much power and the lack of excursion control at the resonant frequency. You'll know this if you've seen and analyzed as many blown drivers as I have.
If you have any speakers that have come apart or experienced other mechanical failure, there is an excellent document on the Loctite web site in the proper adhesives to use in assembling speakers. The man who wrote it is ne of the world's top experts in the field of speaker failure (and a good personal friend).
Contrary to what many of you feel, I design powered speakers for a living and an intimately familiar with real world failure modes of speakers.
It is true that clipping can kill tweeters by generating far more high frequency energy in a signal than was there originally. However, woofers have a natural HF limiting action and woofers are not killed by clipping.
If you look at the impedance vs frequency curve for a bass driver you can see that it peaks at the resonant frequency, then as you go up in frequency the impedance drops back to nominal for a bit and then as you still go further up in frequency, the inductance of the voice coil takes over and the impedance rises with frequency, reducing the amount of energy the voice coil needs to dissipate. Esentially, the crossover the person is talking about is already there.
When we rip open blown speakers, the speaker is damaged by one of two things. Over excursion or thermal failure.
Over excursion is from slaming the voice coil past the ends of save travel. Thermal failure is from too much energy being put into the voice coil for too long a period of time. Over excursion is the biggest killer of poorly made speakers, ones that are not glued together properly, specially a lot of Chinese made speakers.
In our powered professional speakers we do not get either of these failures. We only see them in out unpowered speakers where others set them up with other amplifiers.
To prevent thermal failures, we use a long time constant limiter. This limiter does nothing unless the set point power is exceeded for several seconds. Bsically we use a limiter with an attack time of several seconds. This allows the ful dynamic range of the music to pass unaffected. The limiter only cuts in it the set point is exceeded for several seconds. It never cuts in with only mild overdrive.
The second thing we do is excutsion control. We found that when driven at the speaker's resonant frequency is when the voice coil is most likely to hit the end stops. At other frequencies this id not a problem. So what we do is have an EQ filter in the circuit that is set to the speaker's resonant frequency with a Q of about 2 and about 4 db of cut. This stays out of circuit and is engaged through a resistive opto coupler when the power level reaches a level about 4 to 6 db lower than the overall limiting threshold.
With these measures in our powered speakers, we have stopped losing drivers. We use this in our flagship product which puts out 1 Kilowatt into the LF. We also use excursion control on our MF drivers which get 1/2 kilowatt of power.
And I know some theoretical purists are going to insist that clipping kills bass drivers too. Yes, I can make up theories too. In the real world, clipping can kill HF drivers. Mids and bass drivers are killed by using too much power and the lack of excursion control at the resonant frequency. You'll know this if you've seen and analyzed as many blown drivers as I have.
If you have any speakers that have come apart or experienced other mechanical failure, there is an excellent document on the Loctite web site in the proper adhesives to use in assembling speakers. The man who wrote it is ne of the world's top experts in the field of speaker failure (and a good personal friend).
Dan: I'm building a subwoofer at this time which is composed of a Dayton 15" DVC driver in a 3 ft^3 sealed box. I also plan to implement a Linkwitz Transformer which would give about +9dB at 25Hz sloping from about +1dB at 50Hz. I plan to drive each voicecoil with an independant amplifier that output about 120W at 8ohm (8ohm is also the impedance of each voicecoil). Am I taking a risk of exceeding this drivers maximum excursion or overheating it?
Dayton Speaker
The power level you are proposing is within the limits of the speaker's capacity and you should not see either a thermal or excursion failure. As long as you don't insamely drive the crap out of the amplifier. The power level proposed, 120W per voice coil is about 180W when the amplifier is driven into hard asquare wave, a rare condition. This is the maximum capacity of the speaker, 350W. I say go for it.
Dual voice coils on a dual voice coil speaker are meant to be driven in parallel from one channel of an amplifier. Driving them off separate channels is likely to be just fine though. In theory, phase differences between the channels may cause odd effects but in reality, these are generally of no importance below 1000 Hz or so, well away from the pass band you will be using.
Overall, this seems like a setup that should be fine for you.
The power level you are proposing is within the limits of the speaker's capacity and you should not see either a thermal or excursion failure. As long as you don't insamely drive the crap out of the amplifier. The power level proposed, 120W per voice coil is about 180W when the amplifier is driven into hard asquare wave, a rare condition. This is the maximum capacity of the speaker, 350W. I say go for it.
Dual voice coils on a dual voice coil speaker are meant to be driven in parallel from one channel of an amplifier. Driving them off separate channels is likely to be just fine though. In theory, phase differences between the channels may cause odd effects but in reality, these are generally of no importance below 1000 Hz or so, well away from the pass band you will be using.
Overall, this seems like a setup that should be fine for you.
- Status
- Not open for further replies.