Is adding say a few caps before a full-range driver a definite loss in the revered "purity" of full-range drivers which normally are left unfiltered?
Obviously, not having passive coimponents is technically always beneficial for preservation of signal, but let's say I use quality broadbands in a speaker with a separate bass enclosure w/ bass drivers crossed 1st order at about 200Hz and don't want to over-exert the broadbands unnecessarily below 200Hz, is much of the puristic quality of the broadband now lost?
Obviously this would depend on the quality of the crossover components used, I'm sure the best of components are nearly "invisible" but also expensive.
Let's say I use ClarityCap CSA range. Let's say the amplifier is revealing. A good quality broadband is used. Is it pointless to fret if filtering the broadband with a few such resonable quality caps could audibly detract from its natural peformance?
Obviously, not having passive coimponents is technically always beneficial for preservation of signal, but let's say I use quality broadbands in a speaker with a separate bass enclosure w/ bass drivers crossed 1st order at about 200Hz and don't want to over-exert the broadbands unnecessarily below 200Hz, is much of the puristic quality of the broadband now lost?
Obviously this would depend on the quality of the crossover components used, I'm sure the best of components are nearly "invisible" but also expensive.
Let's say I use ClarityCap CSA range. Let's say the amplifier is revealing. A good quality broadband is used. Is it pointless to fret if filtering the broadband with a few such resonable quality caps could audibly detract from its natural peformance?
Look at the distortion plot of most drivers and you’ll see that much of the distortion is in two places: the low frequencies and above the breakup point. Filtering out <200hz and using subs will likely improve sound quality. Forget what the fullrange purists say; FAST/WAW is the way to go!
If a speaker is pure but sounds bad, are you going to listen to tolerate it?
Go with whatever sounds better.
Go with whatever sounds better.
Hi Arenith,
If you relieve your wideband drivers from bass work and long excursions, they will sound better, so in my books yes to filtering with parts of high quality.
BUT a single cap will most likely not work. At the Fs region, interaction with the impedance spike will happen and the driver will actually emit more at Fs than without anything.
The single cap approach at best results in increased distortions and at worst in the risk of driver destruction at high levels. You will need at least a parallel LCR notch to equalize the impedance at Fs, or a 2nd order filter.
Most widerange drivers will also benefit from filtering of the rise towards uper mid and high frequencies, which is inherent to drivers with a strong motor and a light cone/coil (some control the rise with a very heavy cone and/or a weak motor, but that sounds ugly and makes no sense). The gained advantage in timbral balance far outweighs any negative impacts caused by passive elements of high quality. YMMV.
1st order filtering of a bass driver at 200Hz will most probably not work, due to impedance rise, baffle step and SPL rise towards higher frequencies caused by a strong motor (and again, I don´t like weak motors...). You will at least need to control the impedance.
Don´t trust any online calculators, they don´t work. Measure what you´re doing. Heading for first order acoustic behaviour is a good thing with the right drivers, but you will most probably end up with an electrical filter which is not purely first order. It´s physics.
All the best
Mattes
If you relieve your wideband drivers from bass work and long excursions, they will sound better, so in my books yes to filtering with parts of high quality.
BUT a single cap will most likely not work. At the Fs region, interaction with the impedance spike will happen and the driver will actually emit more at Fs than without anything.
The single cap approach at best results in increased distortions and at worst in the risk of driver destruction at high levels. You will need at least a parallel LCR notch to equalize the impedance at Fs, or a 2nd order filter.
Most widerange drivers will also benefit from filtering of the rise towards uper mid and high frequencies, which is inherent to drivers with a strong motor and a light cone/coil (some control the rise with a very heavy cone and/or a weak motor, but that sounds ugly and makes no sense). The gained advantage in timbral balance far outweighs any negative impacts caused by passive elements of high quality. YMMV.
1st order filtering of a bass driver at 200Hz will most probably not work, due to impedance rise, baffle step and SPL rise towards higher frequencies caused by a strong motor (and again, I don´t like weak motors...). You will at least need to control the impedance.
Don´t trust any online calculators, they don´t work. Measure what you´re doing. Heading for first order acoustic behaviour is a good thing with the right drivers, but you will most probably end up with an electrical filter which is not purely first order. It´s physics.
All the best
Mattes
Sounds like you are building a WAW (Woofer Assisted Wideband).
Crossing that low it is often cheaper, and easier, to biamp.
At those frequencies a passive is bumping into the LF resonance of each driver which makes the design, execution, and function much harder to achieve.
I am a fan of the PLLXO as mentioned by Rayma, especially if your loudspeaker is happy with a first order XO.
dave
Crossing that low it is often cheaper, and easier, to biamp.
At those frequencies a passive is bumping into the LF resonance of each driver which makes the design, execution, and function much harder to achieve.
I am a fan of the PLLXO as mentioned by Rayma, especially if your loudspeaker is happy with a first order XO.
dave
These days I run my waw with a dsp even (Minidsp Flex) as these got that good that they are invisible soundwise. Even with vinyl records they are. And i use a lot of FIR and Biquad eq to tailor the sound to what i want. And active also gives the opportunity to use a tube amp for the fullrange, while using a class D for the woofer i use (where the high damping factor is an advantage). And it sounds a lot better now than before.
But before i used a passive crossover, and on most fullrange (even if it's a one driver system) i use filtering with caps and sometimes also coils in series with the drivers. Those filters make the system sound better for me.
But before i used a passive crossover, and on most fullrange (even if it's a one driver system) i use filtering with caps and sometimes also coils in series with the drivers. Those filters make the system sound better for me.
I can see lots of advantages of going down the WAW route, so I'm thinking of using two identical stereo amps, using one for the full rangers and the other for bass augmentation, and using line level filters between them. I'm considering housing them in one enclosure with the volume control and filter (possibly with a DAC) and another enclosure for the power supply transformers. Hopfuly, this will mean that any power supply droop, or clipping distortion caused by low frequencies will have no effect on the other amp, hence the rest of the frequency range.
If the amps are identical best to use one per channel, on ecannel for bottom and one for top. If it has a common PS you will get better use out of it.
dave
dave
Just a note on crossovers: even a 12dB/octave filter may be considered insufficient, as below crossover, excursion remains constant, until resonance, below which excursion (finally) rapidly diminishes.
My next-next build (after my 12+3 WAW is finally completed) is likely to be a 5" full-range. By then I should've also made some decisions on the amplifier topology, which should help narrow down the filtering scheme.
See, if the output resistance is high, this should help reduce inductance modulation, which could be particularly good for a wide-band driver's top end. However, it may also need some padding, which may be difficult to do passively without bringing the impedance down -- catch 22. So the preferred option is active filtering.
On the other hand, a voltage amplifier in a push-pull or bootstrapped follower kind-of configuration is a safer bet with lower risk of instability or needing extensive debugging or design changes. Then I would add a small inductor (say 500uH-1mH) for demodulation, and add a little treble boost actively. Two different ways to achieve similar results.
See, if the output resistance is high, this should help reduce inductance modulation, which could be particularly good for a wide-band driver's top end. However, it may also need some padding, which may be difficult to do passively without bringing the impedance down -- catch 22. So the preferred option is active filtering.
On the other hand, a voltage amplifier in a push-pull or bootstrapped follower kind-of configuration is a safer bet with lower risk of instability or needing extensive debugging or design changes. Then I would add a small inductor (say 500uH-1mH) for demodulation, and add a little treble boost actively. Two different ways to achieve similar results.